Enzymatic Cleavage of the C-Glucosidic Bond of Puerarin by Three Proteins, Mn2+, and Oxidized Form of Nicotinamide Adenine Dinucleotide

Abstract

We previously isolated the human intestinal bacterium, strain PUE, which can cleave the C-glucosidic bond of puerarin to yield its aglycone daidzein and glucose. In this study, we partially purified puerarin C-glucosidic bond cleaving enzyme from the cell-free extract of strain PUE and demonstrated that the reaction was catalyzed by at least three proteins, Mn²⁺, and oxidized form of nicotinamide adenine dinucleotide (NAD⁺). We completely purified one of the proteins, called protein C, by chromatographic separation in three steps. The molecular mass of protein C was approximately 40 kDa and the amino acid sequence of its N-terminal region shows high homology to those of two putative proteins which belong to Gfo/Idh/MocA family oxidoreductase. Protein C catalyzed hydrogen-deuterium exchange reaction of puerarin to 2″-deuterated puerarin in D₂O condition, which closely resembles those of glycoside hydrolase family 4 and 109.

Glycosidic bonds between an aglycone and a sugar are generally classified into four groups including O-, N-, S-, and C-glycosidic bonds. Glycoside hydrolases that catalyze one of the O, N, S-glycosidic bond hydrolysis are categorized into about 130 families based on the similarities of amino acid sequences (Carbohydrate-Active enZYmes database at http://www.cazy.org/). Most glycoside hydrolases do not require any cofactors, however the glycoside hydrolase family 4 and 109 cleave glycosidic bonds through redox and elimination steps under the specific reaction conditions involving oxidized form of nicotinamide adenine dinucleotide (NAD⁺) as cofactor.^1,2)

Quite a unique characteristic of C-glycosidic bond in C-glycosides is found in the acidic hydrolysis or glycosidase treatments because their anomeric carbon in the sugar is directly attached to the aglycone by C−C bonding. Despite of the chemical stability of C-glycosidic bond, biotransformation of C-glycosides to its aglycone by the human intestinal bacteria were reported.^3–12) A few papers described the partial purification of C-glycosidic bond cleaving enzymes.^13,14) According to the paper of Sanugul et al.,¹⁴⁾ one of the enzymes for C-glycosidic bond cleavage reaction of mangiferin (xanthone C-glucoside) consisted of two proteins which required Mn²⁺ and undetermined low molecular cofactor(s).

C-Glycosides are distributed in a lot of medicinal plants. Among them, puerarin, isoflavone C-glucoside, is a major flavonoid contained in the roots of Pueraria lobata Ohwi (Leguminosae), which is a well-known herbal drug Puerariae Radix, so called “Kakkon” in Japanese and has been used as a diaphoretic, antifebrile and antispasmodic¹⁵⁾ in Japan and East Asia.

After oral administration of puerarin in the case of rats, two metabolic pathways were reported by Prasain et al.¹⁶⁾ One showed that puerarin was first transferred to the portal circulation by the sodium-dependent glucose transporter, and second entered itself into the bloodstream as an unmetabolized form. Another depicted that puerarin was metabolized to its aglycone daidzein, equol and related compounds by the intestinal bacteria. The latter pathway suggested that the key reaction determining the metabolic fate of puerarin was C-glycosidic bond cleavage catalyzed by the intestinal bacteria. To our best knowledge, a dozen of Pueraria isoflavones secondary metabolized from puerarin show various and significant bioactivities, however, the enzyme(s) performed the C-glycosidic bond cleavage in puerarin as the initial key reaction has not been characterized, yet.

We previously isolated the human intestinal bacterium, strain PUE, which can cleave the C-glucosidic bond of puerarin to yield daidzein and glucose^17,18) (Fig. 1). In the course of our investigation for the mechanistic study of biotransformation of C-glycosides, the purification and characterization of the novel C-glycosidic bond cleaving enzyme from strain PUE were partially accomplished. In this paper we describes an elucidation of the new enzyme composed of at least three proteins, and an existence of the cofactors associated with this cleaving reaction. Furthermore, we determined the amino acid sequence of N-terminal region of one purified protein.

