INTRODUCTION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Nitrogen is frequently the limiting nutrient for growth of many crop plants (Schubert, 1986), and tropical legumes such as soybean (Glycine max) obtain the nitrogen that ultimately supports protein synthesis via a particularly sophisticated system. The plants establish a symbiotic relationship with bacteria from the genus Bradyrhizobium, which contain the enzyme nitrogenase. Nitrogenase catalyzes the conversion of atmospheric dinitrogen to ammonia, which can be used to meet the plant's metabolic nitrogen requirements, but first it must be converted into organic forms that can be transported throughout the plant and further metabolized. Isotope-labeling studies demonstrate that ammonia produced from dinitrogen reduction is rapidly converted into allantoin and allantoate, the so-called ureides (Ohyama and Kumazawa, 1978); up to 95% of the nitrogen in the xylem sap in nodulated soybeans is in the form of ureides (Schubert, 1981).

The ureides are derived from the oxidation of purines, and allantoate arises from the hydrolysis of S-allantoin, but details of the origin of allantoin remain unclear. It has been widely reported that allantoin is the product of the urate oxidase (UO) reaction, but recent work has established that UO catalyzes the conversion of urate to 5-hydroxyisourate (Modric et al., 1992; Kahn et al., 1997). 5-Hydroxyisourate (HIU) is relatively unstable, and in vitro decomposes to yield allantoin, but this is unlikely to be the physiologically relevant mechanism of allantoin formation because the half-life of HIU is approximately 20 min and its decomposition yields racemic allantoin. The nonenzymatic pathway for the conversion of HIU to allantoin has been defined, and the first step is hydrolysis of the N1-C6 bond (Kahn et al., 1997).

In an effort to clarify the mechanism of allantoin biogenesis, we searched for an enzyme in soybean root nodules that would use HIU as a substrate. We earlier reported the identification and purification of a novel enzyme that catalyzes the hydrolysis of the N1-C6 bond of HIU, which we designated hydroxyisourate hydrolase (HIUHase; Sarma et al., 1999). We describe here the cloning of the gene that encodes HIUHase from soybean and the subcellular localization of the protein. The gene sequence confirms that the encoded protein is a novel enzyme and provides clues to its mechanism of action. The localization of the protein in the nodule and the temporal pattern of expression of its gene are consistent with HIUHase playing a role in the ureide pathway.

	RESULTS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Isolation, Cloning, and Sequencing of HIUHase cDNA

PCR amplification using degenerate primers designed from the sequences of tryptic peptides yielded a 600-bp fragment of the HIUHase gene. Using RACE procedures, we were able to isolate a cDNA fragment that contained an open reading frame 1,680 bp in length. The 600-bp genomic DNA fragment that was isolated first was identical to a fragment amplified from cDNA with the same primers, indicating the absence of introns in that portion of the gene. The open reading frame codes for a protein 560 amino acids long with a calculated molecular mass of 63,727 D (Fig. 1). The cDNA sequence has been deposited in GenBank and assigned accession number AF486839. The mass of HIUHase isolated from soybean root nodules was estimated at 68 kD, based on SDS-PAGE and gel filtration chromatography; the smaller mass calculated from the deduced amino acid sequence may suggest that the protein is glycosylated, or it may be a consequence of the imprecision of the experimental techniques used to estimate the molecular mass.

View larger version (61K): [in this window] [in a new window]   Figure 1.   Alignment of the deduced amino acid sequence of HIUHase with maize (Zea mays) -glucosidase isozyme 1. Identical residues are indicated with an asterisk and conservative substitutions are indicated with a period. The signal peptide for HIUHase is italicized; the active site Glu are in boldface type, and the SKL motifs in HIUHase are underlined.

Examination of the deduced amino acid sequence revealed the presence of a putative signal peptide at the N terminus and identified a likely cleavage site between Gly31 and Ala32 (Nielsen et al., 1997). N-terminal amino acid sequencing of HIUHase isolated from soybean root nodules yielded the sequence starting from Ala32. The sequences SKI (residues 523-525) and IPLKL (residues 553-557) are recognizable as SKL motifs that define the type I peroxisomal targeting sequence in plants (Johnson and Olsen, 2001).

