INTRODUCTION

N-Acetylneuraminic acid (NeuAc)¹ is an important constituent of the carbohydrate chain of many glycoproteins and glycolipids, and has an important function in many biological recognition processes (1). The biosynthesis of NeuAc has been studied extensively in vivo and in vitro (2, 3). Two kinds of 2-epimerases have been assigned in the first step of the NeuAc biosynthetic pathway. One is N-acyl-D-glucosamine 2-epimerase (GlcNAc 2-epimerase) (EC.5.1.3.8) catalyzing the interconversion of N-acetylglucosamine (GlcNAc) and N-acetylmannosamine (ManNAc). The other one is UDP-N-acetyl-D-glucosamine 2-epimerase (UDP-GlcNAc 2-epimerase) (EC.5.1.3.14). The enzyme catalyzes the direct formation of ManNAc from UDP-N-acetyl-D-glucosamine (UDP-GlcNAc) (4). ManNAc formed by either reaction is eventually converted to cytidine 5-monophospho-N-acetylneuraminic acid (CMP-NeuAc) through the consecutive reactions: GlcNAc or UDP-GlcNAc  ManNAc  ManNAc 6-phosphate  NeuAc 9-phosphate  NeuAc  CMP-NeuAc (5, 6). CMP-NeuAc is utilized as a precursor for the synthesis of connective tissues, blood cells, and other cellular macromolecules.

Of the two 2-epimerases catalyzing the formation of ManNAc, UDP-GlcNAc 2-epimerase has been considered to be essential in the NeuAc biosynthesis (4), and the physiological significance of the GlcNAc 2-epimerase in the formation of NeuAc is not known. However, no definitive evidence denying the participation of the GlcNAc 2-epimerase in the NeuAc biosynthesis together with UDP-GlcNAc 2-epimerase has been presented.

The GlcNAc 2-epimerase has been found in porcine kidney, rat kidney, liver, spleen, brain, intestinal mucosa, thymus, pancreas, and in salivary gland (4, 5). Datta (7) and Gosh and Roseman (8) partially purified the enzyme from porcine kidney and found that the GlcNAc 2-epimerase activity is modulated by the catalytic amount of ATP. Datta (7) also reported that the GlcNAc 2-epimerase possesses two distinct interaction sites, a catalytic site for substrate and an allosteric site for ATP.

To clarify reaction mechanism and function of the GlcNAc 2-epimerase, we have isolated from porcine kidney the enzyme and its gene, and analyzed their properties. Surprisingly, the GlcNAc 2-epimerase was found to be identical with a renin-binding protein (RnBP) isolated from porcine kidney (9, 10). The physiological role of RnBP has been presumed by several investigators. Leckie and McConnell (11) suggested that RnBP is a regulator of renin because RnBP can tightly bind to renin and inhibit the renin activity. Boyd (12) proposed that RnBP is a renin carrier. Murakami et al. (13) showed that the binding of RnBP is highly specific to renin and does not interact with other acid proteases in the kidney. Furthermore, in tissues containing renin, RnBP was always detected (14). Takahashi et al. (15) reported that human RnBP gene is located in chromosome X, spans about 10 kilobases and consists of 11 exons separated by 10 introns. Tribioli et al. (16) and Faranda et al. (17) indicated that the RnBP gene is mapped in distal Xq28 chromosomal band, closely linked to a housekeeping host cell factor 1-encoding gene, and both genes are transcribed in the same direction from telomere to centromere.

In this paper, we report purification and molecular cloning of GlcNAc 2-epimerase from porcine kidney, and demonstrate the renin-binding ability of GlcNAc 2-epimerase.

EXPERIMENTAL PROCEDURES

Materials

Porcine kidney was obtained from the Kyoto wholesale market. Porcine kidney renin was purchased from Sigma. DEAE-cellulose DE-52 was from Whatman. Q Sepharose FF, Sephadex G-100, Superose 12 HR 10/30, and Mono Q HR 5/5 were from Pharmacia Biotech Inc. Butyl-Toyopearl 650M was from Tosoh. Hydroxyapatite and lysyl endopeptidase were from Wako Pure Chemical Industries. The ZAP-cDNA synthesis kit, Gigapack II Gold packaging extract, and the Exo/Mung deletion kit were from Stratagene. Other chemicals were of the highest grade commercially available.

