Peroxisomal and mitochondrial carnitine acetyltransferases of the n-alkane-assimilating yeast Candida Tropicalis. Analysis of gene structure and translation products.

A genomic DNA clone encoding carnitine acetyltransferases (EC 2.3.1.7), localized in two subcellular organelles, peroxisomes and mitochondria of an n-alkane-assimilating yeast Candida tropicalis, was isolated from the yeast lambda EMBL library using a carnitine acetyltransferase CDNA probe. Nucleotide sequence analysis disclosed that the open reading frame was 1881 bp, corresponding to 627 amino acids with a molecular mass of 70760 Da. Comparison of the predicted amino acid sequence of the C. tropicalis enzyme with that of Saccharomyces cerevisiae mitochondrial matrix carnitine acetyltransferase revealed 46.3% identity. It was noticeable that the C. tropicalis enzymes had amino acid sequences similar to both proposed mitochondrial and peroxisomal targeting signals. When the C. tropicalis gene was expressed in S. cerevisiae using its own 5'-upstream region, a 12-fold increase in activity was observed. Western blot analysis revealed the presence of two major proteins whose sizes corresponded to the peroxisomal and mitochondrial proteins detected in C. tropicalis. This suggested that peroxisomal and mitochondrial carnitine acetyltransferases were encoded by one gene, as suggested for the S. cerevisiae enzyme. Furthermore, we have separated and purified these enzymes from peroxisomes and mitochondria of C. tropicalis, and analyzed the amino-terminal amino acid sequences of each. The amino-terminal sequence of the mitochondrial enzyme suggested that a signal sequence had been cleaved during translocation into mitochondria. Concerning the peroxisomal enzyme, the evidence obtained indicated that in vivo the translation was initiated at the second methionine of the open reading frame.