Orally active trifluoromethyl ketone inhibitors of human leukocyte elastase

This paper describes the development a series of peptidyl trifluoromethyl ketone inhibitors of human leukocyte elastase which are found to have excellent pharmacological profiles. Methods have been developed that allow for the synthesis of these inhibitors in stereochemically pure form. Two of these compounds, 1k and 1l, have high levels of oral bioavailability in several species. Compound 1l has entered development as ZD8321 and is presently undergoing clinical evaluation. These compounds demonstrate that peptidyl trifluoromethyl ketone inhibitors can achieve high levels of oral activity and bioavailability, and therefore they may prove useful as therapeutic agents in the treatment of diseases in which elastase is implicated.

Discovery and biological activity of orally active peptidyl trifluoromethyl ketone inhibitors of human neutrophil elastase

Previously we had shown that tripeptidyl trifluoromethyl ketones (TFMKs) possessing an N-terminal diarylacylsulfonamide, such as ICI 200,880 and ICI 200,355, displayed unparalleled protection against the lung damage induced by human neutrophil elastase (HNE) when the inhibitors were administered intratracheally. Since the diarylacylsulfonamides were designed specifically to afford a long residence time in the lung, it was not unexpected that inhibitors from this class of TFMKs were not active when administered orally. Upon evaluating a large number of peptidyl TFMKs possessing a variety of N-terminal groups, several compounds were identified which demonstrated oral activity. Compounds were evaluated for their oral activity by measuring their ability to inhibit the increase in lung weight relative to body weight (Lw/Bw), the increase in red blood cells, and the increase in white blood cells induced by intratracheally administered HNE (100 micrograms/hamster). A number of tripeptidyl trifluoromethyl ketones containing neutral N-terminal groups displayed good oral activity, while those containing basic, acidic, or polar groups did not. Compound 50, possessing an N-terminal 4-(CH3O)C6H4CO group, was particularly effective, reducing Lw/Bw by 77%, red cells by 89%, and white cells by 91% when dosed at 37.5 mg/kg orally. Thus, by modifying the N-terminal group of tripeptidyl TFMKs, inhibitors can be designed which are effective in vivo when administered either orally or intratracheally.

Peptidyl alpha-ketoheterocyclic inhibitors of human neutrophil elastase. 2. Effect of varying the heterocyclic ring on in vitro potency

A series of peptidyl alpha-ketoheterocycles were synthesized and evaluated for their in vitro inhibition of human neutrophil elastase (HNE). Several heterocycles, including oxazoline and benzoxazole, afforded extremely potent inhibitors of HNE (1p-r) with nanomolar to subnanomolar Ki values. The structure-activity relationships revealed that for compounds with a Ki < 1000 nM potency tends to be positively correlated with the sigma I value of the heterocycle. Furthermore, the results in this study support the hypothesis that, in the covalent enzyme-inhibitor adduct, the azole nitrogen atom of the inhibitor heterocycle participates in a hydrogen-bonding interaction with the active-site His-57.

Nonpeptidic inhibitors of human leukocyte elastase. 2. Design, synthesis, and in vitro activity of a series of 3-amino-6-arylopyridin-2-one trifluoromethyl ketones

A series of potent nonpeptidic inhibitors of the enzyme human leukocyte elastase (HLE) is reported. These inhibitors contain a 3-amino-2-pyridone ring as a central template in which the pyridone carbonyl and 3-position NH group are thought to form important hydrogen bonding interactions with the Val-216 residue of HLE. Substitution of the 6-position of the pyridone ring by various alkyl and aryl groups was found to afford increases in the in vitro potency of these inhibitors. A 6-position phenyl group, compound 10f, was found to result in a large increase in binding affinity, which was not obtained when the phenyl group was placed in either the 4- or 5-position of the molecule. Compound 10f was found to have good selectivity for HLE over other proteolytic enzymes, with the exception of bovine pancreatic chymotrypsin (BPC). Substitution of the 6-phenyl group in these molecules was found to decrease binding affinity for BPC without adversely affecting affinity for HLE.

