cDNA cloning, expression, and mutagenesis study of leukotriene B4 12-hydroxydehydrogenase.

Leukotriene B4 12-hydroxydehydrogenase catalyzes the conversion of leukotriene B4 into its biologically less active metabolite, 12-oxo-leukotriene B4. This is an initial and key step of metabolic inactivation of leukotriene B4 in various tissues other than leukocytes. Here we report the cDNA cloning for porcine and human enzymes from kidney cDNA libraries. A full-length cDNA of the porcine enzyme contains an open reading frame consisting of 987 base pairs, corresponding to 329 amino acids. The human enzyme showed a 97.1% homology with the porcine enzyme. Northern blotting of human tissues revealed its high expression in the kidney, liver, and intestine but not in leukocytes. The porcine enzyme was expressed as a glutathione S-transferase fusion protein in Escherichia coli, which exhibited similar characteristics with the native enzyme. Because the enzymes have a homology, in part, with NAD(P)+-dependent alcohol dehydrogenases, a site-directed mutagenesis study was carried out. We found that three glycines at 152, 155, and 166 have crucial roles in the enzyme activity, possibly by producing an NADP+ binding pocket.

Leukotriene B 4 12-hydroxydehydrogenase catalyzes the conversion of leukotriene B 4 into its biologically less active metabolite, 12-oxo-leukotriene B 4 . This is an initial and key step of metabolic inactivation of leukotriene B 4 in various tissues other than leukocytes. Here we report the cDNA cloning for porcine and human enzymes from kidney cDNA libraries. A full-length cDNA of the porcine enzyme contains an open reading frame consisting of 987 base pairs, corresponding to 329 amino acids. The human enzyme showed a 97.1% homology with the porcine enzyme. Northern blotting of human tissues revealed its high expression in the kidney, liver, and intestine but not in leukocytes. The porcine enzyme was expressed as a glutathione S-transferase fusion protein in Escherichia coli, which exhibited similar characteristics with the native enzyme. Because the enzymes have a homology, in part, with NAD(P) ؉ -dependent alcohol dehydrogenases, a site-directed mutagenesis study was carried out. We found that three glycines at 152, 155, and 166 have crucial roles in the enzyme activity, possibly by producing an NADP ؉ binding pocket.
Leukotriene B 4 (LTB 4 ) 1 is a potent chemotactic and proinflammatory factor produced in various tissues (1)(2)(3)(4). Arachidonic acid, released from the cell membrane by various stimuli, is converted to 5-hydroperoxyeicosatetraenoic acid and LTA 4 by 5-lipoxygenase (5)(6)(7)(8). LTB 4 is biosynthesized from LTA 4 by the action of LTA 4 hydrolase (9 -13). In human polymorphonuclear leukocytes, LTB 4 is converted to 20-hydroxy-LTB 4 by a cytochrome P-450 LTB 4 and further to 20-carboxy-LTB 4 (14 -17). The cDNA of cytochrome P-450 LTB 4 was cloned, and the mRNA is detected only in human leukocytes (18,19). LTB 4 is also produced in tissues other than leukocytes (20,21). We reported an alternative pathway for LTB 4 in various porcine tissues and purified a cytosolic LTB 4 12-hydroxydehydroge-nase from the porcine kidney (22). This enzyme converts LTB 4 to 12-oxo-LTB 4 in the presence of NADP ϩ . 12-Oxo-LTB 4 is at least 100 times less potent than LTB 4 in increasing intracellular calcium concentrations in human leukocytes (22). However, the molecular structure of the enzyme as well as its tissue distribution have not been known. Here we report the primary structures of porcine and human LTB 4 12-hydroxydehydrogenases and the putative NADP ϩ -binding domain. We clearly showed that the enzyme is expressed in the kidney, liver, and various tissues but not in leukocytes. Thus, this enzyme represents one of the major pathways of the metabolic inactivation of LTB 4 in tissues other than leukocytes.
