Structural organization of the genes for murine and human leukemia inhibitory factor. Evolutionary conservation of coding and non-coding regions

Leukemia inhibitory factor, LIF, is a glycoprotein with multiple activities in both the adult and the embryo. LIF appears to be encoded by a unique gene in both mouse and man, although the 3'-untranslated region of the mouse LIF gene gives a complex hybridization pattern on Southern blots. The complete nucleotide sequences of both the murine and human LIF genes and their flanking regions (8.7 and 7.6 kilobase pairs, respectively) were determined and compared. Both genes comprise three exons, two introns and an unusually long 3'-untranslated region (3.2 kilobase pairs), specificying a mRNA of approximately 4.1 kilobases. Two start sites of LIF-transcription were determined, by S1-nuclease protection and by a novel approach involving the polymerase chain reaction. S1-nuclease protection revealed a start site 60-64 base pairs upstream of the translational start codon and immediately downstream of a TATA box (TATATAAAT). The PCR approach identified a second transcriptional start site 160 base pairs 5' of the start codon and adjacent to a "TATA-like" element (CATAATTT). A comparison of the murine and human LIF gene sequences revealed a high degree of conservation in the coding regions and in segments of the untranslated and flanking regions. Seven segments displaying greater than 75% homology were identified, with the 5' and 3' ends of the transcription unit revealing the highest degree of homology. These conserved regions represents potential cis-acting control elements.