HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) An addition or correction has been published Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jones, J. M. Articles by Gould, S. J. Search for Related Content PubMed PubMed Citation Articles by Jones, J. M. Articles by Gould, S. J. Social Bookmarking                      What's this?

J Biol Chem, Vol. 275, Issue 17, 12590-12597, April 28, 2000

Identification and Characterization of HAOX1, HAOX2, and HAOX3, Three Human Peroxisomal 2-Hydroxy Acid Oxidases^*

Jacob M. Jones, James C. Morrell, and Stephen J. Gould

From the Department of Biological Chemistry, The Johns Hopkins University School of Medicine, Baltimore, Maryland 21205

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Computer-based approaches identified three distinct human 2-hydroxy acid oxidase genes, HAOX1, HAOX2, and HAOX3, that encode proteins with significant sequence similarity to plant glycolate oxidase, a prototypical 2-hydroxy acid oxidase. The products of these genes are targeted to peroxisomes and have 2-hydroxy acid oxidase activities. Each gene displays a distinct tissue-specific pattern of expression, and each enzyme exhibits distinct substrate preferences. HAOX1 is expressed primarily in liver and pancreas and is most active on the two-carbon substrate, glycolate, but is also active on 2-hydroxy fatty acids. HAOX2 is expressed predominantly in liver and kidney and displays highest activity toward 2-hydroxypalmitate. HAOX3 expression was detected only in pancreas, and this enzyme displayed a preference for the medium chain substrate 2-hydroxyoctanoate. These results indicate that all three human 2-hydroxy acid oxidases are involved in the oxidation of 2-hydroxy fatty acids and may also contribute to the general pathway of fatty acid -oxidation. Primary hyperoxaluria type 1 (PH1) is caused by defects in peroxisomal alanine-glyoxylate aminotransferase, the enzyme that normally eliminates intraperoxisomal glyoxylate. The presence of HAOX1 in liver and kidney peroxisomes and the ability of HAOX1 to oxidize glyoxylate to oxalate implicate HAOX1 as a mediator of PH1 pathophysiology.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The oxidative metabolism of fatty acids can proceed by either of two distinct pathways, -oxidation or -oxidation (1). Fatty acid -oxidation, which shortens the carbon chain by two-carbon increments, generates cellular energy through the release of reducing equivalents and acetyl-CoA. In contrast, fatty acid -oxidation shortens fatty acids in one-carbon increments, and its contribution to energy production, if any, is unknown. One function of fatty acid -oxidation in humans appears to be in converting 3-methyl branched fatty acids, such as phytanic acid, to 2-methyl branched fatty acids, which can then be oxidized further by peroxisomal and mitochondrial -oxidation pathways. Complete -oxidation pathways have been detected in organisms as diverse as plants and mammals (2, 3).

In humans, the first steps the in the -oxidation of fatty acids involve their activation to CoA esters (4) and the subsequent hydroxylation by a dioxygenase, PAHX, generating 2-hydroxyacyl CoAs (5-7). The downstream steps of fatty acid -oxidation are less certain, with two models attracting most support. The currently favored model invokes a 2-hydroxyacyl-CoA lyase to generate formyl CoA and a shortened fatty aldehyde, which can then be dehydrogenated to regenerate a fatty acid (1, 8-11). Alternatively, some reports indicate that fatty acid -oxidation may involve the generation of a 2-keto intermediate (12-15). The 2-keto acid could then undergo decarboxylation to yield a fatty aldehyde, which could then be dehydrogenated to the corresponding acid.

Although it is unclear whether 2-hydroxy acid oxidases contribute to a general mechanism of fatty acid -oxidation, there is abundant evidence that these enzymes can oxidize a broad range of 2-hydroxy acids, ranging from glycolate to long chain 2-hydroxy fatty acids such as 2-hydroxypalmitate (16-18). These enzymes utilize a flavin cofactor and convert 2-hydroxy acids to 2-keto acids with the concomitant reduction of molecular oxygen to hydrogen peroxide. The most well studied enzyme of this family is the plant glycolate oxidase, an enzyme with an essential role in photorespiration (19-21). Glycolate oxidase catalyzes the oxidation of glycolate to glyoxylate as well as the oxidation of glyoxylate to oxalate.

Mammals also contain significant 2-hydroxy acid oxidase activities. These have been detected in both liver and kidney (16-18, 22, 23). The liver isozyme has been reported to be most active on glycolate, while the kidney isozyme is thought to prefer longer chain substrates. Although two distinct rodent genes encoding 2-hydroxy acid oxidases have been described (24-26), there is as yet no thorough analysis of human 2-hydroxy acid oxidases, their patterns of expression, or their substrate specificities. We report here the results of a search for human 2-hydroxy acid oxidases. We identified three human homologs of plant glycolate oxidase, HAOX1, HAOX2, and HAOX3. Each gene encodes a peroxisomal protein with 2-hydroxy acid oxidase activity. However, each gene has a distinct, tissue-specific pattern of expression, and each protein has distinct substrate preferences. These results and their implications for 2-hydroxy acid metabolism are discussed.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Plasmids and Strains-- Candidate HAOX cDNAs were identified by scanning the data base of expressed sequence tags for human cDNAs capable of encoding proteins with sequence similarity to spinach glycolate oxidase (GOX). Multiple cDNA clones were identified, and these corresponded to three distinct genes, which we refer to as HAOX1, HAOX2, and HAOX3. The HAOX1 cDNA was assembled from two overlapping cDNA clones as follows: a BamHI/SalI fragment was excised from the cDNA clone IMAGE: 84824 (GenBank^TM accession no. T74667) and ligated between the BamHI and SalI sites of the cDNA clone IMAGE: 77520 (GenBank^TM accession no. T58830). The HAOX2 cDNA was assembled in a similar manner, by inserting an Asp718/NotI fragment of the cDNA clone IMAGE: 434823 (GenBank^TM accession no. AA703164) between the Asp718 and NotI sites of the cDNA clone IMAGE: 245519 (GenBank^TM accession no. N72519). The entire HAOX3 open reading frame (ORF)¹ was present on a single cDNA clone, IMAGE: 531512 (GenBank^TM accession no. AA075839).

