The glycine cleavage system. Occurrence of two types of chicken H-protein mRNAs presumably formed by the alternative use of the polyadenylation consensus sequences in a single exon.

Several cDNAs encoding H-protein, a constituent of the glycine cleavage system, were cloned from chicken liver cDNA libraries with an antibody raised against rat H-protein or with a nick-translated cDNA of an immunoreactive clone. The structure of the H-protein cDNA consisting of 910 base pairs was determined using clones with an apparent overlap in the nucleotide sequence. The cDNA encodes the precursor form of H-protein that is comprised of 39 amino acid residues for a mitochondrial presequence and 125 amino acid residues for the mature protein, following a 5' untranslated region of 13 base pairs. There are two genuine consensus sequences for the cleavage/polyadenylation of the precursor H-protein mRNA in the 3' untranslated region of the cDNA sequence. We showed by comparison with the delta-aminolevulinate synthase gene that only one copy of the H-protein cDNA occurs in the haploid genome of the chicken. Nevertheless, two types of H-protein mRNAs, which differ by the length of their 3' untranslated region, are produced in liver. The chicken H-protein gene extends over 8 kilobase pairs on the genome and includes 5 exons that encode the entire cDNA sequence. Two AATAAA motifs are coded in the last exon of this gene, suggesting that the differently size H-protein mRNAs are produced by the alternative use of these motifs.