GlcNAc 2-Epimerase Assay

GlcNAc 2-epimerase was assayed by measuring interconversion of GlcNAc and ManNAc. The reaction mixture (0.1 ml) consisted of 100 mM Tris-HCl, pH 7.4, 40 mM ManNAc, 4 mM ATP, 10 mM MgCl₂, and 20 µl of enzyme. After incubation at 37 °C for 30 min, the reaction was stopped by boiling for 3 min. The reaction products were treated with 1-phenyl-3-methyl-5-pyrazolone, and the derivatives were analyzed by HPLC (18). One unit of enzyme activity was the quantity that produced 1 µmol of GlcNAc/1 min under the assay conditions.

Purification of GlcNAc 2-Epimerase

Unless otherwise noted, all operations were carried out at 0-4 °C. The potassium phosphate buffer, pH 7.6 (KPB), used always contained 1.0 mM EDTA and 0.05% 2-mercaptoethanol. Centrifugation was carried out at 16,000 × g for 30 min, and dialysis was for 16 h against 20 mM KPB.

Kidney cortex (5.6 kg) was homogenized in 12 liters of 3.0 mM KPB. The supernatant (11.7 liters) obtained after centrifugation was diluted with 11.7 liters of cold water followed by adding 705 ml of 2.0% protamine sulfate. Precipitated materials were removed, and the supernatant (22.9 liters) was treated again with 2.3 liters of 2.0% protamine sulfate. Precipitates were washed in 5 liters of 10 mM KPB, and then the extract (5.8 l) was treated for 10 min with 58 g of bentonite, which was suspended in 580 ml of 1 mM EDTA. The mixture was centrifuged, and the supernatant was dialyzed. The dialysate (6.5 liters) was put on a DEAE-cellulose DE-52 column (25 × 13 cm). After washing the column with 50 mM KPB containing 100 mM KCl, the GlcNAc 2-epimerase was eluted with 200 mM KPB containing 150 mM KCl. Ammonium sulfate (6.7 kg) was added to the eluate (12 liters) to 80% saturation. The precipitates were dissolved in 200 ml of 20 mM KPB and dialyzed. The dialysate (280 ml) was put on a hydroxyapatite column (2.6 × 9.5 cm) equilibrated with 10 mM KPB, and the enzyme was eluted with 324 ml of the same buffer. The enzyme was treated with ammonium sulfate (80%) and dialyzed. The dialysate (40 ml) was put on a Q Sepharose column (2.6 × 10 cm), and developed with a linear gradient (1 liter) of 100-400 mM KCl in 20 mM KPB. The GlcNAc 2-epimerase was eluted at 180-220 mM KCl. Active fractions were concentrated by ultrafiltration with a ultrafilter UK-10 (Toyo Roshi Kaisha) and dialyzed. The dialysate (12 ml) was put on a Mono Q column. The column was washed with 20 mM KPB containing 200 mM KCl and developed with a linear gradient (20 ml) of 200-300 mM KCl in 20 mM KPB. The GlcNAc 2-epimerase was eluted at about 240 mM KCl. Active fractions were concentrated with ammonium sulfate (80%) and used throughout this study after dialysis.

E. coli XL1-Blue were transformed with a plasmid pEP114 (see ``cDNA Library, Subcloning, and Nucleotide Sequencing'' for subcloning) an aerobically grown for 16 h at 30 °C in Luria-Bertani medium (7 l) supplemented with 1 mM isopropyl--D-thiogalactopyranoside and 0.1 mg/ml ampicillin. The washed cells (about 32 g wet weight) were suspended in 350 ml of KPB, disrupted by sonication. After adding 93 mg of protamine to 340 ml of extract, the mixture was stirred gently for 30 min and centrifuged. The precipitates were discarded, and 3.5 g of bentonite was added to 350 ml of the supernatant. After incubation for 10 min, the precipitates were discarded. The enzyme was treated with ammonium sulfate (80%) and dialyzed. The dialysate was put on a DEAE-cellulose DE-52 column (5 × 20 cm) and then the enzyme was eluted with 200 mM KPB containing 150 mM KCl. Fractions containing GlcNAc 2-epimerase were concentrated with ammonium sulfate (80%) and dialyzed. The dialysate was put on a Sephadex G-100 column (5 × 90 cm), and then the enzyme was eluted with 20 mM KPB containing 200 mM KCl. Active fractions were concentrated with ammonium sulfate (80%) and dialyzed. The dialysate was put on a Q Sepharose column (2.6 × 10 cm), and then the enzyme was eluted with a linear gradient of 100 to 400 mM KCl in 20 mM KPB. Active fractions, which were eluted between 180 and 220 mM KCl, were concentrated with ammonium sulfate (80%) and dialyzed. The dialysate was put on a Butyl-Toyopearl 650 M column (2.6 × 10 cm) and then the enzyme was eluted with a linear gradient of 30 to 0% ammonium sulfate in 20 mM KPB. The enzyme was eluted between 19.5 and 13.5% ammonium sulfate. The active fractions were concentrated with ammonium sulfate (80%), dissolved in 8 ml of 20 mM KPB and then dialyzed.