Two forms of the HAOX1 ORF were amplified from the complete HAOX1 cDNA. The HAOX1 ORF was amplified using the primers 5'-CCCGGATCCATGCTCCCCCGGCTAATTTCTATC-3' and 5'-CCCGCGGCCGCTCACAACTACGAAACGGCCAAAGGATTTTTC-3'. A form of the HAOX1 ORF lacking the final two codons (HAOX1KI) was also amplified using the first primer above together with the primer 5'-CCCGCGGCCGCTCACATCTACGAAACGGCCAAAGGATTTTTC-3'. Both sets of primers append BamHI and NotI sites (underlined) to the 5'- and 3'-ends of the HAOX1 ORF, respectively. The PCR product from each reaction was digested with BamHI and NotI and cloned between the BamHI and NotI sites of pMBP (27), a modified form of the pMALc2 expression vector (New England Biolabs) to make pMBP-HAOX1 and pMBP-HAOX1KI, respectively. The HAOX1 ORF was excised from pMBP-HAOX1 and inserted between the BamHI and NotI sites of pcDNA3-Nmyc (28), a mammalian expression vector designed to express proteins in fusion with an amino-terminal c-myc epitope, creating pcDNA3-Nmyc/HAOX1. The same fragment was also inserted between the BamHI and NotI sites of pET28a (Novagen), a bacterial expression plasmid designed to express proteins in fusion with an amino-terminal 6-histidine tag, creating pET28a-HAOX1.

The HAOX2 ORF was amplified using the primers 5'-CCCGTCGACCATGTCCTTGGTGTGTTTGACAGAC-3' and 5'-CAAGCGGCCGCTTACAGCCTGGAAAACTGGACCAAG-3'. These primers append SalI and NotI sites to the 5'- and 3'-ends of the HAOX2 ORF, respectively. The PCR product was digested with SalI and NotI and cloned between the XhoI and NotI sites of pcDNA3-Nmyc to make pcDNA3-Nmyc/HAOX2. The HAOX2 ORF was excised from this plasmid by digestion with BamHI and NotI and inserted into the corresponding sites of pET28a to make pET28a-HAOX2.

Two forms of the HAOX3 ORF were amplified from the HAOX3 cDNA. The HAOX3 ORF was amplified using the primers 5'-CCCGTCGACGATGTCTTTGCTGTGTGTTTGGCAG-3' and 5'-CCCGCGGCCGCTTATAATCTGGAGAACTGAATCAG-3'. A form of the HAOX3 ORF lacking the final two codons (HAOX3RL) was amplified using the oligonucleotide 5'-CCCGCGGCCGCTTATAATCAGGAGAACTGAATCAG-3' in conjunction with the first primer above. Both sets of primers append SalI and NotI sites to the 5'- and 3'-ends of the ORF, respectively. The HAOX3 PCR product was digested with SalI and NotI and inserted into the corresponding sites of pT7-His₆ (29) to make pT7-His₆/HAOX3. The HAOX3 ORF was then excised from pT7-His₆/HAOX3 by digestion with SalI and NotI and inserted into the XhoI and NotI sites of pcDNA3-Nmyc to make pcDNA3-Nmyc/HAOX3. The HAOX3RL product was digested with SalI and NotI and inserted into the corresponding sites in pMBP to make pMBP-HAOX3KL. The sequences of all polymerase chain reaction-generated clones were confirmed using automated fluorescent sequencing.

All bacterial manipulations were performed in the Escherichia coli strain DH10B (30) except expression of proteins from pET28a and pHis₆ vectors, which was performed in the strain BL21/DE3 (Novagen).

Northern Blot Analysis-- Northern blot analysis was performed using standard protocols (31) and human multitissue Northern blots from CLONTECH (Palo Alto, CA). High stringency wash conditions were used (0.2× SSC, 50% formamide, 45 °C) to reduce the possibility of cross-reactivity. The probe for HAOX1 transcript detection was derived from a polymerase chain reaction product of the entire HAOX1 ORF. Due to the high sequence similarity of HAOX2 and HAOX3, regions of the 3'-untranslated region (UTR) were used to make probes. For HAOX2, the HAOX2 cDNA clone was digested with DraIII and NotI, and a restriction fragment corresponding to the 3'-UTR was used as probe. An HAOX3-specific probe was generated by polymerase chain reaction using the primer 5'-CTACGAGATTGCCTACAAGAG-3' in conjunction with a primer designed to anneal to the T7 promoter site in the HAOX3 cDNA clone.

