Multi-enzymatic one-pot reduction of dehydrocholic acid to 12-keto-ursodeoxycholic acid with whole-cell biocatalysts.

Ursodeoxycholic acid (UDCA) is a bile acid of industrial interest as it is used as an agent for the treatment of primary sclerosing cholangitis and the medicamentous, non-surgical dissolution of gallstones. Currently, it is prepared industrially from cholic acid following a seven-step chemical procedure with an overall yield of <30%. In this study, we investigated the key enzymatic steps in the chemo-enzymatic preparation of UDCA-the two-step reduction of dehydrocholic acid (DHCA) to 12-keto-ursodeoxycholic acid using a mutant of 7ß-hydroxysteroid dehydrogenase (7ß-HSDH) from Collinsella aerofaciens and 3α-hydroxysteroid dehydrogenase (3α-HSDH) from Comamonas testosteroni. Three different one-pot reaction approaches were investigated using whole-cell biocatalysts in simple batch processes. We applied one-biocatalyst systems, where 3α-HSDH, 7ß-HSDH, and either a mutant of formate dehydrogenase (FDH) from Mycobacterium vaccae N10 or a glucose dehydrogenase (GDH) from Bacillus subtilis were expressed in a Escherichia coli BL21(DE3) based host strain. We also investigated two-biocatalyst systems, where 3α-HSDH and 7ß-HSDH were expressed separately together with FDH enzymes for cofactor regeneration in two distinct E. coli hosts that were simultaneously applied in the one-pot reaction. The best result was achieved by the one-biocatalyst system with GDH for cofactor regeneration, which was able to completely convert 100 mM DHCA to >99.5 mM 12-keto-UDCA within 4.5 h in a simple batch process on a liter scale.

One-step synthesis of 12-ketoursodeoxycholic acid from dehydrocholic acid using a multienzymatic system.

12-ketoursodeoxycholic acid (12-keto-UDCA) is a key intermediate for the synthesis of ursodeoxycholic acid (UDCA), an important therapeutic agent for non-surgical treatment of human cholesterol gallstones and various liver diseases. The goal of this study is to develop a new enzymatic route for the synthesis 12-keto-UDCA based on a combination of NADPH-dependent 7ß-hydroxysteroid dehydrogenase (7ß-HSDH, EC 1.1.1.201) and NADH-dependent 3α-hydroxysteroid dehydrogenase (3α-HSDH, EC 1.1.1.50). In the presence of NADPH and NADH, the combination of these enzymes has the capacity to reduce the 3-carbonyl- and 7-carbonyl-groups of dehydrocholic acid (DHCA), forming 12-keto-UDCA in a single step. For cofactor regeneration, an engineered formate dehydrogenase, which is able to regenerate NADPH and NADH simultaneously, was used. All three enzymes were overexpressed in an engineered expression host Escherichia coli BL21(DE3)Δ7α-HSDH devoid of 7α-hydroxysteroid dehydrogenase, an enzyme indigenous to E. coli, in order to avoid formation of the undesired by-product 12-chenodeoxycholic acid in the reaction mixture. The stability of enzymes and reaction conditions such as pH value and substrate concentration were evaluated. No significant loss of activity was observed after 5 days under reaction condition. Under the optimal condition (10 mM of DHCA and pH 6), 99 % formation of 12-keto-UDCA with 91 % yield was observed.

Novel whole-cell biocatalysts with recombinant hydroxysteroid dehydrogenases for the asymmetric reduction of dehydrocholic acid.