Fig. 1. Cleavage of the C-Glucosidic Bond of Puerarin by Strain PUE

Materials and Methods

General

A bacterium was incubated in anaerobic incubator EAN-140 (Tabai Co., Osaka, Japan). UV spectra were measured by a UV-2200 UV/VIS recording spectrophotometer (Shimadzu Co., Kyoto, Japan).

Chemicals and Materials

Strain PUE was isolated from human feces as previously described.¹⁷⁾ Puerarin was isolated from the roots of Pueraria lobata (Willd.) Ohwi. General anaerobic medium (GAM) broth was purchased from Nissui Seiyaku Co. (Tokyo, Japan). NAD⁺ was purchased from Oriental Yeast Co. (Tokyo, Japan). Manganese(II) chloride tetrahydrate was purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan).

Preparation of Cell-Free Extract

Strain PUE was cultured under anaerobic conditions at 37°C for 14 h in 1 L of GAM broth containing 0.3 mm puerarin as enzyme inducer. The bacterial cells were collected by centrifugation and suspended in 50 mL of 50 mm potassium phosphate buffer (pH 7.4). Cells were disrupted by sonication on ice and centrifuged at 10000×g for 60 min at 4°C to obtain a supernatant as cell-free extract.

Purification of Puerarin C-Glucosidic Bond Cleaving Enzyme (Proteins A, B, and C)

All purification procedures were carried out at 0–4°C. All column chromatography steps for enzyme purification were performed using ÄKTA purifier HPLC system (GE Healthcare, Buckinghamshire, U.K.). Prepared cell-free extract was adjusted to 25% saturation of (NH₄)₂SO₄ followed by centrifugation at 10000×g for 10 min. The supernatant was applied to a HiPrep Butyl FF 16/10 column (GE Healthcare) at a flow rate of 5 mL/min. Bound proteins were eluted with a linear gradient of 1–0 m (NH₄)₂SO₄ in 100 mm potassium phosphate buffer (pH 7.3). Two protein fractions (Fr. I and Fr. II) essential for the activity were pooled and further enzyme purification were performed as follows.

Partial Purification of Proteins A and B from Fr. I: Fraction I was concentrated and buffer exchanged into 50 mm potassium phosphate buffer (pH 7.4) using Amicon Ultra-15 centrifugal filter devices (10K MWCO, Millipore). It was applied to a Mono Q 4.6/100 PE column (GE Healthcare) at a flow rate of 1 mL/min. Bound proteins were eluted with a linear gradient of 100–400 mm NaCl in 50 mm potassium phosphate buffer (pH 7.4). Two protein fractions (proteins A and B) essential for the activity were pooled.

Complete Purification of Protein C from Fr. II: Fraction II was concentrated and buffer exchanged into 10 mm potassium phosphate buffer (pH 7.4) using Amicon Ultra-15 centrifugal filter devices. It was applied to a Hydroxyapatite-MP column (Kanto Chemical, Tokyo, Japan, 8×100 mm) at a flow rate of 1 mL/min. Bound proteins were eluted with a linear gradient of 10–300 mm potassium phosphate buffer (pH 7.4). The active fractions were pooled and concentrated. It was applied to a Mono Q 4.6/100 PE column at a flow rate of 1 mL/min. Bound proteins were eluted with a linear gradient of 0–400 mm NaCl in 50 mm potassium phosphate buffer (pH 7.4). The active fractions were pooled and used as a purified protein C in the following experiments.

Measurement of Enzyme Activity

Three protein fractions (proteins A, B and C), Mn²⁺ and NAD⁺ were essential for puerarin C-glucosidic bond cleaving reaction in our result. In the case of determining the active fraction of protein C, beforehand prepared proteins A and B were added to the reaction mixture. The reaction mixture consisted of enzyme solution, puerarin (0.5 mm), MnCl₂ (1 mm) and NAD⁺ (1 mm) in a final volume of 100 µL. It was incubated at 37°C for several hours, and extracted twice with 200 µL of 1-butanol (saturated with water and acidified with 0.1% acetic acid). The extract was evaporated to dryness, dissolved in 50% MeOH (100 µL), and the amount of daidzein (metabolite) was measured by HPLC. HPLC conditions were as follows: column, COSMOSIL 5C₁₈-MS-II (Nacalai Tesque, Kyoto, Japan, 4.6×150 mm); flow rate, 1 mL/min; detection, 256 nm; mobile phase, 0.1% trifluoroacetic acid in 50% MeOH (isocratic); injection volume, 20 µL.