Transfections, Cell Lysates, Enzyme Assays, and Immunoblot-- For protein expression in human skin fibroblasts, GM5756-T cells were cultured and transfected as described (32). Two days after electroporation, cells were washed once in Hanks' balanced salt solution (Life Technologies), harvested by gentle scraping, washed once in 100 mM KP_i (pH 7.5), 10 mM EDTA, and then resuspended briskly in 500 µl of 100 mM KP_i, pH 7.5, 10 mM EDTA, 0.5% Triton X-100 for lysis. 2-Hydroxy acid oxidase activity was measured by a method similar to that of Schuman and Massey (22). Specifically, 50 µl of each lysate was added to a mixture of 100 mM KP_i, 10 mM EDTA, 100 µM 2,6-dichloroindophenol (DCIP). The reaction was started by the addition of a 500 µM concentration of the indicated substrate, and the change in A₆₀₅ was measured over time. Reactions were carried out in a final volume of 1 ml at room temperature (18 °C). An extinction coefficient of 110.26 µM¹ cm¹ for DCIP at 605 nm was used for all calculations. One unit of activity is defined as the reduction of 1 µmol of DCIP in 1 min. The protein concentration of each lysate was determined by Bio-Rad protein assay (New England Biolabs). For immunoblot, 10 µg of protein from each lysate was separated by polyacrylamide gel electrophoresis and transferred to a membrane. Detection was by standard protocols as described (32) using a commercially available anti-c-myc polyclonal antibody (Jackson Immunochemicals).

Expression, Purification, and Assay of Recombinant Proteins-- BL21/DE3 cells harboring the appropriate His₆-HAOX expression plasmid were grown overnight at 37 °C in 50-ml cultures of Luria broth supplemented with 25 µg/ml kanamycin. This culture was added to 1 liter of 2YT medium containing 25 µg/ml kanamycin and grown at 37 °C to an optical density of 0.4. The cultures were then cooled to room temperature (18 °C), induced by the addition of isopropyl-1-thio--D-galactopyranoside to a final concentration of 1 mM, and grown overnight at 18 °C with vigorous shaking. Cells were harvested and resuspended in 25 ml of binding buffer (20 mM NaP_i (pH 7.8), 500 mM NaCl, 5 mM benzamidine HCl) containing 1% Triton X-100, 8 µg/ml DNase, and 8 µg/ml RNase. Cells were frozen in liquid nitrogen, thawed, and lysed by mixing. Lysates were cleared by centrifugation at 25,000 × g for 30 min. The resulting supernatant was then diluted to 50 ml in binding buffer and applied at a rate of 1.5 ml/min to a 2-ml bed of chelating Sepharose fast-flow resin (Amersham Pharmacia Biotech) previously charged with NiCl and equilibrated in binding buffer. The bound resin was washed with 50 bed volumes of pH 7.8 binding buffer followed by 10 bed volumes of binding buffer containing 5 mM EDTA, 50 mM imidazole and adjusted to pH 6.0. an imidazole/EDTA step gradient was used to elute bound proteins. Specifically, three steps of 2.5 bed volumes each of pH 7.8 binding buffer containing 200 mM imidazole, 5 mM EDTA, 300 mM imidazole plus 10 mM EDTA, and 500 mM imidazole plus 50 mM EDTA were applied sequentially to the resin. Fractions were analyzed by polyacrylamide gel electrophoresis for the presence and purity of His₆-HAOX proteins. Fractions containing purified His₆-HAOX proteins were pooled, stabilized with glycerol, and stored at 4 °C. Substrate comparison studies were performed as for whole cell lysates using 15 µg of protein. For kinetic studies, 15 µg of protein was added to a mixture of 100 mM KP_i (pH 7.5), 10 mM EDTA, and 150 µM DCIP. To start the reaction, substrate was added in various concentrations ranging from 5 µM to 1 mM, and the change in A₆₀₅ was monitored with time.

Indirect Immunofluorescence and Fluorescence Microscopy-- Indirect immunofluorescence studies were performed on normal human skin fibroblasts (GM5756-T cells) or the PEX10-deficient fibroblast cell line, PBD100, which was derived from a Zellweger syndrome patient and has been described previously (33). Cell lines were cultured and transfected as described (32). After transfection, cells were transferred to cover glasses and incubated in complete medium at 37 °C for 1 day. The cells were fixed, permeabilized, and processed for indirect immunofluorescence as described (32). Permeabilization was normally performed for 5 min with 1% Triton X-100, which permeabilizes both the plasma and peroxisome membranes. For selective permeabilization of only the plasma membrane, cells were incubated for 5 min in 25 µg/ml digitonin instead of Triton X-100. Guinea pig anti-PMP70 antibodies were raised against a synthetic peptide corresponding to the C-terminal 18 amino acids of human PMP70 (34). The anti-c-myc mouse monoclonal antibody was derived from tissue culture medium of the mouse hybridoma line 1-9E10 (Roche Molecular Biochemicals). Fluorescent secondary antibodies were obtained from commercial sources.

Subcellular Fractionation Studies-- Subcellular fractionation of rat liver was as described by Mihalik (35). Anti-HAOX1 and anti-HAOX3 antibodies were raised in rabbits against MBP-HAOX1KI and MBP-HAOX3RL, respectively. Antibodies were affinity-purified using the appropriate MBP fusion protein coupled to CNBr-activated Sepharose (Amersham Pharmacia Biotech) as described (28). Antibodies obtained were stored in 100 mM Tris-HCl, pH 8.0, 25% glycerol, 0.04% NaN₃ at 4 °C. Immunoblots were performed as described by Crane et al. (36).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