Heterologous Expression and Purification of HIUHase

The HIUHase coding sequence was cloned into the expression vector pET-14b, and the overexpressed, recombinant (His)₆-tagged enzyme was demonstrated to catalyze the hydrolysis of 5-hydroxy-isourate. The recombinant protein was partially soluble; after cell lysis, a large fraction of the HIUHase was present in the pelleted cell debris. However, sufficient HIUHase was soluble that it could be purified and characterized. The (His)₆-tagged protein was readily purified by chromatography on a metal-affinity column charged with nickel. Coomassie Blue staining of samples that had been electrophoresed on SDS-polyacrylamide gels indicated that the purified protein was >95% homogeneous (Fig. 2).

View larger version (60K): [in this window] [in a new window]   Figure 2.   Expression and purification of recombinant soybean root nodule HIUHase. Crude protein extracts were prepared from Escherichia coli cells harvested before (lane 1) and 3 h after (lane 2) induction with isopropyl -D-thiogalactoside. The soluble protein was purified by metal affinity column (lane 3). The purified protein was also analyzed by immunoblotting with HIUHase antiserum (lane 4). The positions of Mr markers (Bio-Rad, Hercules, CA) on the gel are indicated.

Biochemical Properties and Sequence Homology of HIUHase

Recombinant HIUHase lacking the signal peptide catalyzed the hydrolysis of HIU with a turnover number of 24 ± 1.5 s¹; the K_m for HIU was 20 ± 3 µM. These values agree well with those obtained with HIUHase purified from soybean root nodules (Sarma et al., 1999).

The primary sequence of HIUHase is 57% identical over 497 amino acids with an open reading frame in Arabidopsis that is annotated as a -glucosidase-like protein (GenBank accession no. AAF02882). The sequence of HIUHase is 33% identical to maize -gluco-sidase isozyme 1 (GenBank accession no. U25157), for which a crystal structure is available (PDB accession nos. 1ELE and 1ELF; Czjzek et al., 2001). A comparison of the sequences of maize -glucosidase isozyme 1 and HIUHase is shown in Figure 1. Sequence analysis of family 1 -glycosidases has revealed that two Glu residues are highly conserved and appear within the consensus sequences T(F/L/M)NEP and (I/V)TENG (Czjzek et al., 2001). HIUHase contains Glu residues at positions 199 and 408; they lie within the sequences TVNEP and IHENG, which closely resemble the consensus motifs in family 1 -glycosidases. The crystal structure of maize -glucosidase isozyme 1 shows that residues Gln38, His142, Trp457, Glu464, and Trp465 interact with the Glc moiety of the substrate; Trp378 is believed to provide the binding determinant for the aromatic aglycon substituent of the substrate. These active site residues are highly conserved in retaining glycosidases; however, with the exception of Gln54 in HIUHase, which aligns with Gln38 in the maize enzyme, none of these residues are conserved in HIUHase.

The E199A and E408A proteins were prepared by site-directed mutagenesis and purified. The circular dichroism spectra of the mutant proteins were identical to that of the wild-type enzyme, suggesting that the mutants were properly folded. Neither mutant exhibited detectable hydrolytic activity toward 5-hydroxyisourate. Because HIU decomposes relatively quickly even in the absence of HIUHase, we can only conclude that the mutant proteins are at least an order of magnitude less active than the wild-type enzyme.

Expression of HIUHase in Soybean

Tissue-specific and temporal expression of the HIUHase gene was investigated by reverse transcriptase (RT)-PCR analysis. Transcripts for HIUHase could be detected in leaves, stems, roots, and nodules, but were clearly more abundant in nodules than in other tissues (Fig. 3). Furthermore, transcript levels in leaves, stems, and roots did not vary over time, whereas the HIUHase transcript increased with increasing plant age up to 21 d, then decreased. The tissue-specific and temporal expression of UO matched that of HIUHase.

