Beiträge zur Inhibierung der Proproteinkonvertase Furin

The serine protease furin belongs to the family of proprotein convertases and fulfills important physiological functions in the processing of numerous vital precursor proteins (proproteins). On the other hand, furin is also associated with diverse pathological processes, e.g., various bacterial and viral infections. Therefore, furin emerged as an attractive target for drug development. Based on the consensus furin cleavage site comprising the multibasic sequence -Arg-Xaa-Arg/Lys-Arg- (Xaa: any amino acid), new fluorogenic furin substrates and substrate-analogue furin inhibitors were synthesized within this work. All compounds were obtained by a combination of solid phase and solution synthesis. The new substrates are more efficiently cleaved and possess improved kcat/KM values compared to already known derivatives. The improved specificity constants enable more sensitive kinetic measurements at reduced substrate and enzyme concentrations. Compared to the previously used compound Phac-Arg-Val-Arg-Arg-AMC, the new substrate acetyl-Arg-Arg-Tle-Arg-Arg-AMC exhibits a 30 times higher specificity constant and enables a more convenient Ki value determination of highly potent furin inhibitors (which were previously characterized as tight-binding inhibitors) under classical, non-tight-binding conditions. Several cyclic furin inhibitors were developed by selective modifications of a previously published structure with pronounced cell penetrating properties. All these oligoarginine-containing compounds with the benzamidine moiety 4-Amba (4-amidinobenzylamide) in P1 position exhibit a high inhibitory potency towards furin. Moreover, the potency of previously developed linear oligoarginine derivatives with a C-terminal 4-Amba residue could be improved by converting their N-terminal α-amino group into a more basic guanidine. The previously published compound 4-guanidinomethyl-Arg-tert-Leu-Arg-4-Amba (MI-1148, Ki = 5.5 pM), developed in our group, is the most efficient furin inhibitor known so far. In a first approach, its backbone has been C-terminally elongated by replacement of the P1 4-Amba with various stable or cleavable P1-P1’ segments. However, the most potent of these derivatives possesses only a moderate Ki value of 20 nM. In another attempt, a series of stepwise C-terminally truncated derivatives was developed. The loss of the P1 benzamidine led to a dramatic drop in the inhibitory potency. Interestingly, the truncated derivatives showed an unexpected behaviour in the kinetic test, deviating from the usually observed classical competitive inhibition mechanism. As the binding mode on furin was not known, only IC50 values could be determined for these compounds. All effective synthetic furin inhibitors are multibasic compounds. The results of previous work also revealed that a basic P1 residue is of prime importance for a strong furin inhibition by peptidomimetic inhibitors. However, numerous adverse effects have been described in literature for benzamidine derivatives. Therefore, the 4-Amba residue in inhibitor MI-1148 was replaced by various other groups, most of which are less basic. Interestingly, some of these derivatives showed the same aberrant enzyme kinetic behaviour as observed for the truncated peptides. According to mathematical and graphical analysis, it could be classified as a parabolic competitive inhibition. This kinetic model is characterized by the simultaneous binding of two inhibitor molecules on two different sites in one enzyme molecule. The P1 benzylamine and aminopyridine analogues exhibit a relatively high inhibitory potency. Even more active is a fluoro-substituted benzamidine compound (Ki = 22.3 pM), which also effectively inhibits the propagation of Dengue and West Nile viruses and the cleavage of the hemagglutinine precursor of the highly pathogenic avian influenza virus A H7N1 in cell culture. A first study in mice revealed a suboptimal toxicological profile for inhibitor MI-1148. By contrast, its tribasic analogues possessing a neutral P5 phenylacetyl residue (still a strong furin inhibitor) or P1 benzylamide moiety (a poor inhibitor) were better tolerated. The similar toxicity of these two compounds suggests that the low tolerability in vivo does not correlate with the strength of furin inhibition and is rather caused by unspecific off-target effects. The results of the study also suggest that the toxicity depends at least to some extent on the overall multibasic character of the inhibitor MI-1148. However, no suitable alternative for the P1-Amba residue could be identified thus far. Besides, benzamidine groups are found in some marketed drugs. This suggests to retain the benzamidine in P1 position and replace other residues instead. Since the basic P5 and P4 residues are also essential for furin inhibition, only the P2 arginine was replaced by the slightly less basic lysine; the resulting inhibitor is nearly as effective as the P2 arginine derivative in cell culture and shows a better tolerability in vivo.