Preparation of Anti-GlcNAc 2-Epimerase Antibody

A Japanese white rabbit was injected subcutaneously into foot pads on the back with 1 ml of emulsion containing 840 µg of the purified GlcNAc 2-epimerase from porcine kidney cortex and complete Friend's adjuvant. The same amount of GlcNAc 2-epimerase emulsion was injected at 2, 4, and 6 weeks after the first injection. After 8.5 weeks, bleeding was performed. Blood was allowed to clot at 37 °C for 30 min and was then stored at 4 °C overnight. An antiserum was separated from clots by centrifugation, added to 0.01% NaN₃, and stored at 4 °C. An IgG from the antiserum was purified with protein A-Sepharose (Pharmacia).

Inhibition of Renin by GlcNAc 2-Epimerase

Inhibitory effect on the renin activity by GlcNAc 2-epimerase was assayed according to the method of Takahashi et al. (9). Porcine kidney renin (1.4 pmol) was incubated at 37 °C for 30 min in the presence or absence of GlcNAc 2-epimerase in a mixture (0.1 ml) containing 100 mM sodium phosphate buffer, pH 6.5, 1 mM EDTA, 1 µM leupeptin, and 0.05% bovine serum albumin. The reaction was stopped with the addition of 0.9 ml of chilled sodium phosphate buffer, pH 6.5, containing 1 mM EDTA, 1 µM leupeptin, and 0.05% bovine serum albumin. Remaining renin activity was assayed by measuring the rate of formation of angiotensin I from porcine plasma angiotensinogen (19).

Binding of Renin with GlcNAc 2-Epimerase

Porcine kidney renin (1.4 pmol) was incubated at 37 °C for 30 min in the presence or absence of the GlcNAc 2-epimerase (140 pmol) in the mixture (0.1 ml) as described above. After incubation, the mixture was put on a Superose 12 HR 10/30 column equilibrated with 50 mM KPB containing 150 mM KCl and incubation products were eluted with the same buffer at 1.0 ml/min. Renin activity in each fraction (0.5 ml) was determined as described above.

cDNA Library, Subcloning, and Nucleotide Sequencing

Total RNAs were extracted from porcine kidney cortex in acid guanidinium thiocyanate-phenol-chloroform mixture (20), and poly(A)⁺ RNA was fractionated on oligo(dT)-cellulose column chromatography. The cDNA library was constructed as described by Gubler and Hoffman (21) by using  DNA phage ( ZAP) vector and E. coli SURE host. The cDNA library of 1.2 × 10⁶ clones was screened by immunostaining using an anti-GlcNAc 2-epimerase antibody raised in rabbit. Briefly, 64 positive clones were identified using a primary antibody directed against the protein of the GlcNAc 2-epimerase and the alkaline phosphatase-conjugated anti-rabbit IgG (goat). One of the positive clones were converted to phagemids carrying cDNA insert (1.4 kilobase pairs) in the sense orientation between EcoRI and XhoI sites of pBluescript SK() by in vivo excision in E. coli XL1-Blue host with R408 helper phage, and plasmid thus constructed was designated pEPI1. Sequential unidirectional deletion of the recombinant plasmid pEPI1 was carried out by cleavage at a unique SacI site of multicloning site, followed by digestion with exonuclease III and mung bean nuclease. These deletion fragments were self-ligated with T4 DNA ligase, and the recombinant plasmids were used to transform E. coli XL1-Blue. One of the recombinant plasmids thus generated had a deletion of about 60 base pairs of nucleotide sequence localized 5 terminus in cDNA 5-noncoding region and was designated pEP114. The nucleotides of cDNA were sequences in both strands by the dideoxy sequencing method of Sanger et al. (22).

