Abstract

Mammalian ADP-ribosylation factors (ARFs), approximately 20-kDa guanine nucleotide-binding proteins that stimulate cholera toxin ADP-ribosyltransferase activity, were grouped into three classes based on deduced amino acid sequence. Human ARF 1, a class I ARF, is identical with its bovine counterpart, has a distinctive pattern of tissue and developmental expression, and is encoded by a approximately 1.9-kilobase mRNA. ARF 1 cDNAs were isolated from a human fibroblast cDNA library; one arose via an alternative polyadenylation signal (AA-TACA) 84 nucleotides 5' to the polyadenylation signal (AATAAA) used in the 1815-base pair cDNA. The polyadenylation signals, their respective locations, and the surrounding nucleotide sequences are conserved in human and rat. The human ARF 1 gene, with four introns, spans approximately 16.5 kilobases. Exon 1 (46 base pairs) contains only untranslated sequence. Translation initiates in exon 2, which encodes the sequence GXXXXGK involved in phosphate binding (GTP hydrolysis). The sequence DVGG is encoded in exon 3, and NKQD, which is involved in the interaction with the guanine ring, is interrupted following the codon for Q by intron 4. The carboxyl-terminal 53 amino acids and greater than 1110 base pairs of 3'-untranslated region are encoded in exon 5. Primer extension and mung bean and S1 nuclease mapping indicated multiple transcription initiation sites and were consistent with Northern analyses. The 5'-flanking region has a high GC content but no TATA or CAAT box, as found in housekeeping genes. In addition, the two human class I ARF genes, ARF 1 and ARF 3, have similar exon/intron organizations and use GC-rich promoters.