Deletion Mutation in Drosophila ma-l Homologous, Putative Molybdopterin Cofactor Sulfurase Gene Is Associated with Bovine Xanthinuria Type II*

Defective xanthine dehydrogenase (XDH) activity in humans results in xanthinuria and xanthine calculus accumulation in kidneys. Bovine xanthinuria was demonstrated in a local herd and characterized as xanthinuria type II, similar to the Drosophila ma-l mutations, which lose activities of molybdoenzymes, XDH, and aldehyde oxidase, although sulfite oxidase activity is preserved. Linkage analysis located the disease locus at the centromeric region of bovine chromosome 24, where a ma-l homologous, putative molybdopterin cofactor sulfurase gene (MCSU) has been physically mapped. We found that a deletion mutation at tyrosine 257 inMCSU is tightly associated with bovine xanthinuria type II.

Xanthine dehydrogenase (XDH) 1 activity is essential for the degradation of purine bases in mammals catalyzing the oxidation reactions of both hypoxanthine to xanthine and xanthine to uric acid. Defective XDH activity elevates xanthine concentration in plasma and urine, whereas hypoxanthine can be salvaged to inosine by hypoxanthine-guanine phosphoribosyltransferase.
XDH requires molybdopterin cofactor (also referred as molybdenum cofactor, MoCo) for its enzymic activity. The cofactor is also essential for the enzymic activities of aldehyde oxidase (AO) and sulfite oxidase (SO) in mammals (1)(2)(3). Both XDH and AO require the sulfide form of molybdopterin cofactor for their enzymic activities, whereas SO does not in the Drosophila ma-l mutant (4 -7). Molybdopterin cofactor extracted from ma-l mutant flies lacked the sulfide moiety (desulfo form) (7). Resulfuration of desulfo MoCo in vitro reactivated xanthine dehydrogenase of the ma-l mutant (7). Therefore, it was hypothesized that the Drosophila ma-l gene encodes a putative enzyme that catalyzes sulfuration of desulfo MoCo, the last step of MoCo synthesis.
XDH deficiency in humans results in xanthinuria and the accumulation of xanthine calculi in renal tubules that leads to renal dysfunction (1,2). Loss-of-function mutations in the XDH gene and genes responsible for MoCo biosynthesis can result in xanthinuria. In fact, hereditary xanthinuria in humans is classified into three categories (1,2). Xanthinuria type I is caused by a loss-of-function mutation in the XDH gene (8). Xanthinuria type II lacks both XDH and AO activities (1,9); although the causative gene of xanthinuria type II is unknown, it has been suggested to be equivalent to the Drosophila ma-l mutation (4 -6). The third category is found in MoCo deficiency produced by a loss-of-function mutation of the MoCo synthetase gene catalyzing the first steps in MoCo synthesis (10). In humans, this condition is lethal in the perinatal period due to the absence of SO activity, which catalyzes sulfite oxidation (2).
Recently, xanthinuria was demonstrated in a local cattle herd (11). Here, we show that this form of bovine xanthinuria is an autosomal recessive trait by which affected cattle lose XDH and AO activities, although SO activity is preserved, suggesting that this disorder is xanthinuria type II (1,9). We report that a deletion mutation at the Drosophila ma-l orthologue is strongly associated with bovine xanthinuria type II.

EXPERIMENTAL PROCEDURES
Genotyping and Linkage Analysis-Total genomic DNA was prepared from peripheral blood leukocytes using standard protocols with an Easy-DNA TM Kit (Invitrogen). The PCR conditions for microsatellite markers were optimized (12), and additional reaction conditions were set as recommended by the manufacturer. Microsatellite polymorphisms were analyzed using PCR amplification and gel electrophoresis by an ABI 377 DNA sequencer as described (13). Genotype data were captured by means of GENESCAN and Genotyper software (Perkin-Elmer Applied Biosystems), and linkage analysis was performed with the GENEHUNTER package (14).
Assay of Enzymic Activities-Crude liver extracts for XDH and AO activities were prepared as described (15). Briefly, the liver homogenates were treated at 55°C for 11 min followed by 50% ammonium sulfate precipitation. The dialyzed crude extracts were then assayed for XDH and AO activities. XDH activity was estimated by the oxidation of hypoxanthine to uric acid as described (16). AO activity was measured by the oxidation of N 1 -methylnicotinamide to 2-and 4-pyridones in the presence of the XDH inhibitor allopurinol as described (15). The preparation of crude liver extracts for SO activity and the assay were performed as described (3).