Preparation of Peptide and Amino Acid Analysis

GlcNAc 2-epimerase (0.5 mg) from porcine kidney cortex was digested with 5 µg of lysyl endopeptidase in 0.5 ml of 100 mM Tris-HCl, pH 8.0, containing 4 M urea at 37 °C for 12 h. After reaction, the peptides were separated with a reversed-phase HPLC column (µBondasphere 5µ C18-300Å, 3.9 × 150 mm, Waters). N-terminal amino acids of protein or peptide were sequences by automated Edman degradation (23) on an Applied Biosystems protein sequencer (model 477A). C-terminal amino acids of protein were sequenced using carboxypeptidase Y by Klemm et al. (24). Free amino acids and hydrolyzed peptides were constant-boiling hydrochloric acid at 112 °C for 24 h were analyzed with a Hitachi amino acid analyzer (model L8500).

Electrophoresis

Sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) was done with 10-20% gradient gel as described by Laemmli (25). Proteins were stained with Coomassie Brilliant Blue R-250.

Computer Analysis

Homology analyses with other nucleotide and protein sequences were done using the FASTA comparison program (26) with the GenBank/EMBL data base of nucleotide sequence and PIR data base of amino acid sequence.

RESULTS AND DISCUSSION

We have purified the GlcNAc 2-epimerase from porcine kidney cortex (Table I). Overall purification achieved was approximately 630-fold with 3% recovery of activity. Specific activity of the enzyme was 21 units/mg of protein, which was about 3.5-fold higher than that of the preparation of Datta (7). The molecular mass of the enzyme was determined to be 45 kDa on SDS-PAGE (Fig. 1) and 93 kDa by sedimentation equilibrium (data not shown). This result suggested that the enzyme consists of two identical subunits of 45 kDa. The purified enzyme could be stored without loss of activity at 20 °C for at least 6 months in 20 mM potassium phosphate buffer, pH 7.6, containing 1.0 mM EDTA, 0.05% 2-mercaptoethanol, and 5.0% sucrose. The optimum pH and temperature were 6.8 and 47 °C, respectively, and catalyzed the interconversion of GlcNAc and ManNAc with apparent K_m values of 7.4 mM for GlcNAc, 6.3 mM for ManNAc, and 0.18 mM for an effector, ATP. ATP was not essential for the GlcNAc 2-epimerase reaction, but the activity of the enzyme was enhanced about 20-fold in the presence of ATP or deoxy-ATP.

Table I. Purification of GlcNAc 2-epimerase from porcine kidney cortex Purification procedures are shown in the text. Steps Total protein Total activity Specific activity Purification Yield mg units unit mg1 fold % Crude extract 284,000 9,500 0.033 1 100 Protamine concentration 29,000 4,300 0.15 5 45 Bentonite adsorption 5,640 4,270 0.76 23 45 DEAE-cellulose 391 1,150 2.9 88 12 Hydroxyapatite 76 620 8.2 248 7 Q Sepharose 23 463 20.1 609 5 Mono Q 16 332 20.8 630 3

A gene for the GlcNAc 2-epimerase was cloned by immunoscreening from a cDNA library for porcine kidney cortex. The plasmid with cDNA for the GlcNAc 2-epimerase was isolated, designated pEPI1 and used for the structure analysis (Fig. 2a). Fig. 2b shows the 1372-nucleotide sequence of cDNA in pEPI1. Examination of the nucleotide sequence showed an open reading frame starting at position 68 and ending at position 1273. The 1206-nucleotide reading frame encoded 402 amino acids with a predicted polypeptide of 46.4 kDa, which was closely similar to that obtained with the purified GlcNAc 2-epimerase (45 kDa on SDS-PAGE). The 3-terminal noncoding region of the cDNA is 99 nucleotides long, including a poly(A) tail of 18 nucleotides. The polyadenylation signal (AATAAA) is present in nucleotides 1327-1332. There is a potential asparagine-linked glycosylation site conforming to the consensus sequence of Asn-X-Ser at amino acid positions 228-230, although no glycosyl residues were detected in the purified GlcNAc 2-epimerase, when assayed by the method of Kondo et al. (27). N-terminal amino acid was not detectable by Edman method (23). To define N-terminal region of the GlcNAc 2-epimerase, the partial amino acid sequences were determined by using three kinds of peptides (A, B, and C) eluted at 7.0, 15.8, and 17.8 min from a lysyl endopeptidase digestion, respectively, under the HPLC conditions specified. Peptides (B and C) had sequences of Glu-Arg-Glu-Thr-Leu-Gln-Ala-Trp-Lys and Ala-Gly-Gly-Glu-Phe-Leu-Leu-Arg-His-Ala-Arg-Val-Ala-Pro-Pro-Glu-Lys corresponding to the sequences at positions 4-12 and 86-102, respectively (Fig. 2b). The smallest peptide fragment (A) contained Met, Glu, and Lys at an equimolar amounts. In deduced amino acid sequence, the amino acid composition matched only to the N-terminal region (positions 1-3) preceding the amino acid sequence located at positions 4-12. These results on peptide fragment analysis indicated that the amino terminus of the enzyme was methionine or modified methionine, and that the enzyme had no signal sequence, which was also confirmed by hydropathy plot analysis (data not shown). C-terminal amino acid sequence was determined to be -Leu-Ala by carboxypeptidase Y digestion, which corresponded to the sequence at positions 401-402 (Fig. 2b). The amino acid composition deduced from the nucleotide sequence showed a good similarity to that obtained with the GlcNAc 2-epimerase purified from porcine kidney cortex (Table II).