View larger version (27K): [in this window] [in a new window]   Figure 3.   Urate oxidase and HIUHase gene transcripts in different soybean tissues, and at different ages postinfection. L, Leaf; S, stem; R, root; N, nodule. Ubiquitin (Ub) was used as an internal control for the RT-PCR analysis.

Polyclonal antibodies raised against HIUHase purified from soybean root nodules were used to detect HIUHase protein in soybean nodules, roots, and leaves (Fig. 4). HIUHase was most abundant in nodules, although it was detectable in roots and leaves as well. HIUHase levels in the leaves did not vary over time, but in the nodules, the HIUHase level was maximal at 21 d and then declined.

View larger version (42K): [in this window] [in a new window]   Figure 4.   Immunodetection of HIUHase in soybean tissue. L, Leaf; R, root; N, nodule. A, Expression of HIUHase in 21-d-old plants. Each lane contains 30 µg of protein. B, Temporal expression pattern of HIUHase in nodules and leaves. The leaf tissue blot was overexposed relative to the nodule tissue blot to allow detection of HIUHase. Each lane contains 15 µg of protein.

Subcellular Localization of HIUHase

To determine the intracellular distribution of HIUHase in soybean root nodules, immunogold electron microscopy was conducted with thin sections of root nodules collected 21 d postinfection. Substantial labeling was found only within the enlarged peroxisomes of the uninfected cells residing in the infected region of the soybean root nodule (Fig. 5A). The gold particles were uniformly distributed over the matrix that fills the peroxisome (Fig. 5B). Sparse labeling in the cytoplasm was not significant, and there was no evidence for ordered localization within the cytoplasm, which would be expected for proteins associated with the endoplasmic reticulum (Shorrosh et al., 1993). No labeling above background was observed in infected cells. A similar pattern of labeling was obtained with sections probed with anti-UO antibody, i.e. labeling was observed only in peroxisomes of uninfected cells (data not shown). No specific immunolabeling was observed in control sections treated with preimmune serum (Fig. 5C).

View larger version (95K): [in this window] [in a new window]   Figure 5.   Immunogold localization of soybean root nodule HIUHase. Tissue sections were probed with antibody raised against purified HIUHase (A and B) or pre-immune serum (C), followed by incubation with secondary antibodies conjugated to 10-nm gold particles. Bars = 500 nm. A, Portion of the uninfected cells (UC) in the infected region of the nodule. Gold labeling is evident only in the enlarged peroxisomes (px) present in the uninfected cells. Note the absence of labeling in the cytoplasm of the uninfected cells and the adjacent infected cells (IC). B, Labeling in a peroxisome in an uninfected cell; note the associated endoplasmic reticulum (arrow). C, Nodule tissue section containing uninfected cells in an infected region, probed with pre-immune serum.

	DISCUSSION

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

In tropical legumes that are actively fixing nitrogen, the ureide pathway supplies the nitrogen that is required for biosynthetic purposes (Schubert, 1986). The nitrogen is carried in the form of allantoin and allantoate, the ureides. Despite the importance of this metabolic pathway, several of its constituent enzymes remain unidentified or poorly characterized. We recently purified HIUHase from soybean root nodules and demonstrated that it catalyzes the hydrolysis of HIU, the product of the UO reaction, to generate 2-oxo-4-hydroxy-4-carboxy-5-ureidoimida-zoline (OHCU). OHCU is an intermediate in the nonenzymatic formation of allantoin from HIU; decarboxylation and tautomerization of OHCU yields allantoin (Scheme 1).

View larger version (14K): [in this window] [in a new window]

Sequence analysis revealed that HIUHase has significant homology to -glucosidases, members of the family of retaining glycosidases. Enzymes in the -glucosidase family are characterized by the presence of two Glu residues in the active site, and these are present in HIUHase as Glu199 and Glu408. Replacement of Glu199 or Glu 408 with Ala yields inactive protein, which strongly suggests that both residues perform critical functions in the catalytic reaction. In addition to the Glu at the active site, several residues that directly interact with the substrate are highly conserved in -glycosidases. However, only one of these residues is conserved in HIUHase. This result is perhaps unsurprising because HIU bears little resemblance to the natural and artificial substrates for -glucosidases, but it does serve to suggest that we have not simply isolated a -glucosidase that can use HIU as an alternative substrate.