Protease inhibitors decrease rabbit cartilage degradation after meniscectomy

In vitro proteoglycan (PG) synthesis and release were measured on cartilage removed from rabbit knees within 1 week of meniscectomy. Three days following partial lateral meniscectomy, 72% of the femurs and 82% of the tibias had visible ulcers. Cartilage from the weight-bearing areas incorporated 2.0-2.9 times more 35S-sulfate in vitro than cartilage from the opposite, unoperated knees. 3H-thymidine incorporation was 2.5-3.4 times higher for surgical than control groups. 35S-sulfate incorporation by the surgical group was inhibited by 22% in the presence of 10(-4) M U24522, an inhibitor of rabbit chondrocyte metalloprotease (CMP). 3H-thymidine incorporation by the surgical group was inhibited by 28% by 10(-4) M U24522. In vitro PG release from cartilage removed 2 days after surgery was 1.6-3.7 times higher for the surgical than the control group. PG release by the surgical group after 22 h of incubation was reduced to the control level by three CMP inhibitors, U24278, U24279, and U24522. PG release by cartilage from the nonsurgical group was also reduced by these compounds at 22 h. These results suggest that both the anabolic and catabolic processes that are stimulated by surgery can be isolated in vitro and that CMP may be involved in the catabolic process.

Proteoglycan degradation by a chondrocyte metalloprotease. Effects of synthetic protease inhibitors

Synthetic inhibitors of a chondrocyte metalloprotease (CMP) were assessed for potency. Proteoglycan core protein was used as substrate. The IC50 values were between 2 X 10(-6) and 7 X 10(-6) M for two types of inhibitors, thiol tripeptides and N-carboxyalkyl peptides. Hydroxamic acid peptides were more potent, with IC50 values of 3.2 X 10(-8) to 6.0 X 10(-8) M. These results confirm inhibitory concentrations reported using a proteoglycan-polyacrylamide bead assay. The slopes of the dose-response curves for the thiol compounds were steeper than the slopes for the other two types of compounds. All of the culture media tested inhibited CMP to some extent. Some media also interfered with inhibitor activity. In Ham's F10 nutrient medium, minimum CMP inhibition occurred, and all four hydroxamic acid peptides retained their activity for 1-2 days at 37 degrees. One thiol peptide compound assayed lost activity in 1 hr in thiocyanate-treated serum. All four hydroxamic acid peptides assayed retained activity in thiocyanate-treated serum after 3 days at 37 degrees. The hydroxamic acid peptides may provide a way to block endogenous CMP activity in vivo and to assess the role of CMP in normal and experimentally altered cartilage. They are more potent than other known CMP inhibitors. They retain activity in culture media and serum conditions used for in vivo and in vitro tests of CMP activity and toxicity.

Effect of synthetic metalloprotease inhibitors on cartilage autolysis in vitro

The role of chondrocyte metalloprotease (CMP) in mediating cartilage autolysis was studied. Proteoglycan (PG) release and synthesis by rabbit articular cartilage explants were measured. After a 1-day preculture in control medium, 3.3 X 10(-6) M retinoic acid (RET) treatment for 1 day stimulated PG release several fold. RET also caused a large decrease in PG synthesis that returned to the control level after a 3-day recovery period. The effect on PG synthesis was observed at serum levels of 5 and 0.05%. The effect of RET on PG release required protein synthesis, inasmuch as it was lost in cultures maintained in media without amino acids or in a low volume of media. Interleukin-1 (IL-1) and lipopolysaccharide (LPS) treatment for 2 days also stimulated PG release. More PG was released after RET than after IL-1 or LPS, and only RET produced an effect that was evident by day 1. The amount of CMP that produced the same size effect on PG release as these stimulators was below the detection level of PG protease assays. Three potent CMP inhibitors reduced RET-, IL-1- and LPS-stimulated PG release to control levels. These inhibitors did not block another action of RET on chondrocytes, namely the inhibition of PG synthesis by RET immediately after treatment. The inhibitors did not act by reducing cell viability, because recovery of the rate of PG synthesis 3 days post-treatment occurred in inhibitor-treated cultures. These studies suggest that CMP is involved in cartilage autolysis that is stimulated by RET and IL-1.