Isolation and characterization of the human gene for ADP-ribosylation factor 3, a 20-kDa guanine nucleotide-binding protein activator of cholera toxin.

ADP-ribosylation factors (ARFs) are approximately 20-kDa guanine nucleotide-binding proteins that stimulate the ADP-ribosyltransferase activity of cholera toxin in vitro. Five different human ARFs have been identified by cDNA cloning. Northern analysis using ARF 3-specific oligonucleotides identified two mRNAs of 3.7 and 1.2 kilobases (kb). We report here the complete nucleotide sequence of the 3.7-kb ARF 3 mRNA derived from three overlapping cDNAs isolated from human hippocampus and fetal brain cDNA libraries, as well as the structure of human ARF 3 gene. Sequences of two overlapping genomic clones indicated that the ARF 3 gene spans approximately 18.3 kb and contains five exons and four introns. The conserved amino acid sequences involved in guanine nucleotide binding by ARF 3 are distributed among separate exons, as found in other GTP-binding protein genes. Translation initiates in exon 2 which includes the sequence GXXXXGK that probably participates in phosphate binding and GTP hydrolysis. The sequence DVGG in exon 3 coordinates binding of Mg2+ and the beta-phosphate of GDP. In the ARF 3 gene in contrast to those of other GTP-binding proteins, the sequence NKXD (which is thought to contribute to the specificity of interaction with the guanine ring) is divided between exons 4 and 5. The latter encodes the COOH-terminal 53 amino acids of ARF 3 and contains greater than 2500 base pairs of untranslated DNA. The sequence AATTAA is 19 bases 5' to the polyadenylation addition site of the 3.7-kb mRNA. Multiple transcription start sites were identified by primer extension and S1 and mung bean nuclease analyses. The 5'-flanking region of exon 1 contains neither a TATA nor a CAAT box, but is high in GC content (greater than 70%) and includes three potential Sp1-binding sites (GC box), consistent with the promoters described for several housekeeping genes. The 1.2-kb ARF 3 mRNA is shown to arise by use of an alternative polyadenylation signal (AACAAA) at nucleotide 1091 within the ARF 3 cDNA.