Determination of Cofactors

The reaction mixture consisted of partially purified proteins A (61 µg), B (69 µg), and C (21 µg), puerarin (0.5 mm), in the presence of both Mn²⁺ (1 mm) and cofactor (NAD⁺, reduced nicotinamide adenine dinucleotide (NADH), NAD phosphate (NADP⁺) or reduced NAD phosphate (NADPH), 1 mm, respectively) or in the absence of either one, in a final volume of 100 µL. After two hours incubation at 37°C, the enzyme activity was measured as described above.

Effects of Three Protein Fractions on Enzyme Activity

The reaction mixture consisted of a couple of partially purified proteins (A: 61 µg, B: 69 µg, C: 21 µg), puerarin (0.5 mm), Mn²⁺ (1 mm) and NAD⁺ (1 mm) in a final volume of 100 µL. After two hours incubation at 37°C, the enzyme activity was measured as described above.

Protein Measurement

The protein content was determined by the method of Bradford, using Protein Quantification Kit-Rapid (Dojindo Molecular Technologies, Inc., Kumamoto, Japan).

Determination of Molecular Weight

The molecular weight of protein C was analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) using c-Pagel 12.5% gel (ATTO, Tokyo, Japan). Protein Molecular Weight Marker (broad) (TaKaRa Bio Inc., Shiga, Japan) was used as standard. The protein bands were stained with Coomassie brilliant blue R250.

Determination of Amino Acid Sequence in N-Terminal Region

After the solution of purified protein C was desalted, it was spotted onto a polybrene-coated glass fiber disk. The amino acid sequence of its N-terminal region was determined by the Edman degradation method using the protein sequencer PPSQ-21 (Shimadzu, Kyoto, Japan). Homology searches were performed using the BLASTp search of the NCBI database.

Hydrogen-Deuterium Exchange Reaction

To the D₂O (870 µL) was added purified protein C (43 µg), puerarin (0.5 mm), Mn²⁺ (1 mm) and NAD⁺ (1 mm) in a final volume of 1 mL (87% D₂O condition). After the reaction mixture was incubated at 37°C for 20 h, it was extracted three times with 500 µL of 1-butanol (saturated with water and acidified with 0.1% acetic acid). The extract was evaporated to dryness, dissolved in CD₃OD and ¹H-NMR of the residue was taken on Varian 400-MR (Varian Inc., CA, U.S.A.).

Results

Effects of Cofactors on Enzymatic Activity

The cell-free extract of strain PUE showed weak puerarin C-glucosidic bond cleaving activity, while by addition of Mn²⁺ to the reaction mixture the activity was significantly increased. From the follow-up study, we found that the presence of Mn²⁺ together with some low molecular cofactor(s) furnished more accelerated reaction. In order to identify the cofactor(s), we tried to use combined cocktail of the partially purified enzyme fractions for C–C bond cleaving reaction (Fig. 2). The best condition was found in the presence of both Mn²⁺ and NAD⁺, but in the absence of either Mn²⁺ or NAD⁺, the cleavage underwent fairly. The effect of NADH, NADP⁺, and NADPH were also examined in the same conditions, however, these cofactors showed a slight effect on the cleaving activity.

Fig. 2. Effects of Cofactors on Puerarin C-Glucosidic Bond Cleaving Activity of Partially Purified Enzyme

The reaction mixture consisted of partially purified proteins A (61 µg), B (69 µg), and C (21 µg), puerarin (0.5 mm), in the presence of both Mn²⁺ (1 mm) and cofactor (NAD⁺, NADH, NADP⁺ or NADPH, 1 mm, respectively) or in the absence of either one, in a final volume of 100 µL. After 2 h incubation at 37°C, the enzyme activity was measured.

Partial Purification of Puerarin C-Glucosidic Bond Cleaving Enzyme (Proteins A, B, and C)

We purified puerarin C-glucosidic bond cleaving enzyme from the cell-free extract of strain PUE by using column chromatography in several steps (Fig. 3). First step column chromatography was performed using a HiPrep Butyl FF 16/10 column (Fig. 3a), to yield in disappearance of the bond cleaving activity in any of one fraction alone. However, the activity was restored in the mixed condition of two fractions (Fr. I and Fr. II), Mn²⁺ and NAD⁺, indicating that at least two protein fractions were essential for the bond cleaving reaction.