HAOX1, HAOX2, and HAOX3 Are Homologous to Spinach Glycolate Oxidase-- The amino acid sequence of spinach GOX was used to scan the data base of expressed sequence tags for all human homologs. Multiple expressed sequence tags were identified, and these corresponded to three distinct genes, HAOX1, HAOX2, and HAOX3. The human HAOX1 cDNA (Fig. 1) is 1743 bp long and contains a 1113-bp ORF, with 24 bp of 5'-UTR and 606 bp of 3'-UTR. The sequence surrounding the indicated start codon (GTGAAAATGC) is a partial match to the Kozak start site consensus (GCC(A/G)CCATG(G/A)) (37). All subsequent possible start sites are also only in partial agreement with this consensus sequence until the methionine codon at position 312, roughly one-third of the way into the region of high amino acid identity with plant GOX. This, combined with the finding that two in-frame termination codons precede the proposed start site at positions 12 and 18 (Fig. 1, double underline), make the indicated methionine codon the most likely translation initiation site. A consensus polyadenylation site is present in the 3'-UTR (Fig. 1, boldface type). The deduced protein product is 370 amino acids long, has a predicted molecular mass of 41 kDa, and terminates in the sequence serine-lysine-isoleucine, which resembles the type 1 peroxisome targeting sequence (PTS1) (underlined) (38). The region from position 659 to 966 matches the sequence of a sequence-tagged site, SHGC-3321, which has been mapped to the short arm of chromosome 20, 58.49 centirays from the top of the linkage group.

View larger version (49K): [in this window] [in a new window]   Fig. 1.   Nucleotide and predicted protein sequence of the HAOX1 cDNA. The 1113-bp open reading frame, which terminates in the near-consensus type-1 peroxisomal targeting signal serine-lysine-isoleucine-COOH (underlined), is predicted to encode a basic protein (pI 8.46) with a mass of 41 kDa. A consensus polyadenylation site is shown in boldface type, and two upstream, in-frame stop codons are double underlined.

The HAOX2 cDNA sequence (Fig. 2) is 1417 bp long and contains a 1056-bp ORF, with 73 bp of 5'-UTR and 288 bp of 3'-UTR. The sequence surrounding the indicated initiation site (CAGAAAATGT) is only a partial match to the Kozak initiation consensus. However, the next methionine codon occurs at nucleotide position 268, well into a region of highly conserved sequence. In addition, in-frame termination codons are present in the 5'-UTR at positions 12 and 64 (Fig. 2, double underline). The deduced protein product of HAOX2 is 351 amino acids long, has a predicted molecular mass of 39 kDa, and terminates in the consensus PTS1 serine-arginine-leucine (Fig. 2, underline). The region from position 458 to 562 corresponds to the chromosome 1 sequence-tagged site, stSG32829.

View larger version (51K): [in this window] [in a new window]   Fig. 2.   Nucleotide and predicted protein sequence of the HAOX2 cDNA. The 1056-bp open reading frame, which terminates in the consensus type-1 peroxisomal targeting signal serine-arginine-leucine-COOH (underlined), is predicted to encode a basic protein (pI 8.21) with a mass of 39 kDa. Two upstream, in-frame stop codon are double underlined.

The HAOX3 cDNA sequence (Fig. 3) is 1821 bp long and contains a 1062-bp ORF, with a 171-bp 5'-UTR and a 588-bp 3'-UTR. Like the HAOX1 and HAOX2 cDNAs, the sequence surrounding the apparent initiation codon (CCAGCAATGT) is only a partial match to the Kozak initiation consensus. Again, subsequent possible initiation methionines are a significant distance into the region of high amino acid identity with GOX, and the indicated initiation site is preceded by an in-frame stop codon (Fig. 3, position 45, double underline). The deduced protein product is 353 amino acids long, has a predicted molecular mass of 39 kDa, and also terminates in the consensus PTS1, serine-arginine-leucine. The chromosomal location of the HAOX3 gene is unknown.

View larger version (53K): [in this window] [in a new window]   Fig. 3.   Nucleotide and predicted protein sequence of the HAOX3 cDNA. The 1062-bp open reading frame, which terminates in the near-consensus type-1 peroxisomal targeting signal serine-lysine-isoleucine-COOH (underlined), is predicted to encode a basic protein (pI 8.53) with a mass of 39 kDa. A consensus polyadenylation site is shown in boldface type, and an upstream, in-frame stop codon is double underlined.

Alignment of the deduced amino acid sequences of HAOX1, HAOX2, HAOX3 and spinach GOX shows that HAOX1 shares the highest identity with the spinach enzyme (54.6%) (Fig. 4). HAOX2 and HAOX3 are more distantly related to plant GOX and have only 46.7 and 44.8% identity, respectively. However, HAOX2 and HAOX3 are closely related and share 70.4% sequence identity to one another, as opposed to only 46.0 and 45.3% identity to HAOX1, respectively.

View larger version (103K): [in this window] [in a new window]   Fig. 4.   Amino acid alignment of spinach glycolate oxidase, HAOX1, HAOX2, and HAOX3. The predicted protein products of HAOX1, HAOX2, and HAOX3 were aligned with spinach glycolate oxidase (SoGOX) using DNASTAR (Madison, WI) and the PAM 250 matrix. Identical residues are boxed.

