The import of most of peroxisomal proteins into the lumen of their target organelle is driven by C-terminal (PTS1) or N-terminal (PTS2) signals recognized by the Pex5p or Pex7p receptors, respectively. However, some proteins in budding yeast, such as acyl-CoA oxidase (AOx) and carnitine acetyltransferase (Cat2p), are imported into peroxisomes via an alternative route that does not rely on known PTS signals and involves the Pex5p receptor N-terminal region. Here, we show that two other budding yeast peroxisomal proteins, a multifunctional enzyme from the β-oxidation pathway (Fox2p) and catalase A (Cta1p), both of which contain PTS1, can be imported independently of this signal. The I264K amino acid substitution in Pex5p adjacent to its FxxxW diaromatic motif, previously shown to abolish the import of AOx and Cat2p into peroxisomes, also affects Fox2p and Cta1p import. Moreover, we demonstrate that Pex9p, a newly discovered paralog of Pex5p that was recently implicated in the import of malate synthases in budding yeast, also exhibits weak receptor activity towards Fox2p and Cta1p. These findings indicate the need to re-evaluate the peroxisomal import paradigm.This article has an associated First Person interview with the first author of the paper.

The specificity of import of peroxisomal matrix proteins is dependent on the targeting signals encoded within their amino acid sequences. Two known import signals, peroxisomal targeting signal 1 (PTS1), positioned at the C-termini and PTS2 located close to N-termini of these proteins are recognized by the Pex5p and Pex7p receptors, respectively. However, in several yeast species, including Saccharomyces cerevisiae, proteins exist that are efficiently imported into peroxisomes despite having neither PTS1 nor PTS2 and for which no other import signal has been determined. An example of such a protein is S. cerevisiae acyl-CoA oxidase (AOx) encoded by the POX1 gene. While it is known that its import is driven by its interaction with the N-terminal segment of Pex5p, which is separate from its C-terminal PTS1-recognizing tetratricopeptide domain, to date, no AOx polypeptide region has been implicated as critical for this interaction, and thus would constitute the long-sought PTS3 signal. Using random mutagenesis combined with a two-hybrid screen, we identified single amino acid residues within the AOx polypeptide that are crucial for this interaction and for the peroxisomal import of this protein. Interestingly, while scattered throughout the primary sequence, these amino acids come close to each other within two domains of the folded AOx. Although the role of one or both of these regions as the PTS3 signal is not finally proven, our data indicate that the signal guiding AOx into peroxisomal matrix is not a linear sequence but a signal patch.

No abstract