ADP-ribosylation factor 1 (Arf1) is an essential N-myristoylated 21-kDa GTP-binding protein with activities that include the regulation of membrane traffic and phospholipase D activity. Both the N terminus of the protein and the N-myristate bound to glycine 2 have previously been shown to be essential to the function of Arf in cells. We show that the bound nucleotide affects the conformation of either the N terminus or residues of Arf1 that are in direct contact with the N terminus. This was demonstrated by examining the effects of mutations in this N-terminal domain on guanosine 5'-O-(3-thio)triphosphate (GTP gamma S) and GDP binding and dissociation kinetics. Arf1 mutants, lacking 13 or 17 residues from the N terminus or mutated at residues 3-7, had a greater affinity for GTP gamma S and a lower affinity for GDP than did the wild-type protein. As the N terminus is required for interactions with target proteins, we conclude that the N terminus of Arf1 is a GTP-sensitive effector domain. When Arf1 was acylated, the GTP-dependent conformational changes were codependent on added phospholipids. In the absence of phospholipids, myristoylated Arf1 has a lower affinity for GTP gamma S than for GDP, and in the presence of phospholipids, the myristoylated protein has a greater affinity for GTP gamma S than for GDP. Thus, N-myristoylation is a critical component in the construction of this phospholipid- and GTP-dependent switch.

Wild type and eight point mutants of Saccharomyces cerevisiae ARF1 were expressed in yeast and bacteria to determine the roles of specific residues in in vivo and in vitro activities. Mutations at either Gly2 or Asp26 resulted in recessive loss of function. It was concluded that N-myristoylation is required for Arf action in cells but not for either nucleotide exchange or cofactor activities in vitro. Asp26 (homologous to Gly12 of p21ras) was essential for the binding of the activating nucleotide, guanosine 5'-3-O-(thio)triphosphate. This is in marked contrast to results obtained after mutagenesis of the homologous residue in p21ras or Gs alpha, and suggests a fundamental difference in the guanine nucleotide binding site of Arf with respect to these other GTP-binding proteins. Two dominant alleles were also identified, one activating dominant ([Q71L]Arf1) and the other ([N126I]) a negative dominant. A conditional allele, [W66R]Arf1, was characterized and shown to have approximately 300-fold lower specific activity in an in vitro Arf assay. Two high-copy suppressors of this conditional phenotype were cloned and sequenced. One of these suppressors, SFS4, was found to be identical to PBS2/HOG4, recently shown to encode a microtubule-associated protein kinase kinase in yeast.

Our data suggests that duloxetine has a beneficial effect on CIPN caused by oxaliplatin, paclitaxel, vincristine, or bortezomib in Japanese patients.

No abstract