The homology between HIUHase and -gluco-sidases is very informative in terms of potential catalytic mechanisms. Retaining glycosidases catalyze a two-step reaction (Davies et al., 1998). In the first part of the catalytic cycle, one Glu residue acts as a nucleophile toward the anomeric carbon of the substrate, and the second Glu donates its proton to the leaving aglycon. To regenerate free enzyme, the second Glu abstracts a proton from water to facilitate the hydrolysis of the glycosyl-enzyme intermediate. One can envision an analogous mechanism for the HIUHase reaction in which one Glu attacks C6 of HIU to form a tetrahedral intermediate. Collapse of the tetrahedral intermediate and expulsion of N1 would be facilitated by general acid catalysis provided by the second Glu residue, and would yield enzyme covalently attached to the substrate in an anhydride linkage. The free enzyme would be regenerated in the second part of the catalytic cycle using the second Glu, now deprotonated, as a general base to abstract a proton from water for nucleophilic attack on the anhydride linkage. Collapse of the resulting tetrahedral intermediate would yield the observed product and return each Glu residue to its original protonation state.

An alternative mechanistic possibility is suggested by the fact that the type of reaction catalyzed by HIUHase, formally the hydrolysis of an amide bond, is also accomplished by many proteases using paired acidic residues at the active site. Human immunodeficiency virus protease serves as an outstanding example, and uses two Asp residues in the hydrolysis of peptide and protein substrates. Elegant studies of human immunodeficiency virus protease have suggested that no covalent intermediate forms during the catalytic cycle; rather, the Asp residues serve as general acid and general base, and mediate a complex series of proton transfers that facilitate hydrolysis of the substrate by a water molecule that is tightly held at the active site (Hyland et al., 1991). This mechanism is also potentially applicable to the HIUHase reaction; experimentation will be required to determine whether HIUHase adheres to a mechanism similar to that used by enzymes to which it is related by sequence, or whether its mechanism is similar to that used by enzymes that catalyze chemically similar reactions.

The deduced amino acid sequence for HIUHase contains 31 amino acids at the N terminus that are not present in the protein isolated from soybean root nodules. The function of the N-terminal region is not known; however, the N terminus was recognized as a signal peptide by sequence analysis, and the cleavage site that was identified in that analysis matched the experimentally determined N terminus of the protein isolated from the soybean root nodule. Two SKL motifs are present near the C terminus of HIUHase, suggesting that it is targeted to the peroxisome. UO is a peroxisomal protein (Van den Bosch and Newcomb, 1986), and because the product of its reaction is the substrate for the HIUHase reaction, and is unstable as well, it seems logical that HIUHase would also be localized in the peroxisome.

This expectation was borne out by electron microscopy studies, which showed that HIUHase is localized in the peroxisomes of the uninfected cells in soybean root nodules. In mature nodules, remarkable changes in the uninfected cells occur in response to nitrogen fixation; the peroxisomes become enlarged and the endoplasmic reticulum associated with the peroxisomes becomes abundant (Newcomb and Tandon, 1981). There is a distinct division of labor between the infected and uninfected cells, and the uninfected cells in the infected regions of the root nodules are the site of ureide production (Newcomb and Tandon, 1981; Webb and Newcomb, 1987; Smith and Atkins, 2002). The UO reaction and allantoin biogenesis take place in the enlarged peroxisomes of uninfected cells (Hanks et al., 1983; Van den Bosch and Newcomb, 1986; Webb and Newcomb, 1986; Johnson and Olsen, 2001). Allantoinase activity is confined to microsome fractions, including endoplasmic reticulum associated with peroxisomes (Hanks et al., 1981). In this regard, it is particularly interesting to note the close proximity of the endoplasmic reticulum and peroxisomes containing HIUHase (Fig. 5B), although our data do not allow us to determine whether a functional association exists or not.