HAOX1, HAOX2, and HAOX3 Exhibit Different Tissue-specific Patterns of Expression-- Northern blot studies were performed to determine which tissues expressed these three candidate 2-hydroxy acid oxidase genes. A radiolabeled probe specific for each gene was hybridized to poly(A)⁺ RNA from 16 different tissues (Fig. 5). The same filters were used for all three experiments. Long exposures are provided so that low levels of transcript may be seen. For HAOX1 (top panel), a ~2.0-kilobase transcript was present at high levels in liver but at very low levels in kidney. A very small amount of a ~2.4-kilobase transcript can be seen in placenta. A smaller, ~1.2-kilobase transcript was also detected at high levels in pancreas. This transcript may represent an alternatively spliced form of HAOX1 or may reflect the use of one of the multiple near-consensus polyadenylation sites upstream of the consensus polyadenylation motif in the HAOX1 cDNA. A ~2.0-kilobase HAOX2 transcript was detected at high levels in both liver and kidney (middle panel). The finding of similarly sized transcripts of HAOX1 and HAOX2 in kidney raises the possibility of cross-reactivity of the HAOX1 probe with an HAOX2 transcript. This is unlikely, however, since the nucleotide sequence across the HAOX1 probe is only 56% identical to the corresponding HAOX2 sequence and highly stringent conditions were used in blotting. Small amounts of an HAOX2 transcript were also observed in thymus. HAOX3 appeared to be expressed only in pancreas (lower panel). These results are consistent with prior reports of 2-hydroxy acid oxidase activity in human liver and kidney (16, 17, 22, 23) but are novel in identifying the pancreas, placenta, and thymus as possible sites of 2-hydroxy acid oxidase expression.

View larger version (77K): [in this window] [in a new window]   Fig. 5.   Tissue distribution of HAOX1, HAOX2, and HAOX3 expression. A multitissue panel of mRNAs (CLONTECH) was analyzed by Northern blot using probes specific for HAOX1 (top panel), HAOX2 (middle panel), and HAOX3 (bottom panel).

HAOX1, HAOX2, and HAOX3 Display Distinct 2-Hydroxy Acid Oxidase Activities-- The high degree of amino acid sequence identity between HAOX1, HAOX2, HAOX3, and spinach GOX suggested that these proteins might have 2-hydroxy acid oxidase activity. To test this hypothesis, we transfected human skin fibroblasts with plasmids designed to express epitope-tagged versions of HAOX1, HAOX2, HAOX3, and a control plasmid. The three HAOX expression plasmids, pcDNA3-Nmyc/HAOX1, pcDNA3-Nmyc/HAOX2, and pcDNA3-Nmyc/HAOX3, each encode an 18-amino acid c-myc epitope tag immediately upstream of the HAOX open reading frames. Each of the four transfected cell populations were lysed, and the resulting lysates were tested for activity against a panel of 2-hydroxy acids. The lysate of cells expressing Nmyc/HAOX1 showed the highest activity with glycolate as substrate but also exhibited activity toward glyoxylate, 2-hydroxyoctanoate, and 2-hydroxypalmitate (Fig. 6A, upper panel). Nmyc/HAOX1 showed no activity with lactate, 2-hydroxybutyrate, or 3-methyl-2-hydroxybutyrate as substrate. In contrast, lysates of cells transfected with pcDNA3-Nmyc/HAOX2 showed activity only with 2-hydroxyoctanoate or 2-hydroxypalmitate as substrate (Fig. 6A, middle panel), indicating that this enzyme prefers longer chain 2-hydroxy acids. The lysate from cells expressing Nmyc/HAOX3 was only active on 2-hydroxyoctanoate (Fig. 6A, lower panel). Control lysates from human fibroblasts transfected with pcDNA3 lacked activity on any of the substrates tested. Values are the mean ± S.D. of three independent trials. These results indicate that HAOX1, HAOX2, and HAOX3 encode 2-hydroxy acid oxidases with distinct substrate specificities.

View larger version (22K): [in this window] [in a new window]   Fig. 6.   HAOX1, HAOX2, and HAOX3 are 2-hydroxy acid oxidases with distinct substrate specificities. A, lysates of human skin fibroblasts expressing Nmyc/HAOX1 (top panel), Nmyc/HAOX2 (middle panel), and Nmyc HAOX3 (lower panel) were assayed for 2-hydroxy acid oxidase activity versus a panel of 2-hydroxy acids. Oxidase activity was detected by monitoring the coupled reduction of 2,6-dichloroindophenol at 605 nm. Values are the mean ± S.D. of three trials. Lysates of fibroblasts containing pcDNA3 showed no 2-hydroxy acid oxidase activity (data not shown). B, 10 µg of protein from each lysate was separated by polyacrylamide gel electrophoresis and analyzed by immunoblot using anti-myc antibodies. The relative amounts of each enzyme present were detected using MACBAS software, and the results are presented in Table I.

Although each expression plasmid was constructed in a similar manner and an equivalent amount of plasmid was used in each experiment, the actual expression level for each enzyme in the transfected cell populations may differ due to variations in transfection efficiency, protein stability, plasmid purity, etc. Therefore, we examined the levels of Nmyc/HAOX1, Nmyc/HAOX2, and Nmyc/HAOX3 present in the cell lysates by immunoblot (Fig. 6B). Quantitation of the signal obtained indicated that Nmyc/HAOX2 was present at the highest levels, while Nmyc/HAOX3 was present at 69% and Nmyc/HAOX1 at 28% of the Nmyc/HAOX2 level. Adjusting for expression levels, HAOX1 appeared to be the most active on all of the substrates tested (Table I).

We next attempted to examine the intrinsic activities of recombinant HAOX1, HAOX2, and HAOX3 expressed and purified from bacteria. We were unable to express HAOX2 but did obtain some soluble His₆/HAOX1 and His₆/HAOX3 fusion proteins (Fig. 7A). These proteins were tested for activity against the same panel of 2-hydroxy acids tested above. Like the Nmyc/HAOX1-expressing cell lysate, purified His₆/HAOX1 showed highest activity on glycolate and was less active on glyoxylate, 2-hydroxyoctanoate, and 2-hydroxypalmitate (Fig. 7B). No activity was observed on lactate, 2-hydroxybutyrate, or 3-methyl-2-hydroxybutyrate. To examine the kinetics of the His₆/HAOX1 reaction in more detail, we measured the initial rates of glycolate oxidation at various concentrations of glycolate, ranging from 5 to 500 µM (Fig. 7C). His₆/HAOX1 displayed Michaelis-Menten kinetics, and analysis of the data using a double-reciprocal plot and linear regression indicated a specific activity of 0.34 unit/mg and a K_m of 120 µM for glycolate.

