Hidden localization motifs: naturally occurring peroxisomal targeting signals in non-peroxisomal proteins

Table 1 Results of the yeast-two hybrid interaction assays with PEX5

		Yeast PEX5			Human PEX5
Species	Accession	Score*	Activity^† (Units/mg protein)	Standard deviation	Score*	Activity^† (Units/mg protein)	Standard deviation	Carboxyl terminus	Description
Canis familiaris	P81708	-	-	-	0.17	25	2	HCKGKDLSKYLASCNL	Lysozyme
Drosophila melanogaster	P13368	-	-	-	6.70	29	11	PLKDKQLYANEGVSRL	Sevenless protein
Gallus gallus	P00698	-	-	-	2.02	73	4	RCKGTDVQAWIRGCRL	Lysozyme
Rana nigromaculata	Q04604	-	-	-	0.13	91	15	LLMEAEDYQATYQSNL	Tyrosinase
Homo sapiens	P14679	-	-	-	4.01	242	10	LLMEKEDYHSLYQSHL	Tyrosinase
Bos taurus	P80209	-	-	-	7.04	310	58	FDRDQNRVGLAEAARL	Cathepsin D
Saccharomyces cerevisiae	P12687	2.72	482	37	-	-	-	KVEVIARSRRAFLSKL	Mitochondrial ribosomal protein L2, or MRP7
Synthetic construct	DHFR-SKL	11.51	195	45	-	-	-	EKGIKYKFEVYEKSKL	DHFR-SKL
Escherichia coli	P23893	4.81	270	26	11.35	473	57	DINNTIDAARRVFAKL	Glutamate-1-semialdehyde 2,1-aminomutase
E. coli	P78258	-9.46	164	31	5.59	566	70	FAVDQRKLEDLLAAKL	Transaldolase A
Methanopyrus kandleri	NP_613646	6.08	45	8	10.41	358	46	GMGRREGHPDVGPARL	Riboflavin synthase
Archaeoglobus fulgidus	NP_070998	7.57	206	19	-1.36	0	NA	EEVIRKIAEGLNKAKF	2-nitropropane dioxygenase

All eukaryotic target sequences (characterized by species, SWISS-PROT or NCBI-Refseq accession number, score from the PTS1 predictor [7], carboxy-terminal sequence and description) were tested for interaction with the tetratricopeptide (TPR) repeat domain of human PEX5, except for P12687 and DHFR-SKL where the corresponding TPR domains were derived from yeast PEX5. The prokaryotic proteins were assayed using PEX5 from both yeast and human. As the estimated length of the PTS1 signal is 12 carboxy-terminal residues [13], we chose the carboxy-terminal 16-mers to be sure that we have included the complete motif-carrying segment. *A PTS1 prediction score above zero is considered predictive of a functional PTS1 signal; a score between -10 and 0 is considered a 'twilight zone' prediction. It should be noted that the negative score for the DHFR-SKL carboxyl terminus in its context is generated by the PTS1 predictor [7] solely by terms that evaluate its potential accessibility for PEX5. ^†A yeast-two hybrid assay is considered positive if the measured β-galactosidase activity is clearly greater than zero. Experience from previous test series suggests a lower limit of around 10 Miller Units per mg protein [12] for the detection of a productive interaction. The measured β-galactosidase activities (including standard deviations) range from weak (P81708, P13368) to strong (P80209, P12687).

ISSN: 1474-760X