The HIUHase transcripts were present in all the plant tissues we examined. However, there was an enhancement of HIUHase gene expression in nodules when compared with leaves, stems, and roots, where it was constitutively expressed at low levels (Fig. 3). The pattern of HIUHase gene expression detected by RT-PCR matched the expression of the 60-kD immunoreactive HIUHase polypeptide (Fig. 4). The differential expression of the HIUHase gene in nodules suggests that there may be different homologs of HIUHase having roles other than ureide biogenesis. The existence of several structural or functional homologs of nodulin genes including leghemoglobin (Taylor et al., 1994; Jacobsen-Lyon et al., 1995), ENOD40 (van de Sande et al., 1996), nodulin 26 (Miao and Verma, 1993), and UO (Takane et al., 1997, 2000) supports this idea. It has been proposed that the symbiotic genes have diverged from the nonsymbiotic genes through gene duplication (Andersson et al., 1996; Takane et al., 1997), and that these nonsymbiotic genes may have roles in functions common in the plant kingdom like reutilization of nitrogen arising from nucleic acid degradation (Vincentini and Matile, 1993).

The HIUHase transcript was not only abundant in the root nodules, but its level was enhanced during nodule development. The temporal regulation of the HIUHase gene in the nodule that we observed is coincident with the nodule cell expansion phase characterized by an increase in acetylene reduction activity, indicative of increased nitrogenase activity (Anthon and Emerich, 1990) and changes in the activities of several carbon metabolic enzymes that are required to provide energy to the plant to support nitrogen fixation (Karr et al., 1984). The HIUHase expression pattern matched that of the UO gene; it has been established that events during nodule development are important for the regulation of UO gene expression (Padilla et al., 1991). The coordinate induction of the HIUHase and UO genes suggests that they may share at least some regulatory pathways, and strongly supports the hypothesis that HIUHase plays a role in the ureide pathway.

It is not known how OHCU produced by the action of HIUHase is converted to allantoin under physiological conditions. Like HIU, OHCU is relatively unstable, and for in vitro studies, must be generated in situ by the enzymatic oxidation of urate. However, the availability of recombinant HIUHase should allow for more facile preparation of OHCU, which will facilitate the search for an enzyme (or enzymes) that converts OHCU to allantoin.

HIUHase is localized in the peroxisome of the soybean root nodule, and the temporal pattern of its expression matches that of UO and is concomitant with nitrogen fixation activity. We suggest that HIUHase is a constituent of the ureide pathway and serves to catalyze the hydrolysis of HIU to form OHCU. The hydrolytic reaction appears to be mediated by Glu residues at the active site, although it is not known yet whether they function in a manner analogous to the active site glutamates in -glycosidases.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION RESULTS DISCUSSION MATERIALS AND METHODS LITERATURE CITED

Plant Materials and Growth Conditions

Soybean plants (Glycine max cv Williams) inoculated with Bradyrhizobium japonicum strain 2143 were grown as described by Karr et al. (1984). Plants were grown in the greenhouse with a day/night temperature of 28°C (14 h)/25°C (10 h). The light intensity at pot level was maintained at 250 mE m² s¹. Plants were harvested at different ages for collection of the organs. The plants were carefully freed from the potting medium, and the nodules were removed from the roots.