FISH and Microsatellite Isolation-A mouse EST, AA450702, corresponding to Pro-499 3 Val-622 of Drosophila Ma-l, was chosen to design PCR primers (545F, 5Ј-GACCGGAGCTGGATGGTTGTG-3Ј; 596R, 5Ј-CCTCAAGAGGCACCTGGATAGG-3Ј) to amplify a ma-l homologous fragment by RT-PCR using bovine liver mRNA. A 150-bp PCR product corresponding to Asp-545 3 Asp-596 of Drosophila Ma-l was confirmed by direct sequencing. We subsequently designed PCR primers (552F, 5Ј-ATCACAACGGCATTTGCCTGA-3Ј; 588R, 5Ј-CTCC-ATCCCTTGGGCTTTGATGAC-3Ј) from the bovine partial cDNA sequence to amplify a 110-bp fragment corresponding to Asp-552 3 Ile-588 of Drosophila Ma-l using bovine genomic DNA. Direct sequencing * This study was supported by grants from the Japan Racing and Livestock Promotion Foundation. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
The nucleotide sequence(s) reported in this paper has been submitted to the GenBank TM /EBI Data Bank with accession number(s) AB036422 (bovine MCSU).

RESULTS AND DISCUSSION
Bovine xanthinuria in a local herd of Japanese Black cattle was characterized by elevated xanthine secretion in the urine associated with lethal growth retardation at approximately 6 months of age (11). Affected cattle had expanded renal tubules containing xanthine calculi ranging from 1-3 mm in diameter (11). We confirmed that more than 300 xanthinuria-affected cattle have been recorded over the last 20 years and that all parents were descendants of a putative founder sire. Affected male, female, and unknown offspring numbered 177, 148, and 9, respectively. Pedigree analysis in this herd indicates that bovine xanthinuria is inherited as an autosomal recessive trait.
Three types of xanthinuria can be classified based on the activities of the MoCo-requiring enzymes XDH, AO, and SO (1,2). We found that XDH and AO activities in liver extracts were decreased in affected cattle, whereas SO activity was preserved (Fig. 1A). Therefore, we classified this type of bovine xanthinuria as xanthinuria type II (1,9). Twenty-one xanthinuria type II-affected offspring (11 males and 10 females) and their parents, 21 dams and two sires, were collected (Fig. 1B)  covering all bovine autosomes at approximately 15-centimorgan intervals, and showing heterozygosity for at least one of the two sires, was used in an initial genome scan in the family segregated for the disorder. The putative xanthinuria type II locus was mapped at the centromeric region of bovine chromosome (BTA) 24 (Z ϭ 36.6, p Ͻ 1.9 ϫ 10 Ϫ6 ) (Fig. 1C) (14).
Because the Drosophila ma-l orthologue, putative MoCo sulfurase gene has been suggested as causative for xanthinuria type II (4 -6), we investigated whether bovine Drosophila ma-l orthologue is located at the same region of xanthinuria type II locus on BTA24. The deduced amino acid sequence encoded by Drosophila ma-l (GenBank TM accession number AF162681; kindly provided by Victoria Finnerty, Emory University, Atlanta, GA) was subjected to a TBLASTN-search on the dbEST DNA sequence data base of the GenBank TM to collect ESTs derived from mammalian orthologues. One hundred ESTs showing more than 40% identity to the ma-l amino acid sequence were obtained. A mouse EST (AA450702) was chosen to design PCR primers (552F and 588R) to amplify a ma-l homologous fragment using bovine genomic DNA. The direct sequencing of a 110-bp PCR product confirmed the Asp-552 3 Ile-588 of Drosophila Ma-l. The bovine Drosophila ma-l orthologue was physically mapped by FISH using a bovine YAC clone 13H10 identified by PCR-based screening with the 552F and 588R primers. The YAC clone harboring the ma-l orthologue was located at BTA24q13.1-13.3 by FISH (Fig. 1D), where the xanthinuria type II locus genetically mapped (Fig.  1C). Microsatellite DIK124 isolated from a cosmid clone harboring the ma-l orthologue 110-bp DNA fragment was subjected to linkage analysis using the United States Department of Agriculture reference panel (12) and mapped closely to loci CSSM031 ( ϭ 0.00) and ILSTS065 ( ϭ 0.02) on BTA24 at the peak between 25 and 38 centimorgan in Fig. 1C. These results strongly suggest that a bovine xanthinuria type II causative gene is the ma-l orthologue.