Fig. 3. Purification of Puerarin C-Glucosidic Bond Cleaving Enzyme

The arrow represents the active protein fractions (proteins A, B, and C). (a) HiPrep Butyl FF 16/10 column chromatography of cell-free extract, (b) Mono Q 4.6/100 PE column chromatography of Fr. I, (c) Hydroxyapatite-MP column chromatography of Fr. II, (d) Mono Q 4.6/100 PE column chromatography of the active fraction from hydroxyapatite-MP.

Further enzyme purification of Fr. I was conducted using a Mono Q 4.6/100 PE column (Fig. 3b). We found that the essential protein contained in Fr. I was not of single protein and they were separated again to two protein fractions, which were called proteins A and B, respectively. The other essential protein contained in Fr. II was called protein C. At this stage of enzyme purification, the proteins A, B, and C were partially purified proteins.

Figure 4 shows the requirement of three protein fractions for C-glucosidic bond cleaving reaction. We used partially purified proteins A (61 µg), B (69 µg), and C (21 µg). If the three proteins were combined with cofactors, puerarin C-glucosidic bond cleaving activity was 12.5 nmol/h. The absence of any proteins resulted in remarkable decrease of the activity, which suggested that all of the three proteins were exactly essential for the bond cleaving reaction. We had tried to identify any reaction intermediate such as daidzin (daidzein 7-O-β-d-glucopyranoside) by HPLC analysis after incubation of puerarin with the partially purified protein(s), but unfortunately we could not detect it yet.

Fig. 4. Requirement of Three Protein Fractions for C-Glucosidic Bond Cleaving Reaction

The reaction mixture consisted of a couple of partially purified proteins (A: 61 µg, B: 69 µg, C: 21 µg), puerarin (0.5 mm), Mn²⁺ (1 mm), and NAD⁺ (1 mm) in a final volume of 100 µL. After 2 h incubation at 37°C, the enzyme activity was measured.

Complete Purification and Amino Acid Sequence of N-Terminal Region of Protein C

We purified protein C from the cell-free extract of strain PUE by chromatographic separation in three steps, such as a Butyl FF column (Fig. 3a), Hydroxyapatite-MP column (Fig. 3c) and Mono Q column (Fig. 3d). The active fraction obtained from the purification by the Mono Q column showed single band of approximately 40 kDa on SDS-PAGE (Fig. 5) and the amount of obtained protein C was 430 µg.

Fig. 5. SDS-PAGE of the Protein C Obtained by Various Steps of Column Chromatography

Lane M, marker proteins; lane 1, cell-free extract; lane 2, the active fraction (Fr. II) obtained from HiPrep Butyl FF 16/10 column chromatography; lane 3, the active fraction from Hydroxyapatite-MP column chromatography; lane 4, the active fraction from Mono Q 4.6/100 PE column chromatography.

The N-terminal amino acid sequence of protein C was determined by the Edman degradation method as SKLKIGIIGXGGIANQKHFPALKNNADLNE. A BLASTp search showed that the amino acid sequence of N-terminal region showed high homology to those of two putative proteins; 93% identity with oxidoreductase, Gfo/Idh/MocA family from Clostridium sp. D5 (ZP_08128353.1) and 90% identity with hypothetical protein HMPREF0988_02145 from Lachnospiraceae bacterium 1_4_56FAA (ZP_08616560.1). Both putative proteins belong to the Gfo/Idh/MocA family oxidoreductases, in which represented glucose-fructose oxidoreductase,¹⁹⁾ myo-inositol 2-dehydrogenase,²⁰⁾ glycoside hydrolase family 109,²⁾ and so on.