PCR Amplification of Soybean Genomic DNA

General cloning techniques were carried out as described by Sambrook et al. (1989). PCR products were cloned into the pCR 2.1 plasmid (Invitrogen, Carlsbad, CA) and were sequenced using an automated DNA sequencer. HIUHase was purified from fresh or frozen soybean root nodules as previously described (Sarma et al., 1999). Tryptic peptides were generated and sequenced at the University of California-Davis Molecular Structure Facility. Based on three unique tryptic peptide sequences and the sequence of the N terminus, five inosine-containing degenerate oligonucleotides of about 30 bases long were synthesized with a degeneracy of no more than 256-fold and were employed in nested PCR reactions. Primers were designed to have similar annealing temperatures. PCR amplifications were conducted using genomic DNA isolated from soybean according to the method of Dellaporta (Dellaporta et al., 1983). The primers used for the first round of amplification were 5'-gCCgATAAYTAYTCIAgRgATgATTTYCC-3' (primer 1, corresponding to the sense orientation of the 5' end of the coding sequence, deduced from the sequence of the N terminus, ADNYSRDDFPLDFVGSXTSA; the published N-terminal amino acid sequence contains an error: the sequence FVFV for residues 13-16 should be FV), 5'-TATAAT-TTYgATYTICCWCARgTTYTIgA-3' (primer 2F, corresponding to the sense orientation of the coding sequence deduced from the internal peptide LYNFDLPQVLE), 5'-TCIARAACYTgWGGIARATCRAAATTRRTA-3' (primer 2R, the reverse complement of primer 2F), 5'-TTYACWTAYTAYgCWg-AAgTTgAA-3' (primer 3F, corresponding to the sense orientation of the coding sequence deduced from the peptide FTYYAEV), and 5' TTCAA-CTTCWGCRTARTAWGTRAA-3' (primer 3R, the reverse complement of primer 3F). Primers 2F, 2R, 3F, and 3R were used in different combinations to determine their positions in the structural gene. The PCR conditions were: 94°C for 1 min, 45°C for 1 min, and 72°C for 2 min, for 25 cycles, and a final extension at 72° C for 10 min.

cDNA Cloning of HIUHase

Total RNA from soybean root nodule was isolated as described by Ostrem et al. (1987). Nondegenerate primers were synthesized based on the sequences of the amplicons obtained above and were used to amplify the 5' and 3' ends of the cDNA by RACE protocols (Frohman et al., 1988), using a kit from Invitrogen. For the 5'-RACE reaction, gene-specific primers were used for the RT-catalyzed synthesis of cDNA, and terminal transferase was used to add oligo(dC) to the 5' end of the cDNA. PCR reactions were performed with the tailed cDNA using a gene-specific primer and a primer that hybridized to the oligo(dC) tail. An annealing temperature of 50°C was used; other conditions were according to the manufacturer's instructions. The 5' end of the HIUHase gene was reached after three rounds of 5'-RACE.

We were unable to obtain satisfactory cDNA for the 3'-RACE reactions using total RNA, so cDNA was synthesized from mRNA obtained as a generous gift from Virginia Coryell (Northern Arizona University, Flagstaff) using an oligo(dT) primer. PCR reactions were performed with the cDNA using gene-specific primers and oligo(dA), and a 5:1 mixture of Pfu DNA polymerase (Stratagene, Las Jolla, CA) and Taq polymerase (Amersham Biosciences, Piscataway, NJ). Four products were obtained from the 3'-RACE reactions, which were cloned and sequenced. A database search identified an expressed sequence tag that contained the sequence obtained in the 5'-RACE reaction and one of the four products obtained in the 3'-RACE experiment (GenBank accession no. BE806376). Based on this information, primers specific to the 5' and 3' ends of the HIUHase coding sequence were designed; XhoI sites were incorporated into each primer as well. These primers were used with the cDNA as the template to amplify the entire coding sequence of the HIUHase gene.

Expression and Purification of Recombinant HIUHase

The amplicon containing the coding sequence of HIUHase was cloned into the pCR 2.1 vector and the product was digested with XhoI. The XhoI fragment was transferred into the expression vector pET-14b (Novagen, Madison, WI) to make a construct coding for HIUHase with a (His)₆-tag at the N terminus. Inserts with the proper orientation were identified by PCR, and were sequenced to ensure that no mutations had been introduced. A construct lacking the sequence coding for the signal peptide (vide infra) was prepared by PCR amplification in an analogous manner. Overexpression was achieved by transforming BL21(DE3) pLysS cells with the expression plasmid. Cells were grown in Luria-Bertani broth at 30°C, and expression was induced by addition of isopropyl -D-thiogalactoside to a final concentration of 0.4 mM. Cells were harvested by centrifugation 4 to 5 h after induction and were stored at 80°C.