View larger version (27K): [in this window] [in a new window]   Fig. 7.   Activities of purified, recombinant HAOX1 and HAOX3. A, Coomassie-stained SDS-polyacrylamide gel electrophoresis gel showing purified His6/HAOX1 (left lane) and purified His6/HAOX3 (right lane). B, the His6/HAOX1 fusion protein was assayed for 2-hydroxy acid oxidase activity versus a panel of 2-hydroxy acids. C, the activity of the His6/HAOX1 fusion protein was determined at various concentrations of glycolate. His6/HAOX1 displayed Michaelis-Menten kinetics and had a specific activity of 0.34 units/mg and a Km of 120 µM for glycolate. D, the His6/HAOX3 fusion protein was assayed for 2-hydroxy acid oxidase activity versus a panel of 2-hydroxy acids. E, the activity of the His6/HAOX3 fusion protein was determined at various concentrations of 2-hydroxyoctanoate. His6/HAOX3 displayed Michaelis-Menten kinetics and had a specific activity of 0.76 units/mg and a Km of 45 µM for 2-hydroxyoctanoate. For each assay, activity was determined by following the reduction of 2,6-dichloroindophenol at 605 nm. All values are the mean ± S.D. of three trials.

Purified His₆/HAOX3 was also tested against a panel of 2-hydroxy acids and showed the highest activity with 2-hydroxyoctanoate as substrate (Fig. 7D). While this was similar to the results obtained with the Nmyc/HAOX3-containing lysate, the relatively large amounts of purified protein also enabled the detection of lesser activity on 2-hydroxybutyrate and 2-hydroxypalmitate. The detection of activity on 2-hydroxybutyrate combined with the high activity on 2-hydroxyoctanoate and low activity on 2-hydroxypalmitate suggests that this enzyme may prefer medium chain 2-hydroxy acids as substrates. The kinetics of His₆/HAOX3 activity were examined by measuring the initial rates of 2-hydroxyoctanoate oxidation at various concentrations of 2-hydroxyoctanoate, ranging from 5 µM to 1 mM (Fig. 7E). Using a double reciprocal plot as before, we determined a specific activity of 0.76 unit/mg and a K_m of 45 µM for the substrate 2-hydroxyoctanoate.

It should be noted that the purified His₆/HAOX1 obtained in this study displayed an A₂₈₀/A₄₅₀ of 32.3, which, when compared with the reported value of 6.1 for the biochemically isolated human enzyme (18), indicates that approximately 80% of the purified protein may have been inactive due to loss of the FMN cofactor during purification. This is supported by our observation that the characteristic yellow color of the recombinant enzyme faded during the washing of our affinity column. The A₂₈₀/A₄₅₀ ratio obtained for His₆/HAOX3 was 9.39, indicating that it had lost only a small amount of FMN and that the activity of His₆/HAOX3 determined here may be similar to that of the native enzyme.

HAOX1, HAOX2, and HAOX3 Are Peroxisomal Enzymes-- The oxidation of long, medium, and short chain 2-hydroxy acids has long been thought to be a peroxisomal process (16). It was therefore not surprising that the 2-hydroxy acid oxidases identified here each terminate in a match or near match to the consensus PTS1 (serine-lysine-isoleucine-COOH for HAOX1, serine-arginine-leucine-COOH for HAOX2, serine-arginine-leucine-COOH for HAOX3) (Figs. 1-4). To determine the subcellular distribution of these proteins, we expressed the N-terminal myc-tagged forms of HAOX1, HAOX2, and HAOX3 in human skin fibroblasts and examined their distribution by indirect immunofluorescence. We observed that Nmyc/HAOX1, Nmyc/HAOX2, and Nmyc/HAOX3 were localized to peroxisomes, as determined by colocalization with the peroxisomal marker PMP70 (Fig. 8). We also examined the subcellular localization of the myc-tagged forms of these enzymes in cells permeabilized with a limiting amount of digitonin. This method permeabilizes the plasma membrane while leaving the peroxisome membrane intact, rendering the interior of peroxisomes inaccessible to antibodies. The myc-tagged HAOX enzymes could not be seen in peroxisomes, but the cytoplasmic tail of PMP70 was readily detected, indicating that all three enzymes reside in the peroxisome lumen (data not shown).

View larger version (123K): [in this window] [in a new window]   Fig. 8.   HAOX1, HAOX2, and HAOX3 are peroxisomal proteins. Human skin fibroblasts expressing Nmyc/HAOX1 (A and B), Nmyc/HAOX2 (B and C), and Nmyc/HAOX3 (D and E) were processed for double-indirect immunofluorescence by fixing cells and permeabilizing with 1% Triton X-100. The distribution of each HAOX protein was examined using anti-myc antibodies (A, C, and E) and antibodies to the peroxisomal marker protein PMP70 (B, D, and F). Scale bar, 25 µm.

