The cystatin "superfamily" encompasses proteins that contain multiple cystatin-like sequences. Some of the members are active cysteine protease inhibitors, while others have lost or perhaps never aquired this inhibitory activity. In recent years, several new members of the superfamily have been characterized, including proteins from insects and plants. Based on partial amino acid homology, new members, such as the invariant chain (Ii), and the transforming growth factor+ receptor type I1 (TGF-/3 receptor 11) may, in fact, represent members of an emerging family within the superfamily that may have used some common building blocks to form functionally diverse proteins. Cystatin superfamily members have been found throughout evolution and members of each family of the superfamily are present in mammals today. In this review, the new and older, established members of the family are arranged into a possible evolutionary order, based on sequence homology and functional similarities.Keywords: cystatins; cysteine protease inhibitors; evolution; fetuins; kininogens; TGF-/3 receptor I1The concept of a cystatin "superfamily" emerged in the early 1980s, precipitated by an observation that multiple cystatin-like sequences were present in the kininogens and that the stefins were related to both the cystatins and the repeats in the kininogens (Ohkubo et al., 1984). The cystatins are a family of cysteine protease inhibitors with homology to chicken cystatin (Barrett, 1981). Cystatins typically comprise = 1 15 amino acids, are largely acidic, contain four conserved cysteine residues known to form two disulfide bonds (Grubb et a]., 1984), may be glycosylated and/or phosphorylated, and contain a series of conserved residues, most notably the "QVVAG' sequence (reviewed in Rawlings & Barrett, 1990;Turk & Bode, 1991).Protein, cDNA, and genomic sequences of several cystatins have since been reported (e.g., Tsai et al., 1996; reviewed in Rawlings & Barrett, 1990;Turk & Bode, 1991). Their structural genes comprise three exons of characteristic size (e.g., Huh et al., 1995) (Fig. I), and a cluster of these genes is located on human chromosome 20 (e.g., Thiesse et al., 1994) although cystatin genes are also found on other chromosomes (e.g., Pennachio et al., 1996).Cystatin-like sequences identified in the kininogens (Ohkubo et al., 1984) were found to be present as sequence duplications. Subsequently, it was proposed that there were three cystatin-like repeats in the kininogen sequence, the amino terminal repeat being Reprint requests to: K.M Dziegielewska, Department of Anatomy and Physiology, University of Tasmania, GPO Box 252-24, Hobart, Tasmania 7001, Australia. less conserved (Salvesen et al., 1986). This threefold repeat is reflected in the triplicated gene structure (Kitamura et al., 1985). which has been shown to be common to the kininogens (e.g., Cole & Schreiber, 1992). Each of the three cystatin repeats is encoded by three exons of characteristic size (e.g., Cole & Schreiber, 1992).Thus, by the late 1980s the superfa...