To purify recombinant HIUHase, the Escherichia coli cell paste was resuspended in 4 mL of buffer g¹ cell paste; typical preparations were carried out with 15 g of cell paste. The lysis buffer was 50 mM Tris-HCl, pH 7.5, containing 2 mM MgSO₄, 2 mM CaCl₂, 0.5 mM phenylmethylsulfonyl fluoride, and 0.5 mM N-p-tosyl-L-Lys chloromethylketone. Lysozyme was added at 0.2 mg g¹ cell paste, and the suspension was incubated with gentle agitation at 37°C for 1 h. The suspension was briefly sonicated and 0.02 mg mL¹ DNase was added. The suspension was incubated for 5 min at room temperature and the cell debris was removed by centrifugation. The supernatant was placed on ice, and 0.05 mL of protamine sulfate (10%, w/v) mL¹ of solution was added dropwise to the stirred solution. The cell-free extract was clarified by centrifugation and the supernatant was brought to 60% saturation by the addition of saturated ammonium sulfate and was stirred for 20 min. The precipitated protein was collected by centrifugation and dissolved in 50 mM potassium phosphate, pH 7.5, containing 0.5 M NaCl and 10 mM imidazole.

The sample was applied to a 10-mL, nickel-charged chelating Sepharose column in three batches and was equilibrated in the same buffer in which the sample was dissolved. The column was washed with buffer containing 0.10 M imidazole, and HIUHase was eluted by washing the column with buffer containing 0.14 M imidazole. Fractions containing HIUHase activity were pooled and concentrated by ultrafiltration with Centricon concentrators (Amicon, Beverly, MA).

HIUHase Assay

Purified HIUHase was assayed for its ability to catalyze the hydrolysis of HIU. HIU was generated in situ by the addition of 1.5 units of recombinant UO (Kahn and Tipton, 1998) to a 1-mL solution of 0.1 mM urate in 50 mM potassium phosphate, pH 7.5, contained in a 1-cm pathlength cuvette. Sufficient UO was used to convert the urate to HIU within 2 min. When the UO reaction appeared to reach its endpoint, as determined by monitoring the absorbance of the solution at 292 nm, HIUHase was added and the disappearance of HIU was monitored at 312 nm. The concentration of HIU at the time of addition of HIUHase was determined using a value for the extinction coefficient at 312 nm of 6,790 m¹ cm¹. In all cases, the calculated concentration of HIU was within 10% of the expected value based on the amount of urate present in the assay at zero time. The initial velocity data for the HIUHase reaction were fitted to the Michaelis-Menten equation using GraFit (Erithacus Software, Horley, Surrey, UK).

Site-Directed Mutagenesis

The E199A and E408A mutants were prepared using the QuikChange Site-Directed Mutagenesis kit according the directions provided by the manufacturer (Stratagene). The mutant proteins were purified and assayed as described for the wild-type protein.

RT-PCR Analysis of UO and HIUHase Genes

Gene-specific primers were designed for HIUHase (based on the sequence reported here), UO (accession no. AB002809), and ubiquitin (Horvath et al., 1993), and were used for RT-PCR analysis of the expression patterns of the UO and HIUHase genes. As an internal control, PCR was performed simultaneously with the ubiquitin primers. For RT reactions, 5 µg of total RNA was treated with 2 units of DNase I at 37°C for 10 min to remove traces of contaminating DNA. The DNase was inactivated by addition of 2.5 mM EDTA and incubation at 65°C for 15 min. The following primers were used for reverse transcription: 5'-TGCAGTGAAATCTTTCA-ATTGGTTCTTAAG-3' for the HIUHase gene, 5'-AAATCTGTTCAGTGT-GGC-3' for the UO gene, and 5'-ACCACCACGGAGACGGAG-3' for the ubiquitin gene. The cDNA was synthesized using 200 units of RT (Superscript; Invitrogen) in a reaction mixture containing 10 mM dithiothreitol, 1.25 mM dNTPs, and 20 mM Tris, pH 8.4, containing 50 mM KCl and 2.5 mM MgCl₂. The reaction proceeded at 42°C for 50 min and was terminated by incubation at 70°C for 15 min. The RNA present in the samples was removed by digestion with 1 unit of RNase H at 37°C for 20 min. Five microliters of each reaction mixture was used as the template for PCR amplification. The sense primers were designed from the 5' end of each gene, and the antisense primer was designed from a site internal to the one used for the cDNA synthesis. The following primers were used: 5'-GATAACTATAGCAGAG-ATGATTTT-3' for the HIUHase gene, 5'-AGTCTGACGCAGTTGAGCA-TAGAG-3' for the UO gene, and 5'-ATGCAGATATTTGTGAAGAC-3' for the ubiquitin gene. Each reaction mixture contained 5 µL of first strand cDNA, 0.25 mM dNTPs, 1.5 mM MgCl₂, 0.2 µM primers, and 1.25 units of Taq polymerase. The PCR-cycling conditions comprised an initial denaturation step at 94°C for 2 min, 30 cycles at 94°C for 30 s, 60°C for 30 s (for the HIUHase gene) or 55°C for 30 s (for the UO gene), and 72°C for 45 s, and a final elongation step at 72°C for 10 min. Control reactions omitting the reverse transcription step were also run.

