Recently the classical view of the compartmentalization of

Recently, the classical view of the compartmentalization of the plant isoprenoid metabolism has been re-evaluated, following the demonstration of the partial peroxisomal localization of the mevalonic Tolcapone australia (MVA) pathway, generally regarded as cytosolic. Using GFP-tagging approaches, the last two enzymes of this pathway, 5-phosphomevalonate kinase and mevalonate 5-diphosphate decarboxylase from Arabidopsis thaliana and Catharanthus roseus (Simkin et al., 2011) as well as the Arabidopsis short isoforms of isopentenyl diphosphate isomerase (Sapir-Mir et al., 2008) were shown to be targeted to peroxisomes. As FPS is usually related to the MVA pathway and has been reported to be localized within the peroxisome in mammals (Krisans et al., 1994, Olivier et al., 2000, Kovacs et al., 2007), we focused the present study on the cDNA cloning, functional characterization and subcellular localization of the Madagascar periwinkle (C. roseus) FPS.
Materials and methods
Results
Discussion In this study, we cloned a C. roseus cDNA encoding CrFPS that displays high similarities with plant trans-type FPS. The CrFPS protein was shown to have FPS activity by functional complementation in the yeast S. cerevisiae. Due to the lack of transit peptide at the N-terminal end of the protein, CrFPS is related to short FPS isoforms found in plants. In addition, when compared to the three Arabidopsis FPS isoforms (AtFPS1S, AtFPS1L and AtFPS2) encoded by two genes (AtFPS1 and AtFPS2) (Cunillera et al., 1996, Cunillera et al., 1997), CrFPS was more closely related to AtFPS2. This observation suggests that the C. roseus genome likely contains another FPS gene encoding a long and short FPS isoforms. Closa et al. (2010) reported that double knockout mutants of Arabidopsis AtFPS genes exhibit a lethal phenotype. In contrast, no significant developmental and metabolic differences were observed in single mutants, showing a partial redundancy of FPS1 and FPS2 isozymes in the biosynthesis of FPP-derived products. However, these results are not supported by studies on subcellular compartmentalization and so far, only the long AtFPS1 isoform was clearly localized in the mitochondria (Cunillera et al., 1997). Consequently, due to multiple localizations reported for FPS in plants, without strong evidence about the compartmentalization of short FPS isoforms, we decided to evaluate the subcellular distribution of CrFPS via an YFP fusion-protein biolistic approach. This study constitutes the first report of the subcellular localization of a short FPS isoform in plants. Our results showing that CrFPS is targeted to peroxisomes (Fig. 3E–G) are in agreement with recent reports on the subcellular distribution of enzymes involved in the early steps of isoprenoid biosynthesis for both plants and animals. For instance, in plants, the two short isoforms of A. thaliana isopentenyl diphosphate isomerase (IDI) are localized to peroxisomes in agreement with the presence of a peroxisomal targeting signal (PTS) type 1 at their C-terminal end (Sapir-Mir et al., 2008). Two types of PTS (PTS1 and PTS2) mediate the peroxisome import of proteins by the PTS1 receptor PEX5 and the PTS2 receptor PEX7 (Kaur et al., 2009). PTS1 is commonly a C-terminal tripeptide with the consensus sequence (S/C/A)(K/R/H)(L/M) and PTS2 is a nonpeptide located internally or near the N-terminus, and having the usual consensus sequence (R/K)(L/V/I/Q)X5(H/Q)(L/A/F) (Petriv et al., 2004). In addition to the short IDI isoforms, bioinformatic search of PTS motifs and arguments based on substrate channeling lead to a proposed model in which part of the MVA enzymes, IDI and FPS are targeted to the peroxisome in plants (Sapir-Mir et al., 2008). We recently confirmed this prediction for the last two enzymes of the MVA pathway, 5-phosphomevalonate kinase (PMK) and mevalonate 5-diphosphate decarboxylase (MVD) from C. roseus and from A. thaliana, which were shown to be targeted to peroxisomes (Simkin et al., 2011).