Zellweger syndrome is a lethal inherited disorder caused by defects in peroxisome biogenesis (39). Although some Zellweger syndrome patients are defective in both peroxisomal membrane protein import and peroxisomal matrix protein import (40), most, such as those who have mutations in PEX10, are defective only in peroxisomal matrix protein import (33, 41). As an independent test of whether HAOX1, HAOX2, and HAOX3 are imported into the peroxisome matrix, we expressed Nmyc/HAOX1, Nmyc/HAOX2, and Nmyc/HAOX3 in PBD100 fibroblasts derived from a PEX10-deficient Zellweger syndrome patient. The myc-tagged forms of HAOX1, HAOX2, and HAOX3 all accumulated in the cytosol of these cells rather than in the peroxisomes (Fig. 9), indicating that these proteins are normally imported into the peroxisome lumen.

View larger version (134K): [in this window] [in a new window]   Fig. 9.   HAOX1, HAOX2, and HAOX3 are mistargeted in cells deficient in peroxisomal matrix protein import. PEX10-deficient cells from the Zellweger syndrome cell line PBD100 expressing Nmyc/HAOX1 (A and B), Nmyc/HAOX2 (C and D), and Nmyc/HAOX3 (E and F) were processed for double indirect immunofluorescence by fixing cells and permeabilizing with Triton X-100. The distribution of each HAOX protein was examined using anti-myc antibodies (A, C, and E) and antibodies to the peroxisomal marker protein PMP70 (B, D, and F). Scale bar, 25 µm.

We also examined the distribution of HAOX1 by subcellular fractionation. A postnuclear supernatant was prepared from rat liver and separated by Nycodenz density gradient centrifugation. Fractions across the gradient were assayed for peroxisomal, mitochondrial, and cytosolic markers. The distribution of HAOX1 was determined by immunoblot using affinity-purified anti-HAOX1 antibodies (Fig. 10). These antibodies detected a ~42-kDa protein in rat liver peroxisomes. The small amount of HAOX1 detected in cytosolic fractions at the top of the gradient probably represents the lysis of some peroxisomes during homogenization and fractionation and is also observed for the peroxisomal marker enzyme, catalase. These studies were also performed using affinity-purified anti-HAOX3 antibodies. However, as predicted by the tissue-specific expression pattern of HAOX3, no HAOX3 was detected in the liver samples used (data not shown).

View larger version (28K): [in this window] [in a new window]   Fig. 10.   Anti-HAOX1 antibodies recognize a 42-kDa protein in rat liver peroxisomes. A rat liver postnuclear supernatant was fractionated by Nycodenz gradient centrifugation. Equal amounts of fractions were assayed for catalase (dark bars) and succinate dehydrogenase (light bars) activity. The bar graph shows the relative amounts of the peroxisomal and mitochondrial marker enzymes in each fraction. Equal amounts of each fraction were also processed by immunoblot using affinity-purified anti-HAOX3 antibodies. The lower panel shows the distribution of the ~42-kDa rat protein recognized by the anti-HAOX3 antibody.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

-Oxidation of fatty acids has been observed in a wide variety of organisms, but its contributions to mammalian metabolism are unclear. To improve our knowledge of -oxidation in humans, we searched for cDNAs capable of encoding proteins similar to plant glycolate oxidase, a prototypical 2-hydroxy acid oxidase. Our identification of HAOX1, HAOX2, and HAOX3 provides a structural explanation for human 2-hydroxy acid oxidase activities previously known in the liver and kidney and demonstrates that 2-hydroxy acid oxidases are also expressed in pancreas, placenta, and thymus. Furthermore, our inability to identify other related human genes in the available sequence data bases suggests that HAOX1, HAOX2, and HAOX3 may represent the complete repertoire of human 2-hydroxy acid oxidases. It is possible, however, that there may exist additional related or unrelated 2-hydroxy acid oxidases in humans.

Purified, recombinant HAOX1 was active on glycolate as well as longer chain substrates. These findings are consistent with the high sequence similarity between HAOX1 and spinach GOX and also support prior reports that mammalian glycolate oxidase is active on a broad spectrum of substrates (17, 18). The HAOX1 transcript was found at high levels in liver and pancreas and at low levels in kidney. Prior reports have generally focused on the presence of glycolate oxidase activity in mammalian liver, while some have also observed glycolate oxidase activity in the kidney (16, 23, 42). To the best of our knowledge, however, our observation that this enzyme is highly expressed in the pancreas is novel.

HAOX1 is the only one of the the three human HAOXs to be active on glycolate and glyoxylate. Thus, it is likely to play a role in the pathophysiology of primary hyperoxaluria type 1 (43). This disease is caused by defects in alanine-glyoxylate aminotransferase, a peroxisomal enzyme that normally eliminates the bulk of glyoxylate by transamination of alanine and glyoxylate to glycine and pyruvate. Loss of alanine-glyoxylate aminotransferase leads to glyoxylate accumulation, and HAOX1 is clearly able to convert glyoxylate to oxalate. The high oxalate present in primary hyperoxaluria type 1 patients causes formation of calcium oxalate precipitates in the kidney, which can be lethal.

In our search for human 2-hydroxy acid oxidases, we expected to find one homolog of the rat long chain 2-hydroxy acid oxidase (25). Instead, we found two. HAOX2 and HAOX3 are both very similar to the rat enzyme (74.0 and 82.6% identity, respectively). HAOX2 is highly expressed in liver and kidney, and its gene product is most active on long chain 2-hydroxy acids. HAOX3 was only detected in pancreas, and its gene product is most active on a medium chain 2-hydroxy acid. The inability of HAOX2 and HAOX3 to oxidize either glycolate or glyoxylate indicates that these two enzymes are clearly 2-hydroxy fatty acid oxidases.