Protein Expression and Western Analysis

Crude extracts were obtained by grinding leaves, roots, or nodules of soybean in 30 mM MOPS, pH 7.5, containing 2 mM EDTA, 1 mM phenylmethylsulfonyl fluoride, and 2% (w/v) polyvinylpolypyrrolidone. The extracts were filtered through two layers of Miracloth and the extracts were centrifuged at 12,000g for 20 min at 4°C. The supernatants were desalted by ultrafiltration using Centricon-10 apparatuses (Amicon). The proteins in the supernatants were resolved by electrophoresis on 7.5% polyacrylamide SDS gels, and were then electroblotted onto polyvinylidene difluoride membranes. Immunoblots were processed following standard protocols using a 1:1,000 dilution of rabbit polyclonal antibody raised against purified HIUHase (Sarma et al., 1999) and detected with goat anti-rabbit secondary antibody coupled to horseradish peroxidase (1:20,000; Amersham Biosciences). For protein visualization, the gels were stained with Coomassie Brilliant Blue, and the polyvinylidene difluoride membranes were stained with Ponceau S. The total soluble protein concentration in each sample was determined by the method of Bradford (Bradford, 1976).

Immunogold Electron Microscopy

Electron microscopy was conducted at the University of Missouri Electron Microscopy Core Facility. Twenty-one-day postinfection soybean nodules were fixed for 2 h at room temperature in 4% (w/v) paraformaldehyde and 0.5% (v/v) glutaraldehyde in 100 mM potassium phosphate buffer, pH 7.0. The nodules were then washed once for 15 min in the phosphate buffer and twice in ultrapure distilled water. Samples were dehydrated in an ethanol series (20%, 50%, 70%, 90%, and 100%, w/v) and were embedded in LR White resin. Polymerization of the resin was conducted at 60°C for 30 h. Ultrathin sections (100 nm) were cut on an ultramicrotome (Ultracut UCT; Leica, Wetzlar, Germany) and were picked up on 300-mesh nickel grids. The sections were immediately immunolabeled by placing the grids in blocking buffer containing 5% (w/v) bovine serum albumin, 5% (w/v) normal goat serum, 0.1% (w/v) coldwater fish skin gelatin, and 10 mM sodium azide in 100 mM phosphate-buffered saline for 1 h. The grids were washed for 10 min in 100 mM potassium phosphate buffer, pH 7.0, containing 150 mM NaCl, 0.1% (w/v) Aurion bovine serum albumin-c, and 10 mM sodium azide, and were then incubated for 2 h with a 1:100 dilution of the primary antibody (rabbit anti-HIUHase antibody or rabbit anti-UO antibody. The anti-UO antibody was the generous gift of Prof. Mary Alice Webb (Purdue University, West Lafayette, IN) Control grids were incubated in preimmune serum. The grids were washed and incubated for 1 h in a 1:50 dilution of goat anti-rabbit antibodies conjugated to 10-nm gold particles. Sections were poststained in 3% (w/v) aqueous uranyl acetate for 10 min and in lead citrate for 5 min before observing using a transmission electron microscope (1200 EX; JEOL, Tokyo).