Amino acid sequences alignment of rat L-CPT I and CPT II reveals in L-CPT I the presence of an extended N-terminal domain (about 150 amino acids) which bears no similarity to CPT II.6 Deletion of the N-terminal domain of L-CPT I abrogates the ability of the protein to interact with the mitochondrial receptors, and thus to be specifically imported into rat liver mitochondria.51 Conversely, fusion of the L-CPT I N-terminal domain to the cytosolic mouse dihydrofolate reductase (DHFR) or to the mature form of CPT II (which lacks the matrix-targeting signal and is thus incompetent for import) allows them to be targeted to mitochondria both in vitro and in vivo, and to be inserted into the OMM.51 Thus, the OMM targeting signal of L-CPT I resides within the first 147 residues of the protein and explains the inability of L-CPT IA31-148 to be associated with mitochondria.48

Hydrophobic cluster analysis (HCA) is a very efficient method to analyse and compare protein sequences.92,93 Schematically, HCA uses a highly degenerated code for the sequences, where only two main states are initially considered: hydrophobic andhdrophilic.

A number of structural features of a globular protein, with or without transmembrane segments, can be derived from the simple examination of its HCA plot and concerns the structural segmentation, the identification of structural domains and some indications on the secondary structure and loops. For the identification of structural segmentation, the plot is examined to analyse the horizontal distribution and size of the various hydrophobic clusters that have been automatically drawn by the program. The HCA method is more efficient when several homologous sequences can be compared. Comparison of the HCA plots of the rat and human L-CPT I and M-CPT I is shown in Figure 2.

A segment containing mostly hydrophobic (VILFWMY) amino acids and virtually no DENQHKR residues is often indicative of completely buried secondary structure, mostly transmembrane a-helices (when the length is about 20 residues). Two of such hydrophobic clusters are recovered in the HCA plots of the four known CPT I proteins

(Fig. 2) and correspond to the two transmembrane segments (H1 and H2) of these proteins. According to the examination of the HCA plots, H1 comprises residues 47-75 (L-CPT Is) or 53-75 (M-CPT Is), whereas H2 corresponds to residues 103-123 (L-CPT Is) or 105-124 (M-CPT Is). Mosaic (zig-zag) clusters that contain highly hydrophobic residues alternating with hydrophilic ones are often associated with amphiphilic P-strands. Two putative highly conserved amphiphilic P-strands (pi, 8-14 amino acids; P2, 19-23 amino acids) are observed in the extreme N-terminus of all CPT I proteins (Fig. 2). Finally, longer horizontal clusters often denote amphiphilic a-helices. At least three amphiphilic a-helices (al, a2, and a3) seem to be present in the N-terminal domain of the liver CPT I isoforms (Fig. 2). They are successively comprised within residues 25^47, 76-102 and 123-161. For the muscle form, al and a3 are also present and are included within residues 25-52 and 125-161, respectively. Residues 76-104 of the M-CPT I proteins are enriched in proline (H) and glycine (u) (Fig. 2). Proline introduces the largest constraints in a polypeptide chain and is considered to be a systematic break in the clusters. Proline often stops or distorts helices and P-strands whereas glycine has a large conformational flexibility. Thus, residues 76-104 are unlikely to adopt an a-helix but could be considered as a hinge region between H1 and H2. Whether this structural discrepancy between the liver and muscle isoforms has some repercussions on the functional properties of the enzymes remains to be established. In conclusion, the use of the HCA method reveals that the N-terminal domains of the four known CPT I proteins not only show a high degree of amino acid identity but also exhibit similar structural segmentations (Fig. 2). We propose the following prediction for the secondary structural segmentation of the N-terminal domain of the CPT Ifamily: pi-P2-al-Hl-a2 or loop-H2-a3. As other secondary structure prediction methods, the 2-D structural informations obtained by the HCA method leads to the elaboration of a working 2-D model that could be helpfull for the comprehension of the structure-function relationships of the CPT I family.

As described above, L-CPT I contains two hydrophobic transmembrane segments. The first import mechanism hypothesis which can be formulated is that either of the transmembrane segments could function as a signal anchor sequence. Whether H1 and H2 play an equivalent role or only one of them acts as a specific signal anchor sequence still remains to be elucidated. Alternatively, H1 or H2 could function as a stop-transfer sequence in cooperation with a matrix-signal sequence which could be ensured by either al, a2 or a3. Thus, further studies will be required to determine whether OMM insertion of L-CPT I follows the signal anchor sequence or stop-transfer model.

0 0

Post a comment