A series of isoquinolin-1-ones and quinazolin-4-ones and related derivatives were prepared and evaluated for their ability to inhibit tumor necrosis factor alpha (TNFalpha) production in human peripheral blood monocytes stimulated with bacterial lipopolysaccharide (LPS). In an effort to optimize the TNFalpha inhibitory activity, a homologous series of N-alkanoic acid esters was prepared. Several electrophilic and nucleophilic substitutions were also carried out. Alkanoic acid esters of four carbons were found to be optimum for activity in both the isoquinoline and quinazoline series. Ring substituents such as fluoro, bromo, nitro, acetyl, and aminomethyl on the isoquinoline ring resulted in a significant loss of activity. Likewise, similar groups on the quinazoline ring also reduced inhibitory activity. However, the 6- and 7-aminoquinazoline derivatives, 75 and 76, were potent inhibitors, with IC(50) values in the TNFalpha in vitro assay of approximately 5 microM for each. An in vivo mouse model of pulmonary inflammation was then used to evaluate promising candidate compounds identified in the primary in vitro assay. Compound 75 was selected for further study in this inhalation model, and was found to reduce the level of TNFalpha in brochoalveolar lavage fluid of LPS-treated mice by about 50% that of control mice. Thus, compounds such as 75, which can effectively inhibit proinflammatory cytokines such as TNFalpha in clinically relevant animal models of inflammation and fibrosis, may have potential as new antiinflammatory agents. Finally, a quinazoline derivative suitable to serve as a photoaffinity radiolabeled compound was prepared to help identify the putative cellular target(s) for these TNFalpha inhibitors.

No abstract

The alternatively spliced type III connecting segment (IIICS) of fibronectin (Fn) contains an amino acid sequence, CS-1, which is recognized by the integrin receptor, alpha 4 beta 1. Plasma Fn inhibits alpha 4 beta 1-dependent binding of lymphocytes and monocytes to CS-1 containing Fn derivatives poorly, suggesting limited exposure of the CS-1 sequence in Fn. To test the availability of CS-1 in plasma Fn, an antibody was raised to the synthetic peptide CS-1. The CS-1 sequence was found to be minimally exposed in plasma Fn; and immobilization of Fn, a model of matrix deposition, caused only a modest increase in its exposure. Digestion of Fn with selected proteases, however, induced substantial expression of the CS-1 sequence. The acid protease cathepsin D generated fragments of 31-33.5 kDa from the COOH-terminal heparin-binding domain of Fn which possessed high immunoreactivity with anti-CS-1. Digestion of Fn with cathepsin B also resulted in the exposure of CS-1 sequence in a 140 kDa fragment. Although the digestion of Fn with neutral proteases (neutrophil elastase, cathepsin G, chymotrypsin, trypsin) generated fragments from the COOH-terminal heparin-binding domain of similar molecular weight as with cathepsin D, the exposure of CS-1 did not occur. Exposure of the CS-1 region by the cathepsins was supported by cell adhesion experiments; digestion of Fn with cathepsins D and B transformed inert plasma Fn to an effective inhibitor of adhesion of lymphoblastoid B and T cells (Ramos, Jurkat, Molt-4) to an immobilized CS-1 conjugate. These results suggest that exposure of the CS-1 sequence in plasma Fn by proteolysis with cathepsins D and B, enzymes implicated in several pathological processes, may serve a regulatory function in cell adhesion. The adhesive function of the CS-1 region in intact Fn appears to be suppressed by the native conformation of the molecule.