Drug therapy for the major inflammatory skin diseases, which include atopic dermatitis, psoriasis and allergic contact dermatitis, is often inadequate due to poor efficacy, toxicity, or both. Much research has focused on the macrolactam T cell inhibitors as a promising new class of agents for immunotherapy, and medicinal chemistry efforts to design novel ascomycin analogs have produced clinically promising agents. A synthetic program to modify the ascomycin nucleus to alter its physicochemical properties and promote systemic clearance is described. A biologic screening strategy to identify analogs with reduced systemic activity and rapid pharmacokinetic elimination led to identification of the clinical candidate, ABT-281. A swine contact hypersensitivity model was used as a stringent indicator of skin penetration as human doses of topical corticosteroids produced inhibition only in the 50% range and ED50 values were 100-fold less potent than in rat. Also, cyclosporine was confirmed to be topically inactive in swine, as seen in human. ABT-281 had topical potency equal to tacrolimus (FK506) despite a severalfold lower potency for inhibiting swine T cells in vitro, consistent with superior skin penetration. ABT-281 was found to have a shorter duration of action after i.v. dosing in monkeys using an ex vivo whole blood IL-2 production assay. Systemic potency was reduced by 30-fold or more in rat popliteal lymph node hyperplasia and contact hypersensitivity assays. Following i.v. or i.p. administration in the swine contact hypersensitivity model, ABT-281 was 19- and 61-fold less potent, respectively, than FK506. Pharmacokinetic studies showed that ABT-281 had a shorter half life and higher rate of clearance than FK506 in all three species. The potent topical activity and reduced systemic exposure of ABT-281 may thus provide both efficacy and a greater margin of safety for topical therapy of skin diseases.

Ascomycin 2, a close analogue of the immunosuppressant FK506 1, was modified to incorporate a hydroxyl group at the C-33 position. This increased the aqueous solubility of ascomycin by a hundred-fold at pH 7.4 and by approximately 300-fold at pH 6.5. Ascomycin 3 also exhibited an excellent immunosuppressive activity in vitro, as tested in a human mixed lymphocyte proliferation (HuMLR) assay, and in vivo using a rat popliteal lymph node (rPLN) hyperplasia assay.

Comparing nephrotoxicity of numerous drug analogs is impractical with chronic in vivo models. We devised a new cisplatin potentiation assay (CISPA) that sensitively detects renal injury as a serum creatinine increase when only one dose of test compound is followed by cisplatin. Reference nephrotoxins known to act on various sites in kidney tubules, glomeruli or renal papilla were all detected by the CISPA at single doses that without cisplatin gave little change, which showed that this simple, sensitive assay has broad potential utility for mechanistic studies of nephrotoxicity. We used the CISPA both to probe the nephrotoxic mode of action of immunosuppressants and to search for safer compounds. Although several non-nephrotoxic immunosuppressants were inactive, cyclosporine, FK506, ascomycin (C21-ethyl-FK506) and rapamycin were nephrotoxic in the CISPA at single doses equal to the daily amounts required to reduce creatinine clearance with 14 days of treatment. Similar therapeutic indices were derived comparing toxicity by either method to prevention of rat ear-heart allograft rejection. C18-OH-ascomycin, an FK506-binding protein (FKBP) antagonist, reversed in vivo immunosuppression by FK506 and ascomycin in the rat, and pretreatment in the CISPA blocked FK506 and ascomycin nephrotoxicity, which showed a common immunophilin dependence. Rapamycin nephrotoxicity was unaffected (as with cyclosporine), which indicated that binding to FKBP was not required, Rapamycin nephrotoxicity thus appears mechanistically unrelated to its immunosuppressive mode of action. Screening with the CISPA enabled discovery of A-119435, a less nephrotoxic ascomycin analog having a 10-fold higher therapeutic index.

The immunosuppressive agent FK-506 has received much attention due to its efficacy and potency in the areas of transplant rejection and autoimmune disease. Calcineurin, a Ca(2+)-calmodulin activated phosphatase, was recently implicated in the immunosuppressive mechanism of FK-506. In our ongoing search for superior immunosuppressive agents, we have synthesized several analogues of FK-506 and tested their mechanistic and immunosuppressive actions. It was found that C-18 hydroxyl analogues of ascomycin, an analogue of FK-506 also called FR900520, bound tightly to immunophilin FKBP-12, but do not show any immunosuppressive activity in vitro or in vivo despite good bioavailability. Further, they reverse the inhibition of calcineurin caused by FK-506/FKBP-12 complex.

The 3D structure of two unlabeled FK506 analogs, (R)‐ and (S)‐[18‐OH]ascomycin, when bound to [U‐13C,14N]FKBP were determined by isotope‐filtered 2D NMR experiments. The structures for the R and S isomers that bind tightly to FKBP but lack immunosuppresive activity are compared to each other and to the conformation of the potent immunosuppressant, ascomycin, when bound to FKBP. The results are interpreted in terms of calcineurin binding to the FKBP/ascomycin complex.

The synthesis and structure-activity relationships of C-terminal octapeptide analogues of anaphylatoxin C5a have been studied. The introduction of hydrophobic amino acids into the N-acetylated native octapeptide (N-Ac-His-Lys-Asp-Met-Gln-Leu-Gly-Arg-OH) (1) has led to an analogue with 100 times more activity than the native octapeptide in inhibiting the binding of 125I-labeled anaphylatoxin C5a to human neutrophil membrane receptors. The observed apparent binding Ki's for the compounds (8-10) are in the range of 1-3 microM, and they possess nearly full agonist activity, despite the fact that these analogues are one-eighth or -ninth the size of the natural ligand anaphylatoxin C5a.

A number of C5a modifications were tested to determine effects on receptor binding to polymorphonuclear leukocyte (PMNL) membrane receptors and triggering of PMNL chemokinesis and myeloperoxidase (MPO) release. Site-directed mutagenesis was used to probe relationships of key C-terminal residues, and suggested a role for additional sites, particularly Lys19-20. A synthetic peptide based on C5a 19-30, weakly inhibited C5a binding. Potency of the C-terminal octapeptide, a full agonist, was markedly improved by a single Phe substitution for His67, and a Phe point mutation at this site was shown to enhance activity of the full recombinant protein.

C5a is a 74 amino acid polypeptide that likely plays an important role in the pathogenesis of a number of inflammatory diseases. Therefore, the discovery of a C5a antagonist is of considerable therapeutic interest. A series of peptides designed to survey various regions of the molecule was synthesized by solid-phase peptide synthesis and evaluated for receptor-binding activity with polymorphonuclear leukocyte membranes. The C-terminal octapeptide (Ac-His-Lys-Asp-Met-Gln-Leu-Gly-Arg-OH) was identified as the smallest fragment which possessed reasonable C5a receptor binding activity.