Structural characterization of the gene for human histidine-rich glycoprotein, reinvestigation of the 5'-terminal region of cDNA and a search for the liver specific promoter in the gene.

Abstract

Genomic DNA libraries were screened for the human histidine-rich glycoprotein (HRG) gene and a sequence of 15,499 nucleotides was determined. The gene is composed of 7 exons and 6 introns, and all the exon-intron boundaries match the consensus GT/AG sequence for donor and acceptor splice sites. Each of cystatin-like domains I and II of HRG is encoded by three exons, exons I to III and exons IV to VI, respectively, like those of other members of the cystatin superfamily. The entire C-terminal half of the molecule is encoded by the largest exon, VII. The first 103 nucleotides of the cDNA sequence reported for human HRG [Koide, T., Foster, D., Yoshitake, S. , and Davie, E.W. (1986) Biochemistry 25, 2220-2225] could not be found in the determined gene sequence. A homology search of this sequence against a database showed the complete matching to a part of the yeast mitochondrial DNA encoding 21S ribosomal RNA. Rapid amplification of cDNA 5' ends (5'-RACE) analysis revealed that the cDNA has multiple 5'-ends and that a possible starting point is nucleotide 104 of the reported cDNA sequence. These results suggest that the first 103 nucleotides of the cDNA sequence reported for human HRG originated from yeast mitochondrial DNA and were incidentally incorporated into the HRG cDNA in the process of the construction of a cDNA library. Various fragments obtained on restriction endonuclease digestion of the 5'-noncoding region of the HRG gene were ligated to the chloramphenicol acetyltransferase (CAT) gene and then transfected into HepG2 and 293 cells to analyze the promoter activity. The sequence between -262 and -21 from the putative translation initiation site supported the expression of CAT in HepG2 cells but not in 293 cells, suggesting that this segment promotes the liver-specific transcription of the human HRG gene.