An active tryptic fragment of hydrogenase 2 from Escherichia coli has been isolated from the periplasmic face of the cytoplasmic membrane, and the large and small subunits N-terminally sequenced. The large subunit is encoded by the hybC gene and shows no N-terminal processing, other than removal of the initiator methionine during its biosynthesis. Both N-terminal and the subsequent internal trypticfragment amino acid sequence indicate that the small subunit is neither encoded by hybA, a gene previously identified as encoding the small subunit [Menon et al. (1994) J. Bacteriol. 176, 4416Ϫ4423], nor any of the remaining genes in the hyb operon. Genome sequence analysis revealed the presence of an open reading frame which could potentially encode the peptide sequences of the proteolysed small subunit. The gene, designated hyb0, lies directly upstream of, and is separated by two nucleotides from, the start of the hybA gene. Hyb0, which shares an approximate 40% identity with other hydrogenase small subunit amino acid sequences, is synthesised with an N-terminal signal sequence containing a twin-arginine motif which is probably required for export of the enzyme. In the mature enzyme the small subunit is proteolytically cleaved after Ala37. Immunological analysis of strains overproducing either recombinant Hyb0 or HybA using antibodies specific for hydrogenase 2, readily identified Hyb0 as the small subunit. In a pleiotropic hypB mutant, which is unable to insert nickel into the active site, both the large and small subunits accumulate as unprocessed, soluble forms, consistent with the two subunits being assembled and processed in a coordinated manner during biosynthesis.