N-terminal and Internal Amino Acid Sequences of LTB 4 12-Hydroxydehydrogenase-The porcine LTB 4 12-hydroxydehydrogenase was purified as described previously (22). The purified enzyme (100 g) was digested with 5 g of trypsin in 100 mM Tris-HCl, pH 8.5, at 37°C for 8 h. Digested fragments were purified by reversed phase HPLC using a Pharmacia Smart System equipped with a RPC C 2 /C 18 column (2.1 ϫ 100 mm). The digested enzyme was injected onto a RPC C 2 /C 18 column previously equilibrated with 0.1% trifluoroacetic acid in water and eluted by a linear gradient to 80% acetonitrile with 0.1% trifluoroacetic acid for 3.8 ml at a flow rate of 100 l/min. The eluted peptide fragments were monitored at 215 nm, and 31 fractions were collected. LTB 4 12-hydroxydehydrogenase (5 g) and six of 31 peptide fragments (Fractions 8,19,24,25,27,and 37) were loaded on polyvinlidene difluoride membranes with Prospin (Perkin Elmer) and sequenced by Edman degradation using an automated protein sequencer PPSQ-10 (Shimadzu, Kyoto). SWISS PROT protein data base was used to search for homologous proteins using a BLAST program (23).
cDNA Cloning of Porcine LTB 4 12-Hydroxydehydrogenase-Degenerative reverse transcriptase-polymerase chain reaction using mixed oligonucleotide primers was performed to obtain a partial cDNA fragment for screening of the library. Mixed oligonucleotide primers were designed according to the amino acid sequences of N-terminal and Fraction 19. Each primer was synthesized by Sawaday Technology (Tokyo), and the sequences of sense and antisense primers were 5Ј-GTGCGCGCCAAGTCCTGGACCCTGAA(A/C)AA(A/C)CA(T/C)TT(T/C) GT-3Ј (corresponding to N-terminal, 38 mers) and 5Ј-GCGGGCCACCT-GCTC(A/G/C/T)CCCATCATCAT(A/G)TC-3Ј (corresponding to the peptide of Fraction 19, 30 mers), respectively.
The conditions of polymerase chain reaction were as follows: denaturation at 94°C for 1 min, annealing at 50°C for 2 min, and elongation at 72°C for 3 min. After 5 cycles, the annealing temperature was changed to 55°C. After 30 cycles of polymerase chain reaction, the products were ethanol-precipitated and separated on an 1% agarose gel, and 4 different bands were recovered from the gel using a QIAGEN gel purification kit. Each band was ligated into a T-vector (Promega) by a T 4 DNA ligase, and the resulting constructs were transformed into Escherichia coli strain JM 109 (Competent high, TOYOBO, Tokyo). Plasmids were purified by an alkaline lysis method and sequenced with an ABI automated DNA sequencer 373A (Perkin Elmer). A band of 220 base pairs encoded the 5Ј end of the cDNA and was used as a probe to screen the library.
An oligo(dT)-primed Zap-II (Stratagene) porcine kidney cDNA library was constructed from 4 g of poly(A) ϩ RNA with Superscript II Choice System (Life Technologies Inc.) according to the manufacturer's manual. The library yielded 1.6 ϫ 10 6 independent clones. Full-length cDNA clones were obtained by a plaque hybridization method. 1.0 ϫ 10 6 clones were transferred to 10 sheets of Hybond N ϩ filters, and then the filters were alkaline-denatured and fixed by baking at 80°C for 2 h. The insert cDNA was digested out from the vector, randomly labeled by [ 32 P]dCTP using a Multiprime Labeling System (Amersham Corp.), and used as a probe for hybridization. After hybridization in Rapidhybri solution (Amersham Corp.) at 65°C for 8 h, each filter was washed extensively three times in 0.1 ϫ SSC, 0.1% SDS at 65°C for 20 min. Three rounds of screening gave three positive clones named pBDH 9, 14, and 15. Each clone was excised in vivo into a pBluescript II SK(Ϫ) phagemid by ExAssist helper phage (Stratagene), mapped using various restriction enzymes, and sequenced as described previously. All the clones showed the same restriction patterns, and sequencing confirmed that these three clones code for full-length cDNAs of LTB 4 12-hydroxydehydrogenase. Ten deletion mutants were prepared by exonuclease III from pBDH 15, and both strands were sequenced. In addition, six internal sequencing primers were synthesized, and the sequences were confirmed.
cDNA Cloning of Human LTB 4 12-Hydroxydehydrogenase-Human cDNAs of LTB 4 12-hydroxydehydrogenase were isolated from a human kidney gt 11 cDNA library (Clonetech) by a cross-hybridization method with a porcine full-length cDNA (pBDH 15) as a probe. 6 ϫ 10 5 clones were transferred to Biodyne nylon membranes (Pall) and hybridized at 55°C for 12 h with a [ 32 P]dCTP-labeled full-length porcine cDNA (pDBH 15). Each filter was washed three times in 2 ϫ SSC, 0.1% SDS at 55°C for 10 min. Three rounds of screening gave two phage clones named hBDH4 and 8, which were then purified, digested by EcoRI, subcloned into a pBluescript II SK(ϩ) vector, and sequenced. Homology search was performed against the GenBank, EMBL, and SCOP (structural classification of proteins) data bases using a BLAST program (23). The three-dimensional data of crystallized proteins were obtained from the SCOP data base and analyzed using a RASMOL program (25). An open reading frame and the deduced amino acid sequence were determined by a Genetyx Mac 6.0.2. software (Software Development, Tokyo, Japan).