Hydrogen-Deuterium Exchange Reaction Catalyzed by Protein C

¹H-NMR analysis of puerarin was performed after incubation of puerarin with protein C, Mn²⁺, and NAD⁺ in 87% D₂O condition (Fig. 6B). The intensity of H-2″ signal (δ 4.11 ppm) was shortened compared to the ¹H-NMR spectrum of the standard puerarin (Fig. 6A), indicating hydrogen-deuterium exchange reaction was occurred at the position of H-2″. This result was also confirmed by analyzing the spin-spin coupling pattern of H-1″ signal (δ 5.09 ppm) which was changed from doublet to singlet. A control experiment was conducted by the same experimental procedure except for the absence of protein C in the reaction mixture, but no reaction was observed. These results suggested that protein C catalyzed hydrogen-deuterium exchange reaction of puerarin to 2″-deuterated puerarin in D₂O condition.

Fig. 6. Hydrogen-Deuterium Exchange Reaction Catalyzed by Protein C

(A) ¹H-NMR spectrum of puerarin standard, (B) ¹H-NMR spectrum of puerarin after incubation with protein C, Mn²⁺, and NAD⁺ in 87% D₂O condition.

Discussion

A human intestinal bacterium, strain PUE can cleave the C-glucosidic bond of puerarin, however the enzyme involved in this reaction have not been characterized. In this study, we partially purified puerarin C-glucosidic bond cleaving enzyme and demonstrated that the reaction was catalyzed by at least three proteins, Mn²⁺, and NAD⁺ (Figs. 2, 4). Another intestinal bacterium, strain MANG, cleaved the C-glucosidic bond of mangiferin and the enzyme involved in this reaction was partially purified.¹⁴⁾ These two C-glucosidic bond cleaving reactions have some similarities, such as the requirement of multiple proteins, Mn²⁺ and some cofactors. These results suggested that the C-glucosidic bond cleavage in both puerarin and mangiferin might proceed via the similar reaction mechanisms.

We purified a single protein, called protein C, from the cell-free extract of strain PUE. The molecular mass of protein C was approximately 40 kDa and it may belong to Gfo/Idh/MocA family, known as oxidoreductases based on the amino acid sequence of its N-terminal region (Fig. 5). Hydrogen-deuterium exchange reaction was observed at the C-2″ of puerarin when it was incubated with protein C in D₂O condition (Fig. 6). The results is closely resembled those of glycoside hydrolase family 4 and 109.^1,2) These glycoside hydrolases involve in redox reaction of carbohydrate at C-3 using NAD⁺ as cofactor. After the oxidation of hydroxyl at C-3, hydrogen at C-2 was exchanged to deuterium from D₂O because the acidity of H-2 is increased by an adjacent carbonyl group. Finally, the reaction was completed by reduction of ketone at C-3 to hydroxyl using the hydride generated from NADH. The enzymatic function of protein C showed some similarities to the glycoside hydrolase family 4 and 109, reasonably supporting the fact that puerarin is hydrolyzed to daidzein and glucose.¹⁸⁾

Puerarin C-glucosidic bond cleavage reaction still remained some unknown reaction mechanism such as unclear roles of three proteins, together with the interaction of Mn²⁺ and NAD⁺. Protein C alone could not cleave the C-glucosidic bond without proteins A and B. Further studies will be needed to conclude the mechanism of puerarin C-glucosidic bond cleavage reaction, and to complete the purification and characterization of proteins A and B.