Table II. Amino acid composition of GlcNAc 2-epimerase purified from porcine kidney cortex and deduced from the DNA sequence Residue Number of residues Identified with purified enzyme Deduced from nucleotide sequence Asp 27a 21 Asn 4 Thr 10 9 Ser 18 18 Glu 51b 36 Gln 18 Gly 36 34 Ala 35 34 Val 23 23 Cys 6 11 Met 15 15 Ile 6 6 Leu 50 48 Tyr 13 13 Phe 21 20 Lys 15 13 His 16 15 Arg 35 35 Pro 18 16 Trp 7 13 Total 402 402 a Value of Asp + Asn. b Value of Glu + Gln.

In order to confirm that the GlcNAc 2-epimerase is the product of the cloned cDNA, the cDNA was expressed in E. coli XL1-Blue having no GlcNAc 2-epimerase activity (Table III). The recombinant plasmid pEPI1 constructed contains a 1.4-kilobase pair cDNA fragment between EcoRI and XhoI sites of a vector pBluescript SK(). When the host cells were transformed with the plasmid, the cells apparently showed the GlcNAc 2-epimerase activity, although the specific activity of the GlcNAc 2-epimerase was almost the same as that of the enzyme in porcine kidney homogenate. To enhance the expression level of the cDNA in E. coli host, about 60 base pairs of nucleotide sequence was deleted from 5-noncoding region in the cDNA. When the deletion mutant, designated plasmid pEP114, was introduced into the E. coli host, the specific activity of the GlcNAc 2-epimerase increased about 24-fold compared with that in the cells with pEPI1. The GlcNAc 2-epimerase expressed in E. coli XL1-Blue carrying pEP114 was purified approximately 57-fold from cell extracts with 19% of activity yield (Table IV). The purified GlcNAc 2-epimerase was homogeneous on SDS-PAGE (Fig. 1) and was identical with the GlcNAc 2-epimerase purified from porcine kidney cortex in molecular size (45 kDa on SDS-PAGE), affinity for substrates GlcNAc (K_m = 7.5 mM) and ManNAc (K_m = 7.8 mM), activity and stability dependences on pH and temperature, and in behavior toward an allosteric effector ATP (K_m = 0.12 mM). These results entirely support that the GlcNAc 2-epimerase purified from porcine kidney cortex is the same as that encoded by the cloned cDNA.

Table III. Expression of cDNA for GlcNAc 2-epimerase in E. coli cells Origins Activity Fold unit mg1 Porcine kidney cortex 0.03 1 E. coli XL1-Blue/pBluescript 0 0 E. coli XL1-Blue/pEPI1 0.03 1 E. coli XL1-Blue/pEP114 0.73 24

Table IV. Purification of GlcNAc 2-epimerase from E. coli XL1-Blue carrying pEP114 Steps Total protein Total activity Specific activity Purification Yield mg units unit mg1 fold % Crude extract 4,600 3,300 0.7 1 100 Bentonite adsorption 3,200 3,200 1.0 1 97 DEAE-cellulose 480 3,100 6.5 9 94 Sephadex G-100 203 2,700 13.3 19 82 Q Sepharose 96 1,300 13.5 19 39 Butyl-Toyopearl 16 640 40.0 57 19

Comparison of the nucleotide and deduced amino acid sequences for the GlcNAc 2-epimerase (Fig. 2b) with other known genes demonstrated the striking similarity to the renin-binding protein (RnBP) from porcine kidney (GenBank/) (10). The identities of nucleotide and amino acid sequences of the enzyme to those of RnBP were 99.6% and 99.0%, respectively. Only 5 nucleotides in the GlcNAc 2-epimerase gene at positions 514 (C), 667 (A), 932 (C), 1018 (C), and 1019 (G) were substituted for G, C, T, G, and C, respectively, at the same nucleotide positions in the RnBP gene. Nucleotide substitutions at positions 514, 932, 1018, and 1019 would also result in the amino acid substitutions of Asp-149 for Glu, His-289 for Tyr, Ser-317 for Arg, and Glu-318 for Gln, respectively. These results strongly imply that, in porcine kidney, GlcNAc 2-epimerase is the RnBP.