Expression of LTB 4 12-Hydroxydehydrogenase as a GST Fusion Protein-The porcine cDNA insert was digested out from pBDH 15 by EcoRI and subcloned into a Pharmacia pGEX-1 expression vector (pGEX-LTB12DH). An E. coli strain JM-109 (TOYOBO) was transformed by heat shock, and then the recombinant protein was induced with 0.1 mM isopropyl-1-thio-␤-D-galactoside. The procedure was basically as described in the manufacturer's manual, except that the protein was induced at 20°C overnight with 0.1 mM isopropyl-1-thio-␤-D-galactoside. E. coli was collected, resuspended in PBS (Ϫ) containing 2 mM EDTA-Na 2 , 1 mM dithiothreitol, 0.1 mM phenylmethylsulfonyl fluoride, 0.2 g/ml pepstatin-A, 2 g/ml leupeptine, and disrupted by sonication. The sonicates were centrifuged for 10 min at 10,000 ϫ g, and 200 l (for 1 liter of E. coli culture) of GSH-sepharose (Pharmacia) was added to the supernatant. After washing with PBS (Ϫ), the protein was eluted in 50 mM Tris-HCl, pH 8.0, containing 10 mM GSH, and the purity was checked by SDS-PAGE. The 7.5% polyacrylamide gel was stained with Coomassie Brilliant Blue G, and the quantity of the recombinant protein was measured by scanning with bovine serum albumin as a standard. The enzyme activity of the recombinant protein was measured as described previously (22).
Peptide Antibody against LTB 4 12-Hydroxydehydrogenase-A peptide (ESLEETLKKASPEG, corresponding to the amino acid residues 197-210 of the porcine enzyme) was synthesized as a multiple antigen peptide (Fmoc MAP-peptide, 8-Branch, Applied Biosystems). An aliquot of 0.5 mg of the peptide was emulsified with an equal volume of Freund's complete adjuvant and injected into 3-9-month-old female New Zealand white rabbits. After three immunizations at one-month intervals, blood samples were collected and the serum was obtained by centrifugation. The anti-serum was purified by affinity chromatography. The recombinant LTB 4 12-hydroxydehydrogenase (1 mg) was coupled to 0.5 g of Epoxy-activated Sepharose (Pharmacia) in the coupling buffer (50 mM sodium-bicarbonate buffer, pH 9.0) at 25°C for 10 h. The Sepharose was loaded on a Poly-Prep Chromatography Column (Bio-Rad). After washing with 5 ml of the coupling buffer, 5 ml of the blocking buffer (50 mM Tris-HCl, 0.1 M ethanolamine, pH 8.0) was added to block the unbound resin. After washing with 10 ml of water, 10 ml of the elution buffer (0.1 M glycine-HCl, pH 2.5), and 10 ml of the wash buffer (20 mM Tris-HCl, 1 M NaCl, 1% Triton X-100, pH 7.5), the column was equilibrated with PBS (Ϫ). 2 ml of anti-serum was applied on the column and allowed to stand at 25°C for 1 h. After washing the column with 10 ml of PBS (Ϫ), 30 ml of the wash buffer, 30 ml of PBS (Ϫ), 30 ml of 0.15 M NaCl, the antibody was eluted in 2 ml of the elution buffer. The eluate was immediately neutralized with 100 l of 1 M Tris-HCl, pH 8.0. The concentration of the purified antibody was 236 g/ml. The affinity-purified antibody is termed ␣2 antibody hereafter.