References

• 1)  Yip VL, Varrot A, Davies GJ, Rajan SS, Yang X, Thompson J, Anderson WF, Withers SG. An unusual mechanism of glycoside hydrolysis involving redox and elimination steps by a family 4 beta-glycosidase from Thermotoga maritima. J. Am. Chem. Soc., 126, 8354–8355 (2004).
• 2)  Liu QP, Sulzenbacher G, Yuan H, Bennett EP, Pietz G, Saunders K, Spence J, Nudelman E, Levery SB, White T, Neveu JM, Lane WS, Bourne Y, Olsson ML, Henrissat B, Clausen H. Bacterial glycosidases for the production of universal red blood cells. Nat. Biotechnol., 25, 454–464 (2007).
• 3)  Hattori M, Kanda T, Shu YZ, Akao T, Kobashi K, Namba T. Metabolism of barbaloin by intestinal bacteria. Chem. Pharm. Bull., 36, 4462–4466 (1988).
• 4)  Hattori M, Shu YZ, el-Sedawy AI, Namba T, Kobashi K, Tomimori T. Metabolism of homoorientin by human intestinal bacteria. J. Nat. Prod., 51, 874–878 (1988).
• 5)  Hattori M, Shu Y, Tomimori T, Kobashi K, Namba T. A bacterial cleavage of the C-glucosyl bond of mangiferin and bergenin. Phytochemistry, 28, 1289–1290 (1989).
• 6)  Che QM, Akao T, Hattori M, Kobashi K, Namba T. Metabolism of aloesin and related compounds by human intestinal bacteria: a bacterial cleavage of the C-glucosyl bond and the subsequent reduction of the acetonyl side chain. Chem. Pharm. Bull., 39, 704–708 (1991).
• 7)  Che QM, Akao T, Hattori M, Tsuda Y, Namba T, Kobashi K. Barbaloin stimulates growth of Eubacterium sp. strain BAR, a barbaloin-metabolizing bacterium from human feces. Chem. Pharm. Bull., 39, 757–760 (1991).
• 8)  Che QM, Akao T, Hattori M, Kobashi K, Namba T. Isolation of a human intestinal bacterium capable of transforming barbaloin to aloe-emodin anthrone. Planta Med., 57, 15–19 (1991).
• 9)  Meselhy MR, Kadota S, Hattori M, Namba T. Metabolism of safflor yellow B by human intestinal bacteria. J. Nat. Prod., 56, 39–45 (1993).
• 10)  Li Y, Meselhy MR, Wang LQ, Ma CM, Nakamura N, Hattori M. Biotransformation of a C-glycosylflavone, abrusin 2″-O-beta-d-apioside, by human intestinal bacteria. Chem. Pharm. Bull., 48, 1239–1241 (2000).
• 11)  Sanugul K, Akao T, Li Y, Kakiuchi N, Nakamura N, Hattori M. Isolation of a human intestinal bacterium that transforms mangiferin to norathyriol and inducibility of the enzyme that cleaves a C-glucosyl bond. Biol. Pharm. Bull., 28, 1672–1678 (2005).
• 12)  Braune A, Blaut M. Deglycosylation of puerarin and other aromatic C-glucosides by a newly isolated human intestinal bacterium. Environ. Microbiol., 13, 482–494 (2011).
• 13)  Saito K. Enzyme-catalyzed cleavage of the C-glycosidic linkage to the aromatic ring-A of 3′,4′,5,7-tetrahydroxyflavone 8-C-glycoside. Biochim. Biophys. Acta, 1035, 340–347 (1990).
• 14)  Sanugul K, Akao T, Nakamura N, Hattori M. Two proteins, Mn²⁺, and low molecular cofactor are required for C-glucosyl-cleavage of mangiferin. Biol. Pharm. Bull., 28, 2035–2039 (2005).
• 15)  Harada M, Ueno K. Pharmacological studies on pueraria root. I. Fractional extraction of pueraria root and identification of its pharmacological effects. Chem. Pharm. Bull., 23, 1798–1805 (1975).
• 16)  Prasain JK, Jones K, Brissie N, Moore R, Wyss JM, Barnes S. Identification of puerarin and its metabolites in rats by liquid chromatography-tandem mass spectrometry. J. Agric. Food Chem., 52, 3708–3712 (2004).
• 17)  Jin JS, Nishihata T, Kakiuchi N, Hattori M. Biotransformation of C-glucosylisoflavone puerarin to estrogenic (3S)-equol in co-culture of two human intestinal bacteria. Biol. Pharm. Bull., 31, 1621–1625 (2008).
• 18)  Nakamura K, Nishihata T, Jin JS, Ma CM, Komatsu K, Iwashima M, Hattori M. The C-glucosyl bond of puerarin was cleaved hydrolytically by a human intestinal bacterium strain PUE to yield its aglycone daidzein and an intact glucose. Chem. Pharm. Bull., 59, 23–27 (2011).
• 19)  Zachariou M, Scopes RK. Glucose-fructose oxidoreductase, a new enzyme isolated from Zymomonas mobilis that is responsible for sorbitol production. J. Bacteriol., 167, 863–869 (1986).
• 20)  Berman T, Magasanik B. The pathway of myo-inositol degradation in Aerobacter aerogenes. Dehydrogenation and dehydration. J. Biol. Chem., 241, 800–806 (1966).