S(+)-4-(1-Phenylethylamino)quinazolines as inhibitors of human immunoglobulin E synthesis: potency is dictated by stereochemistry and atomic point charges at N-1.

Since the pathogenesis of allergic diseases is associated with elevated levels of immunoglobulin E (IgE), we developed a high throughput reporter gene assay in a human B-cell line to screen for low molecular weight IgE inhibitory compounds. Monitoring the IL-4 driven IgE-germline promoter activity (IgE-GLP), we discovered 4-(1-phenylethylamino)qinazolines as potent inhibitors of IgE-germline gene expression. Testing of the individual enantiomers (1, 2) revealed that only the S(+) enantiomer 1 was active. A cell viability assay done in the same cell line in parallel discriminated the dose-dependent inhibition from a general antiproliferative effect. The observed correlation of the inhibitory potencies found in the reporter gene assay with those measured by IgE-ELISA in primary human splenocytes provided evidence that the blockade of IgE synthesis is the direct consequence of IgE-germline gene inhibition, thereby validating the reporter gene assay. Parallel synthesis in solution rapidly provided a series of analogues of compound 1 with modifications in the phenethylamine side chain and the quinazoline core for SAR studies. Increasing the lipophilicity of the arylalkylamine moiety yielded S(+)-4-(1-(2-naphthyl)ethylamino)quinazoline (6) as the most potent inhibitor (IC(50) of 14 nM) while the R(-) enantiomer was again found to be inactive. Within the set of S enantiomers, quantum mechanical calculations revealed that the IgE inhibitory activity can be quantitatively described by the charge at N-1 of the heterocyclic core and to a lesser extent by the molar refractivity. These results demonstrate the importance of electron-deficient fused 4-aminopyrimidines and lipophilic side chains for biological activity. The strong preference for the S configuration of the phenethylamine side chain is remarkable insofar as biological activity for fused 4-(1-phenylethylamino)pyrimidines has been published for the R enantiomers only (EGFR tyrosine kinase inhibition).

Humanized anti-Lewis Y antibodies: in vitro properties and pharmacokinetics in rhesus monkeys.

ABL 364 is a murine monoclonal IgG3 antibody directed against the Lewis Y carbohydrate antigen (Le(y)) expressed on the surface of many epithelial cell tumors. The antibody mediates cytotoxicity via activation of human complement or human effector cells, and has been evaluated in several clinical trials including two Phase I/II trials in relapsed small cell lung cancer and metastatic breast cancer. To improve the effector functions of the antibody, increase its half-life in circulation, and avoid the human antimouse antibody response, two chimeric and several humanized antibodies were constructed for evaluation. The chimeric IgG1 is more potent than the murine IgG3 in tumor cell lysis via activation of human peripheral mononuclear cells (10-fold), but somewhat less effective in complement-dependent lysis (2-3 fold). The chimeric IgG3 is slightly less potent than the IgG1. A humanized IgG1 was constructed by combining the complementarity-determining regions of the ABL 364 antibody with human framework and constant regions. Several additional variants were subsequently constructed to improve the binding affinity and increase expression of the antibody. Two of the variants, designated I and K, differ by a single amino acid at position 75 of the heavy chain. Both variants have affinity within 2-fold of the chimeric IgG1 antibody and retain the cytolytic activities toward tumor cell lines. However, it was possible to express variant K at a significantly higher level (5- 10-fold) than variant I. Pharmacokinetics of the humanized ABL 364 antibody variant K was compared with that of the parent murine antibody in rhesus monkeys. It was shown that the terminal half-life of the humanized antibody in rhesus monkeys is 14-20 days, with a mean of 16.3 days, while that of the parent murine antibody is only 1.9 days.

Enzymatic synthesis and comparative biological evaluation of a phosphonate analogue of the lipid A precursor.

Phosphonate analogue 5 of the lipid A precursor 4 has been prepared from phosphonate 2 and nucleotide 3 with the help of lipid A synthase, isolated from the overproducing Escherichia coli mutant MC 1061 (delta 2512) or JB1104 (delta 2514). The biological properties of phosphonate 5 and phosphate 4 are quite similar to each other as compared in the limulus amoebocyte lysate assay, by the activation of the RAW264 murine macrophagelike cell line (determined by stimulation of ornithine decarboxylase), and by the pyrogenicity in rabbits. Hydrolytic removal of the 1-phosphate group of 4 is thus not a prerequisite for its biological activity.