The rat alpha-fetoprotein and albumin genes. Transcriptional control and comparison of the sequence organization and promoter region.

Functional and structural approaches were used to characterize the transcription units of the rat alpha-feto-protein (AFP) and albumin genes. A cell-free nuclear transcription assay and several genomic clones were used to show that: 1) the rate of transcription of these genes is closely related to the levels of corresponding mRNAs in the yolk sac and during rat liver development, indicating that the expression of the albumin and AFP genes is mainly regulated at the transcriptional level in the rat, and 2) the in vivo 5' end boundaries of the rat AFP and albumin transcription domains were mapped near the respective first exons. Due to the presence of repeated sequences, the 3' end boundary of both genes could not be accurately defined in the same manner. 3) No transcription could be detected until 7 kilobases upstream from the cap site of these genes. In addition, the organization of the rat AFP gene was analyzed by restriction endonuclease mapping, S1 nuclease mapping, and nucleotide sequencing. Our results indicate that: 1) the rat AFP gene is 20 kilobase pairs long and is split into 15 exons by 14 intervening sequences; 2) the transcription initiation site of the rat AFP gene is heterogenous; 3) the 5'-flanking region upstream from the rat AFP gene exhibits 60-90% similarity with the mouse and human AFP genes while no major nucleotide identity is found with the rat albumin gene; 4) a 90-base pair sequence present as one copy upstream from the rat and mouse AFP genes is present as two copies in the human genome; 5) several inverted repeats are mapped in the 5'-flanking region indicating potential stem-loop structures. One highly conserved structure encompasses an enhancer-like core sequence and the sequence recognized by the TGGCA-binding protein.