This conclusion was further evidenced by the following facts. The molecular mass of 45 kDa (SDS-PAGE) or 46.4 kDa (calculated from deduced amino acid sequence: Fig. 2b) determined for the GlcNAc 2-epimerase was closely similar to the reported molecular mass for porcine kidney RnBP (42 kDa) (9). The nucleotide sequence of porcine kidney RnBP contains four leucine residues at positions 185, 192, 199, and 206 comprising a proposed leucine zipper motif region (10, 28), which is believed to play an essential role in the formation of RnBP homodimer and RnBP-renin heterodimer (28). The same arrangement for leucine zipper motif was also recognized in the GlcNAc 2-epimerase (Fig. 2b). Furthermore, the amino acid sequence of the porcine kidney GlcNAc 2-epimerase was highly homologous to that of RnBPs from human and rad kidneys (29), with identity of 87.8% and 83.1%, respectively.

To directly confirm that the GlcNAc 2-epimerase is a RnBP, the purified porcine kidney enzyme was incubated with porcine kidney renin and renin activity was determined (Fig. 3). As anticipated, the renin (1.4 pmol) activity was apparently inhibited, and the inhibition was 50% in the presence of 10 pmol of the enzyme. Similar results were also confirmed when the GlcNAc 2-epimerase purified from E. coli XL1-Blue carrying pEP114 was used in place of the enzyme purified from porcine kidney cortex. Although the inhibition was lower level, the porcine kidney renin activity was also inhibited by rat kidney GlcNAc 2-epimerase, which was isolated by the same methods for the purification of porcine kidney enzyme (Fig. 3). This was presumably due to the low affinity of rat kidney GlcNAc 2-epimerase toward porcine kidney renin.

By the incubation of porcine kidney GlcNAc 2-epimerase with porcine kidney renin, higher molecular mass protein complex was formed (Fig. 4). Similar result was also obtained when rat kidney GlcNAc 2-epimerase was incubated with porcine kidney renin (data not shown). The molecular mass of the complex was determined to be about 70 kDa. Higher molecular mass renin (60 kDa) has been isolated from porcine kidney (12), and it was shown to be a heterodimer of renin and RnBP subunit (9, 19). The formation of higher molecular mass renin complex has also been confirmed in vitro by incubating RnBP with renin (9). Therefore, the observed complex formation between renin and the GlcNAc 2-epimerase from porcine or rat kidney may represent the higher molecular mass renin, although the molecular mass of the complex (70 kDa: Fig. 4) was slightly low compared with that calculated from the molecular masses of renin (36 kDa) (30) and the GlcNAc 2-epimerase (45 kDa, SDS-PAGE). The similar discrepancy in molecular mass has also been indicated in the case of renin and RnBP from porcine kidney (9), and the reason for this discrepancy is attributed to the unique hydrodynamic features such as leucine zipper motif and hydrophobic domain of the RnBP molecule (10, 29). It was suggested that hydrodynamic features in the RnBP molecule mediate the formation of both the RnBP-renin heterodimer (higher molecular mass renin) and the RnBP homodimer (28).

All the data including sequence similarity confirmed that GlcNAc 2-epimerase was the RnBP. However, there is a discrepancy between the tissue sources of GlcNAc 2-epimerase and RnBP. The GlcNAc 2-epimerase has been found in kidney, liver, spleen, brain, intestinal mucosa, thymus, pancreas, and in salivary gland (4, 5), but RnBP was very slightly in liver or salivary gland (14). This question should be elucidated with further investigations.

In conclusion, we have shown that, in porcine and rat kidneys and possibly in other tissues, the RnBP is the enzyme GlcNAc 2-epimerase. The identification of the RnBP as an enzyme GlcNAc 2-epimerase may open new insight into the physiological function of the RnBP, since no definitive conclusion as to the intrinsic function of RnBP has been elucidated.