The expression of HAOX2 in kidney and liver is consistent with prior reports of 2-hydroxy fatty acid oxidase activity in these tissues (16-18, 22, 23). However, these earlier reports documented a 2-hydroxy acid oxidase that acted primarily on 2-hydroxybutyrate and had less activity on 2-hydroxyoctanoate, a substrate preference that we do not observe for HAOX2. One possible explanation for this discrepancy is that there may be yet another 2-hydroxy acid oxidase that prefers this shorter substrate. Alternatively, it could be that differences in expression and assay systems between this and previous reports affect the observed substrate preferences of this enzyme. The nearly exclusive expression of the HAOX3 transcript in the pancreas, a tissue not previously associated with 2-hydroxy acid oxidase activity, suggests that this enzyme is truly novel.

Like plant glycolate oxidase, we found that HAOX1, HAOX2, and HAOX3 all contain type 1 peroxisome targeting signals and localize to peroxisomes. This is in agreement with earlier reports that mammalian 2-hydroxy acid oxidase activities are peroxisomal (44, 45). Cytochemical and immunogold electron microscopic techniques have also localized two 2-hydroxy acid oxidases to mammalian peroxisomes (16). In fact, the marginal plate of kidney peroxisomes, a dense suborganellar structure, consists primarily of crystals of a 2-hydroxy acid oxidase (42, 46).

While it is clear that HAOX1, HAOX2, and HAOX3 are all capable of oxidizing 2-hydroxy fatty acids, it is not generally accepted that these enzymes participate in the -oxidation of fatty acids. The initial steps in fatty acid -oxidation generate 2-hydroxyacyl CoAs, and recent studies suggest that a 2-hydroxyacyl-CoA lyase converts these directly to fatty aldehydes and formyl-CoA (9, 10). However, the data available are not fully explained by this model. For instance, some authors have reported the existence of a 2-keto phytanic acid intermediate in peroxisomes following the addition of 2-hydroxyphytanoyl-CoA (12-15). Furthermore, it is difficult to imagine why peroxisomes would contain multiple enzymes capable of oxidizing 2-hydroxy fatty acids to 2-keto fatty acids but not utilize these enzymes for -oxidation of fatty acids. The work presented here opens the way for future investigations of whether 2-hydroxy acid oxidases contribute to a general mechanism of fatty acid -oxidation.

	ACKNOWLEDGEMENTS

We thank Katherine Sacksteder, Brian Geisbrecht, and Stephanie Mihalik for helpful comments and suggestions during the course of these studies.

	FOOTNOTES

^* This work was supported by National Institutes of Health Grants DK45787 and HD10981 (to S. J. G.).The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom all correspondence should be addressed: Dept. of Biological Chemistry, Johns Hopkins University School of Medicine, Baltimore, MD 21205. Tel.: 410-955-3085 or 410-955-3424; Fax: 410-955-0215; E-mail: sgould@jhmi.edu.

	ABBREVIATIONS

The abbreviations used are: ORF, open reading frame; UTR, untranslated region; DCIP, 2,6-dichloroindophenol; GOX, glycolate oxidase; PTS1, peroxisome targeting sequence type 1; bp, base pair(s).

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

CiteULike

Complore

Connotea

Del.icio.us

Digg

Reddit

Technorati

This article has been cited by other articles:

				V. Foulon, M. Sniekers, E. Huysmans, S. Asselberghs, V. Mahieu, G. P. Mannaerts, P. P. Van Veldhoven, and M. Casteels Breakdown of 2-Hydroxylated Straight Chain Fatty Acids via Peroxisomal 2-Hydroxyphytanoyl-CoA Lyase: A REVISED PATHWAY FOR THE {alpha}-OXIDATION OF STRAIGHT CHAIN FATTY ACIDS J. Biol. Chem., March 18, 2005; 280(11): 9802 - 9812. [Abstract] [Full Text] [PDF]

				K. J. Wallace, R. H. Wallis, S. C. Collins, K. Argoud, P. J. Kaisaki, A. Ktorza, P. Y. Woon, M.-T. Bihoreau, and D. Gauguier Quantitative trait locus dissection in congenic strains of the Goto-Kakizaki rat identifies a region conserved with diabetes loci in human chromosome 1q Physiol Genomics, September 16, 2004; 19(1): 1 - 10. [Abstract] [Full Text] [PDF]

				S. Reumann, C. Ma, S. Lemke, and L. Babujee AraPerox. A Database of Putative Arabidopsis Proteins from Plant Peroxisomes Plant Physiology, September 1, 2004; 136(1): 2587 - 2608. [Abstract] [Full Text] [PDF]

				G. Dodt, D. Warren, E. Becker, P. Rehling, and S. J. Gould Domain Mapping of Human PEX5 Reveals Functional and Structural Similarities to Saccharomyces cerevisiae Pex18p and Pex21p J. Biol. Chem., November 2, 2001; 276(45): 41769 - 41781. [Abstract] [Full Text] [PDF]

				J. M. Jones, J. C. Morrell, and S. J. Gould Multiple Distinct Targeting Signals in Integral Peroxisomal Membrane Proteins J. Cell Biol., June 4, 2001; 153(6): 1141 - 1150. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) An addition or correction has been published Submit a Letter to Editor Alert me when this article is cited Alert me when eLetters are posted Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Request Permissions Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Jones, J. M. Articles by Gould, S. J. Search for Related Content PubMed PubMed Citation Articles by Jones, J. M. Articles by Gould, S. J. Social Bookmarking                      What's this?