Site-directed Mutagenesis of the Putative NADP ϩ -binding Domain-A mutagenesis study was performed by an oligonucleotide-derived mutagenesis method (26) using a Transformer site-directed mutagenesis kit (Clonetech). The mutagenetic primers were designed as follows: Each mutagenetic primer (10 ng) and a selection primer (10 ng, Aat II/EcoRV, 5Ј-GTGCCACCTGATATCTAAGAAACC-3Ј) were annealed simultaneously to 10 ng of pGEX-LTB12DH, and the first strand was synthesized with 4 units of T 4 DNA polymerase and 6 units of T 4 DNA ligase in 30 l at 37°C for 2 h. AatII (20 units) was added to selectively linearize the parental DNA. 40 l of the electrocompetent BMH71-18 mutS strain (Clonetech, CA) was transformed with 2 l of 5ϫ diluted reaction mixture using a Gene Pulser Unit (Bio-Rad). The condition of electroporation was 1.8 kV, 25 microfarad, 100 ⍀. After shaking the culture in 10 ml of TB medium overnight, the plasmids were recovered by an alkaline lysis method, and 100 ng of plasmids were digested with AatII (10 units) again. JM 109 cells were transformed with 10 ng of digested plasmids by heat shock, and colonies were isolated. Each mutated plasmid was sequenced entirely to check for unexpected mutations. The mutant proteins were purified as GST fusion proteins as described previously. Purified proteins (1 g/lane) were separated on a 7.5% SDS-PAGE gel and transferred to a Hybond ECL membrane (Amersham Corp.). It was blotted with ␣2 antibody (200 ϫ dilution) or rabbit anti-GST antibody (Pharmacia) as the first antibody and visualized using an Amersham ECL system.
The V max and K m values against LTB 4 and NADP ϩ were determined as described previously (22) six times in three independent experiments. 4 12-Hydroxydehydrogenase-Screening of 1.0 ϫ 10 6 porcine clones with the probe coding for the 5Ј end gave three independent positive clones, pBDH 9, 14, and 15. Three clones were excised in vivo into pBluescript II SK(Ϫ) and mapped using several restriction enzymes. All the inserts gave an identical restriction map, and DNA sequencing confirmed that these three clones coded for full-length cDNAs of LTB 4 12-hydroxydehydrogenase. pBDH 15 was further sequenced by deletion with exonuclease III and 6 internal sequencing primers. Screening of a human kidney cDNA library (6 ϫ 10 5 clones) with pBDH 15 gave two independent clones, hBDH 4 and 8, which were identical. The primary structures of porcine and human LTB 4 12-hydroxydehydrogenases are shown in Figs. 1 and 2. The deduced amino acid sequences of the porcine enzyme contained all the amino acid sequences of seven peptide fragments obtained from the native porcine kidney enzyme (Fig. 1). The cDNA of pBDH 15 contained a polyadenylation signal after the stop codon (Fig. 1), showing that it codes for a full-length LTB 4 12-hydroxydehydrogenase. pBDH 15 contains an open reading frame of 987 base pairs and coded for 329 amino acids. The calculated M r of the porcine enzyme is 35,761, a value similar to that of the native enzyme (22). Because hBDH 4 and 8 lack the stop codon, the human enzyme seems to have additional amino acids in the C-terminal. Several trials to acquire the full-length clones using rapid amplification of cDNA ends were unsuccessful. The identity between the porcine and human enzymes was 83.5% at the amino acid level and 84.7% at the nucleotide level. Amino acid homology was 97.1%. Both porcine and human enzymes showed a high homology (94.5 and 96.1%, respectively) with a previously reported rabbit protein, AdRab-F, which is expressed only in adult rabbit small intestine but not in the baby (27). Function of the AdRab-F protein has not been documented (27), but it seems to be a rabbit homologue of LTB 4 12-hydroxy-dehydrogenase judging from the high homology. These three proteins contain a proline-rich motif (250 -257 residues) in the C-terminal half (Figs. 1 and 2).

cDNA Cloning of Porcine and Human LTB
In addition, LTB 4 12-hydroxydehydrogenases have a weak homology with NAD ϩ /NADP ϩ -dependent short chain alcohol dehydrogenases (28, 29) and a -crystallin (30), identity being 30 -35%. Especially, a fragment from 149 to 166 of the porcine LTB 4 12-hydroxydehydrogenase has a relatively high homology (ϳ50%) with these dehydrogenases. Because this domain is considered to be a NAD ϩ /NADP ϩ -binding domain in these dehydrogenases (28,29), a mutagenesis study was carried out to determine the putative NADP ϩ -binding domain of LTB 4

12hydroxydehydrogenase (see below).
Northern Blot Analysis- Fig. 3 shows the tissue distribution of mRNA of LTB 4 12-hydroxydehydrogenase in human tissues. The mRNA is expressed most abundantly in the kidney and liver, followed by small intestine and colon. It was absent in human leukocytes. The distribution of mRNA matches the tissue distribution of the enzyme activities studied in various porcine tissues (22).
