The nucleotide (nt) sequence of the Escherichia coli gcvP gene was determined. The polypeptide deduced from the DNA sequence has an M(r) of 104,375 (957 amino acids). In a minicell system, gcvP encodes a polypeptide that migrates at 93.3 kDa on sodium dodecyl sulfate-polyacrylamide gels. After the coding region, there is a 39-nt sequence followed by a T-rich sequence within which transcription appears to terminate. This region is preceded by a G/C-rich sequence that could form a stable stem-loop structure once transcribed, and is characteristic of Rho-independent transcription terminators. A Northern analysis identified an approx. 4700-nt RNA molecule, large enough to encode the T-, H-and P-proteins of the glycine cleavage enzyme complex. Analyses of gcvP::lacZ fusions with and without stop codons in gcvT, the first gene in the operon, confirmed gcvT, gcvH and gcvP lie in an operon. RNA slot blot analyses indicated that induction of gcv by glycine, and PurR-mediated repression of gcv occur at the level of transcription.