Front Matter

LOX. The exchanged region contained the amino acids I418 and M419, which had

been previously characterized as primary determinants for the positional specificity

of the human 15-LOX (Sloane et al., 1995). The chimeric enzyme converted AA

mainly to 15-HETE, indicating that important sequence determinants for the posi-

tional specificity were present in the fragment inserted. Using the inverse strategy

(insertion of a comparable fragment of the porcine 12-LOX into the rabbit 15-LOX),

the rabbit 15-LOX was converted to a 12-lipoxygating species. A similar effect was

seen when a comparable piece of the murine enzyme was inserted into the rabbit

LOX. Next, the size of the inserted fragment was gradually reduced; the results

obtained suggested that additional sequence determinants, which are different

from those described before (I418 and F419 of the human 15-LOX), appear to

be important for the positional specificity of the murine 12-LOX. These determi-

nants must be localized in an 86-amino acid fragment (amino acids 270 to 355)

of the murine leukocyte-type 12-LOX (Borngra ̈ber et al., 1996). Modeling the en-

zyme/substrate interaction of LOX using the X-ray coordinates for the soybean en-

zyme we found that only the seven C-terminal amino acids of this peptide fragment

may line the substrate-binding cavity (positions 349–355). Alignment of this short

amino acid sequence revealed just one conserved amino acid exchange (F353L)

among the LOXs of interest. In the leukocyte-type 12-LOXs of mice and rats

this amino acid is a small leucine, whereas a bulky phenylalanine is located at

this position in the human and rabbit 15-LOXs and in the porcine leukocyte-type

12-LOX. In order to test the hypothesis that amino acid 353 may constitute a se-

quence determinant for the positional specificity, site-directed mutagenesis was car-

ried out. Indeed, we found that the F353L mutation converted the rabbit enzyme to a

12-lipoxygenating species (12-/15-HETE ratio of 2 : 1). These data suggested that

the amino acids at positions 353, 418, and 419 form the bottom of the substrate-

binding cleft (Figure 9) and the following conclusions were drawn:

* Mammalian LOXs containing a small amino acid at position 353 are 12-LOX,

independent of the size of the residues 418 and 419 (rat and mouse leukocyte-

type 12-LOX).

* When a space-filling amino acid is localized at position 353, the residues 418 and

419 become important for the positional specificity.

* When bulky amino acids are localized at positions 353, 418 and 419, 15-lipoxy-

genation is favored (human and rabbit 15-LOX). However, when a bulky residue

at position 353 is combined with a less space-filling amino acid at positions 418

and 419, AA is oxygenated at C-12 (porcine leukocyte-type 12-LOX).

After the crystal structure of various plant and mammalian LOXs became available

(Boyington et al., 1993; Minor et al., 1996; Skrzypczak-Jankun et al., 1997; Gillmor

et al., 1998), the model of the substrate-binding pocket in mammalian LOXs was

substantially improved. In fact, it was calculated that the volume of the sub-

strate-binding pocket of 12-LOXs is about 6 % bigger than that of 15-LOXs

(Skrzypczak-Jankun et al., 1997; Browner et al., 1998). Interestingly, 5-LOX

have an even larger substrate-binding pocket (20 % larger than 15-LOXs). More-

over, the X-ray coordinates predicted that I593 may be involved in defining the

size and shape of the substrate-binding cleft. To test this hypothesis, site-directed

15.3 The structural bases of the positional specificity of LOXs 325