Expression of LTB 4 12-Hydroxydehydrogenase as a GST Fusion Protein-The recombinant porcine LTB 4 12-hydroxydehydrogenase was overexpressed as a GST fusion protein in the E. coli system. When the transformed E. coli was cultured at 37°C, most of the recombinant protein was precipitated by centrifugation at 10,000 ϫ g, possibly existing in bacterial inclusion bodies. By decreasing the culture temperature to 20°C, good yields of a soluble protein were obtained. The recombinant protein was purified by affinity column chromatography using a GSH-sepharose column. A typical yield was 3 mg of protein from 1 liter of bacterial culture, and the purity was around 70% judging from SDS-PAGE. The purified GST fusion protein exhibited characteristics similar to the native enzyme, with a V max value of forming 6 nmol of 12-oxo-LTB 4 / min/mg fusion protein. The K m values of the recombinant enzyme were 20 M against LTB 4 and 10 M against NADP ϩ . After digesting out GST from the fusion protein with thrombin, the specific activity of the enzyme remained unchanged. GST only had no apparent enzyme activity (date not shown). These results indicate that the cDNA of pBDH 15 codes for LTB 4

12-hydroxydehydrogenase.
Site-directed Mutagenesis of the Putative NADP ϩ -binding Domain-A computer-assisted homology analysis revealed that the fragment 149 -166 of LTB 4 12-hydroxydehydrogenase was homologous to the NAD ϩ /NADP ϩ -binding domain of other short chain alcohol dehydrogenases (28,29). To determine whether this domain is essential for binding to NADP ϩ and for the enzyme activity, a site-directed mutagenesis study was done. Ala 149 , Ala 150 , Gly 152 , Gly 155 , Gly 159 , or Gly 166 was con-verted to Val or Glu, and the mutant proteins were expressed in E. coli as GST fusion proteins. The recombinant proteins were purified with a GSH-sepharose column and quantified on Coomassie Brilliant Blue G-stained SDS-PAGE gels, and the enzyme activities were measured. The K m and V max values were determined by changing the concentrations of LTB 4 and NADP ϩ .

FIG. 2. Amino acid alignment of LTB 4 12-hydroxydehydrogenases and AdRab-F protein.
The human sequence is considered to be partial. The hypothetical AdRab-F protein (27) was also aligned as rabbit. The asterisk indicates amino acids that are identical among three species. ⅐ indicates amino acids that are identical in two species. A similar conversion of LTB 4 was reported in the human lung (49), kidney (50), keratinocytes (51), and the guinea pig kidney and liver. 2 The enzyme purified from the porcine kidney is a monomeric protein with an M r of 35,000 and an isoelectric point over 9.5. It specifically recognizes the 12(R)-hydroxymoiety of LTB 4 , thus it was named LTB 4 12-hydroxydehydrogenase (22). The product, 12-oxo-LTB 4 , was at least 100 times less potent in increasing the intracellular calcium concentration in human leukocytes (22). Recently, the method of the chemical synthesis of 12-oxo-LTB 4 was established (52), thus enabling us to determine the biological activity and precise metabolism of this compound.
In the present study, we cloned a cDNA for the porcine LTB 4 12-hydroxydehydrogenase by screening a kidney cDNA library using a probe obtained from its partial amino acid sequences of the purified enzyme (Fig. 1). Porcine LTB 4 12-hydroxydehydrogenase cDNA contained an open reading frame of 987 base pairs and coded for 329 amino acids. The deduced amino acid sequence contained all the sequences from Lys-C-digested peptide fragments (Fig. 1). The calculated M r of the porcine en-zyme is 35,761, which agrees well with that of the native enzyme. In addition, we obtained a cDNA of the human enzyme by cross-hybridization with the porcine cDNA. The primary structures of the porcine and human enzymes are similar, with an amino acid homology of 97.1%. Both enzymes were highly homologous (94.5 and 96.1%) with a rabbit AdRab-F hypothetical protein (Fig. 2), the mRNA of which was expressed only in the adult rabbit and not in the baby (27). The function of AdRab-F protein has not been reported, but it seems to be a rabbit homologue of LTB 4 12-hydroxydehydrogenase. Further studies are required to determine the developmental change of the expression of LTB 4 12-hydroxydehydrogenase.
