Although the third component of complement, C3, has been isolated and its primary structure determined from most living classes of vertebrate, limited information is available on its structure and function for aves, which represent a significant stage in complement evolution. In this study, we present the complete cDNA sequence of chicken C3, the cDNA sequences of the thioester region for two chicken alpha 2-macroglobulin (alpha 2M)-related proteins, a simplified method for purifying chicken C3, and an analysis of the C3 convertase and factor I-mediated cleavages in chicken C3. Using the reverse-transcriptase PCR, with degenerate oligonucleotide primers derived from two conserved C3 sequences (GCGEQN/TM, TWLTAY/FV) and liver mRNA as template, we isolated three distinct 220-bp PCR products, one with a high degree of sequence similarity to C3 and two to alpha 2M and pregnancy zone protein from other species. The complete cDNA sequence of chicken C3 was obtained by screening a chicken liver lambda gt10 library with the C3 PCR product and probes from the 5' end of the partial-length C3 clones. The obtained sequence is in complete agreement with the protein sequence of several tryptic peptides of purified chicken C3. Chicken pro-C3 consists of an 18-residue putative signal peptide, a 640-residue beta-chain (70 kDa), a 989-residue alpha-chain (111 kDa), and an RKRR linker region. It contains an internal thioester and three potential N-glycosylation sites, all in the alpha-chain. The convertase cleavage site, predicted to be Arg-Ser, was confirmed by sequencing the zymosan-bound C3 fragments generated upon complement activation. NH2-terminal sequencing of the purified C3 chains showed that 1) pro-C3 is indeed cleaved at the RKRR linker sequence to generate the mature two-chain molecule, and 2) the beta-chain of chicken C3 is blocked. The deduced amino acid sequence shows 54, 54, 54, 53, 52, 57, and 55% amino acid identities to human, mouse, rat, guinea pig, rabbit, cobra, and Xenopus C3, respectively, and an identity of 44, 31, and 33% to trout, hagfish, and lamprey C3, respectively. The identities to human C4, C5, and alpha 2M are 31, 29 and 23%, respectively. A phylogenetic tree for C3, C4, C5, and alpha 2M-related proteins was constructed based on the sequence data and is discussed.