To isolate a putative homologous Drosophila ma-l MoCo sulfurase gene, we extended the core fragment of the ma-l orthologue 110 bp using internal primers and bovine liverderived cDNA in 5Ј-and 3Ј-RACE, respectively. We detected an open reading frame (ORF) with an in-frame upstream stop codon. The ORF has 2547 nucleotides and encodes a protein of 849 amino acids. Because the sequence revealed approximately 40% amino acid similarity to Drosophila ma-l (GenBank TM accession number AF162681) and Aspergillus hxB (Gen-Bank TM accession number AF128114; Ref. 19) (Fig. 2A), we designated this gene as MoCo sulfurase (MCSU). Analysis of the genomic organization indicated that MCSU consists of at least 15 exons spanning 25 kilobases, with each exon flanked by canonical splice donor and acceptor sequences (Table I). We estimated MCSU expression by semiquantitative RT-PCR and demonstrated ubiquitous expression among diverse normal tissues (Fig. 2B). Normal lung tissue showed the highest expression. Affected cattle also expressed MCSU in the liver at levels equivalent to normal, suggesting that the level of MCSU expression was not impaired. MCSU expression was not detected in normal tissues by Northern blot, probably because of a low level of expression.
To detect a mutation associated with xanthinuria type II in MCSU, we compared ORF sequences from normal individuals with those from affected cattle. A two-base substitution at 478 and 479 resulting in the conversion of Pro-160 3 Gly-160 was detected among normal cattle. This substitution is not related to the xanthinuria type II. A three-base deletion at 769 -771 encoding Tyr-257 was identified in affected cattle (Fig. 3A). PCR primers (714F and 800R) were designed to amplify the 87-bp fragment containing Tyr-257 using genomic DNA as template. Twenty-one affected offspring as shown in Fig. 1B were demonstrated to harbor the homozygous Tyr-257 deleted 84-bp allele, whereas heterozygous 87-and 84-bp alleles were detected in their parents, which were expected carriers (Fig.  3B). In contrast, more than 100 normal cattle from various cattle breeds, such as Japanese Black (unrelated to the xanthinuria type II founder sire), Limousine, Hereford, Holstein, and Angus, contained only the 87-bp normal alleles. This mutation detection method by simple PCR is applicable for practical diagnosis to detect heterozygous carrier cattle.
To further address the issue of whether Tyr-257 is essential for MCSU function, MCSU sequences around Tyr-257 were compared between cattle, human, mouse, pig, fly, and fungus. Fig. 3C shows that Tyr-257 was widely conserved and was situated in a highly conserved region in mammals. Even in the fly and fungus, two phenylalanine aromatic amino acid residues are preceded by basic amino acids and followed by a four-amino acid stretch, GGGT, conserved in mammals. These data strongly suggest that deletion of Tyr-257 results in the loss of MCSU function, leading to the occurrence of type II xanthinuria.
Although we do not as yet have direct evidence demonstrating the enzymic nature of MCSU, loss-of-function mutations of ma-l and hxB have been reported to be MoCo sulfurationdeficient in Drosophila and Aspergillus, respectively (7,19). MoCo sulfuration activities of these gene products have not yet been confirmed biochemically. Wahl et al. (7) reported that application of the resulfuration procedure to crude extracts of Drosophila ma-l mutants reactivates XDH and AO and proposed that the mutants are defective in the sulfuration of desulfo MoCo. Interestingly, Ma-l protein has a weak homology to NifS protein, which is involved in nitrogen fixation of bacteria and has a transulfurase activity (6). MCSU protein does not have significant homology to known proteins except for tRNA splicing protein SPL1 of Candida maltosa, sharing a 25.5% amino acid identity in the region from position 17 to 261 of MCSU. Twenty-one amino acid residues from 223 to 243 of MCSU (ADFVPISFYKIFGFPTGLGAL) have a similarity to the pyridoxal phosphate binding motif ((LIVFYCHT)-(DGH)-(LIVMFYAC)-(LIVMFYA)-x 2 -(GSTAC)-(GSTA)-(HQR)-K-x 4, 6-G-x-(GSAT)-x-(LIVMFYSAC); Prosite motif ID, PS00595) such as that of C. maltosa SPL1 (IDLLSISSHKIYGPKGIGAC). Corresponding regions of Ma-l and HxB are highly conserved (PDYVCLSFYKIFGYPTGVGAL and PDFTVLSFYKIFGFP-DLGAL, respectively). Because most enzymes that have transulfurase activity, such as cystathionine ␥-lyase, require pyridoxal phosphate as a cofactor, we suggest that MCSU protein may bind pyridoxal phosphate and catalyze the transulfuration reaction of MoCo. Further investigation will be needed to confirm MCSU transulfurylase activity.