The tissue distribution of mRNA of human enzyme corresponds well to the distribution of the enzyme activities studied in porcine tissues (22), with the highest expression in the kidney and liver, followed by colon and small intestine (Fig. 3). It is important to note that the mRNA is not expressed in the human leukocytes where the -oxidation pathway is present. LTB 4 12-hydroxydehydrogenases were homologous with other NAD ϩ /NADP ϩ -dependent short chain alcohol dehydrogenases. Although the total homology was 35% or less, there was a relatively highly homologous domain (Fig. 6). Among these homologous proteins, three enzymes were crystallized, and the structures were well studied (53)(54)(55). Crystal structure analyses revealed that this domain forms a compact ␤-sheet-␣-helix-␤-sheet structure and was determined to form a NAD ϩ / NADP ϩ -binding domain (Fig. 6). An acidic residue adjacent to this domain is supposed to bind to the 2Ј and 3Ј hydroxyl groups of the adenine ribose of NAD ϩ /NADP ϩ (29). In addition, mutagenesis studies of the other dehydrogenases indicate that the GXGXX(G/A)XXXGXXXXXXG consensus sequence is important to maintain a close contact between the coenzyme and the enzyme by forming an ␣-helix structure (29,56). By changing two Gly in this domain of NAD ϩ -dependent pyruvate dehydrogenase to Ala, the enzyme activity was decreased (57). This domain is highly conserved in the porcine and human LTB 4 12-hydroxydehydrogenases and AdRab-F hypothetical protein, and the consensus sequence is 149 AAXGXXGXXXGXXX-XXXG 166 (Figs. 2 and 6).
To determine which amino acids are required for the enzyme activity, a site-directed mutagenesis was carried out. Ala 149 , Ala 150 , Gly 152 , Gly 155 , Gly 159 , and Gly 166 were changed into Val, which has a longer side chain than Ala and Gly, or to Glu, which is negatively charged, and the enzyme activities were measured. M6 (G166V) and M7 (A149V, A150V, G152V, G155V, and G159V) mutants readily cleaved into shorter peptides, as shown in Fig. 4. The K m values against LTB 4 and NADP ϩ of the recombinant wild type enzyme were 20 M and 10 M, respectively. Because these values of the native enzyme purified from the porcine kidney were 10 M and 1 M (22), respectively, the recombinant protein may contain a slight change in the three-dimensional structure. To exclude the influence of the enzyme instability and degradation, the quantities of mutant proteins were standarized on Coomassie Brilliant Blue G-stained SDS-PAGE gels (Fig. 4). The enzyme activities of mutant proteins were measured as the relative activities toward the wild type enzyme. Similar results were obtained from the mutagenesis study in the short-chain alcohol dehydrogenase (58). These results suggest that the longer side chain of Val may inhibit the NADP ϩ binding in G152V, G155V, and G166V mutants. Because the enzyme activity remains partially in A149V and G159V mutants, the conformational change of the binding pocket might be moderate in these mutants. In contrast, by changing Ala 149 to Glu, most of the enzyme activity was lost, possibly due to the negative charge of Glu. These results indicate that Gly 152 , Gly 155 , and Gly 166 of LTB 4 12-hydroxydehydrogenase are essential for the enzyme activity, probably by forming an NADP ϩ binding pocket (Fig. 6). As seen in Fig. 6, these three Gly are well conserved among LTB 4 12-hydroxydehydrogenases and other homologous proteins, suggesting that these Gly are important. Ala 149 and Gly 159 seem to play some roles in the enzyme activity but are not essential. Ala 150 seems to have only a little role in the enzyme activity.
There is a proline-rich motif that is conserved among three species in the C-terminal half of LTB 4 12-hydroxydehydrogenase (Fig. 2, 250 -257 residues). The proline-rich motif was reported to play crucial roles by binding src homology 3 (SH3) domains in the signal transduction system of tyrosin-kinase type receptors (59). Recently, the binding of proline-rich domains to SH3 domain was reported to be involved in the translocation and activation of 5-lipoxygenase (60), which catalyzes the initial step of biosynthesis of leukotrienes. The role of the proline-rich domain of LTB 4 12-hydroxydehydrogenase remains to be clarified.
In conclusion, LTB 4 12-hydroxydehydrogenase cDNAs were isolated from the porcine and human kidney, and their primary structures were identified. Northern blotting revealed that the mRNA was expressed in the kidney, liver, small intestine, and colon but not in leukocytes. By a site-directed mutagenesis study, we found that three Gly residues at 152, 155, and 166 play important roles in the enzyme activity. The acquisition of the cDNA and the antibody paves the way for the further analysis of the cellular localization and the biological significance of the enzyme under various physiological and pathological conditions.