388 Although farnesylation is required for a number of abscisic acid mediated responses in plants, knowledge of how this lipid modification of proteins regulates specific developmental and physiological processes remains unclear. Recent information from the Arabidopsis genome-sequencing project in combination with mutants deficient in farnesylation should unravel the role(s) of this process in plant signaling. Addresses Department of Botany, University of Toronto, 25 Willcocks Street, Toronto, Canada M5S 3B2 *e-mail: mccourt@botany.utoronto.ca Current Opinion in Plant Biology 1999, 2:388–392 1369-5266/99/$ — see front matter © 1999 Elsevier Science Ltd. All rights reserved. Abbreviations FPP farnesyl diphosphate FTase farnesyltransferase MVA mevalonate Introduction The post-translational modification of proteins through phosphorylation, methylation or glycosylation permits the rapid relay response that is the signature of a signal trans- duction cascade. Over the past decade advances in animals and fungi have now shown that lipidation of signaling mol- ecules is also essential for many of these proteins to function [1–4]. The attachment of a lipid to a protein pro- motes a hydrophobic interaction between the signaling molecule and membrane lipids and/or other proteins [5,6]. The most common lipid modifications involve fatty acids (myristate and palmitate), isoprenoids (farnesol and ger- anylgeranol) and glycosyl-phosphatidyl inositol anchors [1]. Of these, farnesylation has garnered much of the sci- entific attention because proteins involved in yeast mating and mitogenic responses in mammals such as the γ-subunit of heterotrimeric G-proteins and members of the RAS superfamily of proteins need to be farnesylated to function correctly [2–4]. Both in yeast and in animal cells, inhibition of protein farnesylation blocks the correct localization of RAS proteins to the plasma membrane. In the case of ani- mals, the ability of oncogenic RAS to cause transformation of cell lines is dependent upon farnesylation. Since 30% of human tumors contain mutated RAS protein, this result suggests farnesylation may be an excellent pharmacologi- cal target for cancer therapy. It takes three to farnesylate Studies on farnesylation in plants began with simple ques- tions such as: does this reaction occur in plants and are target proteins similar to yeast and animals? To a first approximation, the answer to this first question, at the structural and biochemical level, is yes [7–12]. As with other systems, plant protein farnesylation appears to be strictly dependent on the presence of a carboxy-terminal CaaX motif. The ‘C’ is a cysteine and the ‘aa’ are usually, but not always, two aliphatic amino acids. In animal and yeast studies the ‘X’ is preferentially a cysteine, methion- ine, serine, alanine, or glutamine. On the basis of present data, it appears that these five terminal residues also serve as farnesylation substrates in plants [8,9,13]. In yeast and animals, once a farnesyl lipid is attached to the cysteine via a thioester bond the last three amino acids are often prote- olytically removed and the now terminal cysteine is methylated [14–17]. Although carboxyl methylation has been shown in plants [18 • ], proteolysis of the last three amino acids has not been demonstrated. Interestingly, a BLAST search of the Arabidopsis DNA database does identify a closely related gene (Accession No. AAB61028) to the human CaaX phenyl protease, suggesting this enzy- matic function does exist. The protein farnesyltransferase (FTase) which carries out the attachment of the lipid to the target protein, is a het- erodimeric enzyme that contains an α- and β-subunit. Both of these genes have been cloned in Arabidopsis, pea and tomato and a β-subunit has been identified in tobacco [19–23]. In cultured tobacco cells, FTase activity has been linked to the cell cycle [20] and levels of transcription of the β-subunit gene does correlate with the cell cycle [24]. Consistent with these studies, both the pea α-and β-sub- unit genes are more highly expressed in actively dividing tissues such as roots and nodules. Expression of the pea β- subunit gene appears to be positively regulated by auxin, and negatively regulated by light and sugar [24]. Together these results indicate that the amount of FTase protein and its activity may link cell division rates with environ- mental signals. Plant FTases — like their animal counterparts — use mevalonate (MVA)-derived farnesyl diphosphate (FPP) as a substrate. However, unlike other organisms, plants pro- duce a large number of unique MVA-derived compounds, including the plant hormones cytokinins, gibberellins and abscisic acid (ABA). Although little is known about the metabolic regulation and compartmentation of the MVA pathway in plants, novel mechanisms might exist that con- trol the synthesis and use of FPP. Studies on the influence of cholesterol metabolism on RAS farnesylation in animals support this possibility [25]. So many targets, so little time Interest in farnesylation has now shifted to defining the targets of farnesylation in plants and determining whether they are involved in plant growth and development. To date, no true RAS or G-protein γ-subunit homologues have been identified in plants. The first biochemically Protein farnesylation in plants: a greasy tale Eiji Nambara and Peter McCourt*