In mammalian melanocytes, a membrane-anchored high molecular weight protein complex consisting of tyrosinase (monophenol monooxygenase, EC 1.14.18.1) and two tyrosinase-related proteins (TRP-1 and TRP-2, respectively), catalyzes the conversion of tyrosine into the pigment eumelanin. These three proteins are members of a large gene family of binuclear copper proteins (Halaban and Moellmann, 1990). TRP-2 has recently been shown (Tsukamoto et al., 1992; Solano et al., 1996) to possess DOPAchrome tautomerase activity (EC 5.3.2.3). On the other hand, the function of TRP-1, the most abundant glycoprotein expressed in human melanocytes remains controversial. This protein, encoded by the brown locus, exhibits more than 50% sequence identity with tyrosinase (Cohen et al., 1990) and shows some tyrosine hydroxylase activity, but may specifically act as 5,6-dihydroxyindole-2-carboxylic acid (DHICA) oxidase (Jimenez-Cervantes et al., 1994; Kobayashi et al., 1994). However, this protein has also been suggested to act as a melanocyte-specific catalase entitled TRP-1 or catalase B (Halaban and Moellmann, 1990). Several observations support this: the oxidation of 5,6dihydroxyindole and DHICA is associated with the generation of hydrogen peroxide, which is in turn known to destroy melanin and some of its precursors. H2O2 may thus regulate the expression of tyrosinase in melanocytes (Karg et al., 1993), and obviously the course of melanogenesis is strongly affected by factors modulating the level of intracellular peroxides (Nappi and Vass, 1996). Furthermore, H2O2 is a rather potent reversible inhibitor of tyrosinase (Schallreuther et al., 1991), and the oxy-form of this enzyme even shows a limited catalatic potential (Wood and Schallreuther, 1991). The evolutionary relationships among different members of the tyrosinase family have been studied (Morrison et al., 1994), but so far there was no indication of any structural as well as evolutionary relatedness between TRP-1 (catalase B) and the family of heme containing catalases (EC 1.11.1.6). This is caused mainly by the fact that the sequences of brown locus proteins (which obviously bind copper) and those of heme catalases exhibit no easily detectable sequence similarities. For instance, the sequence identity between human TRP-1 (Cohen et al., 1990) and Saccharomyces cerevisiae catalase A (Cohen et al., 1988) is only about 16% in the 400 amino acid overlap (alignment not shown). The presumably related function of these two supposedly structurally unrelated enzymes has evoked the idea on their eventual very distant evolutionary relationship. We have recently shown that one of the most strictly conserved structural parts among heme catalases is formed by the major substrate channel allowing the diffusion