C1r and C1s, the constituent proteins of C1s-C1r-C1r-C1s, the Ca2+ -dependent catalytic unit of C1, are homologous serine proteinases that share a common activation pattern and have similar structural organizations at the monomeric level. In both cases, activation occurs through cleavage of a single Arg-Ile bond, which converts the single-chain proenzymes into active proteinases comprising two chains linked by a single disulphide bridge. Both NH2-terminal A chains are sub-divided into five structural units (I-V) including a single copy of an Epidermal Growth Factor-like segment (II) and two different pairs of internal repeats (I/III and IV/V). Regions I and III have no equivalent in other proteins, whereas regions IV and V are homologous to short consensus repeats found, in particular, in complement proteins C2, B, H, C4b-binding protein and CR1. The COOH-terminal B chains are homologous to the catalytic chains of serine proteinases, but lack the "histidine-loop", a disulphide bridge common to all other known mammalian serine proteinases. Overall sequence comparison of C1r and C1s reveals 40% amino acid identity and conservation of all cysteine residues. In contrast, C1r and C1s widely differ from each other by their glycosylation patterns: both proteins contain Asn-linked carbohydrates, but four glycosylation sites are present on C1r, and only two on C1s.(ABSTRACT TRUNCATED AT 250 WORDS)