Chapter 3: Loop-directed mutagenesis of the blue copper protein amicyanin and its effect on the structure and the activity of the type-1 copper site

Five loop-mutants of the blue copper protein amicyanin from Paracoccus versutus have been constructed and characterized. The mutations replaced the loop containing three (Cys93, His96, Met99) of the four copper ligands in amicyanin by the 'ligand loops' of P. aeruginosa azurin (AmiAzu), A. faecalis pseudoazurin (AmiPaz), P. nigra plastocyanin (AmiPcy), T. ferrooxidans rusticyanin (AmiRus) and P. aureofaciens nitrite reductase (AmiNiR). The copper centers of all variants appear to be perfect type-1 Cu sites although the AmiNiR variant exhibits diminished stability. The optical spectra of the AmiAzu and AmiPaz variants display a significant dependence on temperature. The reduction potentials of the four stable variants are all higher than that of wt amicyanin. The reduced forms of the loop-mutants AmiAzu, AmiPaz and AmiPcy protonate at the C-terminal histidine, with pKa values of 5.63, 5.4 and 5.71, respectively (6.8 for wt amicyanin). In contrast, no protonation could be observed for the C-terminal histidine of AmiRus, which makes it the first known amicyanin mutant that does not undergo this protonation. The electron selfexchange rate constants at 25 °C are about one order of magnitude lower for the loop mutants than for wt amicyanin (1.2x10 Ms). The results are interpreted by taking into consideration that three metal ligands are intimately connected with the stable β-sandwich structure of the cupredoxin. This leaves considerable freedom in positioning the fourth one, the C-terminal His, on the 'ligand loop'. This explains the ease by which the Cterminal histidine-ligand can be excised and replaced by external ligands without loosing the metal binding property of the protein. The results also help understand the remarkable evolutionary success of the combination of cupredoxin fold and Cu-site for mediating biological electron transfer. Loop mutants of amicyanin 42 Introduction Whenever nature uses copper to promote biological electron transfer (ET), the metal appears to be encased in what has come to be known as the cupredoxin fold [Adman, 1991]. The question why this combination of metal site and fold is such a unique and successful vehicle for biological ET is addressed in this paper. The fold is usually made up of 6-8 βstrands in a Greek key folding motif. The strands arrange themselves into two β-sheets which face each other to form a β-sandwich [Adman, 1991]. Within a cupredoxin domain the so-called type-1 copper binding site is located excentrically at a distance of 5-7 Å from the outer surface. The metal is anchored in the protein structure by strong bonds to three ligands, a cysteine and two histidines which coordinate by their S-atom and their N-atoms, respectively, in a trigonal planar or a trigonal pyramidal geometry. A weaker fourth ligand, consisting of a methionine or, sometimes, a glutamine, coordinating with its Sor O-atom, respectively, occupies the axial position (see Figure 1) [Romero et al., 1994]. In some cases (azurins) a second weakly interacting axial group in the form of a backbone carbonyl oxygen is found at the opposite axial position [Baker, 1988; Nar et al., 1991a]. Cupredoxin domains also occur as part of multi-copper, multi-domain enzymes. Examples are the blue oxidases (laccase [Ducros et al., 1998], ascorbate oxidase [Messerschmidt, 1993], ceruloplasmin [Zaitseva et al., 1996]) and nitrite reductase (NiR) [Adman et al., 1995; Dodd et al., 1998; Inoue et al., 1998; Murphy et al., 1997b]. Again the copper in the ET site is bound to a cysteine and two histidines, while a methionine, leucine or phenylalanine may occupy the fourth (axial) position. A stunning illustration of the versatility of the cupredoxin fold, when it comes to incorporating copper, is provided by subunit II (SUII) of cytochrome oxidase [Blackburn et al., 1997; Iwata et al., 1995; Figure 1. Active-site structure of wt amicyanin (Paracoccus versutus) [Kalverda et al., 1994; Romero et al., 1994].