Ursodeoxycholic acid is an important pharmaceutical so far chemically synthesized from cholic acid. Various biocatalytic alternatives have already been discussed with hydroxysteroid dehydrogenases (HSDH) playing a crucial role. Several whole-cell biocatalysts based on a 7α-HSDH-knockout strain of Escherichia coli overexpressing a recently identified 7ß-HSDH from Collinsella aerofaciens and a NAD(P)-bispecific formate dehydrogenase mutant from Mycobacterium vaccae for internal cofactor regeneration were designed and characterized. A strong pH dependence of the whole-cell bioreduction of dehydrocholic acid to 3,12-diketo-ursodeoxycholic acid was observed with the selected recombinant E. coli strain. In the optimal, slightly acidic pH range dehydrocholic acid is partly undissolved and forms a suspension in the aqueous solution. The batch process was optimized making use of a second-order polynomial to estimate conversion as function of initial pH, initial dehydrocholic acid concentration, and initial formate concentration. Complete conversion of 72 mM dehydrocholic acid was thus made possible at pH 6.4 in a whole-cell batch process within a process time of 1 h without cofactor addition. Finally, a NADH-dependent 3α-HSDH from Comamonas testosteroni was expressed additionally in the E. coli production strain overexpressing the 7ß-HSDH and the NAD(P)-bispecific formate dehydrogenase mutant. It was shown that this novel whole-cell biocatalyst was able to convert 50 mM dehydrocholic acid directly to 12-keto-ursodeoxycholic acid with the formation of only small amounts of intermediate products. This approach may be an efficient process alternative which avoids the costly chemical epimerization at C-7 in the production of ursodeoxycholic acid.

Triketocholanoic (dehydrocholic) acid. Hepatic metabolism and effect on bile flow and biliary lipid secretion in man.

[24-(14)C]Dehydrocholic acid (triketo-5-beta-cholanoic acid) was synthesized from [24-(14)C]cholic acid, mixed with 200 mg of carrier, and administered intravenously to two patients with indwelling T tubes designed to permit bile sampling without interruption of the enterohepatic circulation. More than 80% of infused radioactivity was excreted rapidly in bile as glycine- and taurine-conjugated bile acids. Radioactive products were identified, after deconjugation, as partially or completely reduced derivatives of dehydrocholic acid. By mass spectrometry, as well as chromatography, the major metabolite (about 70%) was a dihydroxy monoketo bile acid (3alpha,7alpha-dihydroxy-12-keto-5beta-cholanoic acid); a second metabolite (about 20%) was a monohydroxy diketo acid (3alpha-hydroxy-7,12-di-keto-5beta-cholanoic acid); and about 10% of radioactivity was present as cholic acid. Reduction appeared to have been sequential (3 position, then 7 position, and then 12 position) and stereospecific (only alpha epimers were recovered). Bile flow, expressed as the ratio of bile flow to bile acid excretion, was increased after dehydrocholic acid administration. It was speculated that the hydroxy keto metabolites are hydrocholeretics. The proportion of cholesterol to lecithin and bile acids did not change significantly after dehydrocholic acid administration. In vitro studies showed that the hydroxy keto metabolites dispersed lecithin poorly compared to cholate; however, mixtures of cholate and either metabolite had dispersant properties similar to those of cholate alone, provided the ratio of metabolite to cholate remained below a value characteristic for each metabolite. These experiments disclose a new metabolic pathway in man, provide further insight into the hydrocholeresis induced by keto bile acids, and indicate the striking change in pharmacologic and physical properties caused by replacement of hydroxyl by a keto substituent in the bile acid molecule.

Glucose reabsorption from bile. Evidence for a biliohepatic circulation.

Glucose is absent from human bile and present in low concentrations in bile from the rat. To study the mechanisms of this blood-bile glucose concentration difference, infusions of glucose were administered i.v. to 300-400 g male Sprague-Dawley rats with ligated renal pedicles and to two postcholecystectomy patients with indwelling t-tubes. Glucose was assayed in plasma, bile, and rat liver by a hexokinase method specific for D-glucose. In man, glucose was detected in bile when plasma glucose increased above 350 mg/100 ml. In animals studies, low concentrations of bile glucose were observed at plasma levels between 100 and 300 mg/100 ml. However, when plasma concentrations increased between 400 and 900 mg/100 ml, glucose appeared more rapidly in bile, defining by extrapolation an apparent plasma glucose threshold of 280 mg/100 ml. Intraportal phlorizin, a competitive inhibitor of glucose transport, significantly increased bile glucose concentrations. Plasma-bile concentration differences were also observed in rats after i.v. [3-14C]O-methyl glucose (3-O-MG) but not after [3H]mannitol. Hepatic glucose levels were never lower than plasma levels and liver-plasma 3-O-MG ratios were 0.92 +/- 0.22 indicating that entry of glucose and 3-O-MG into hepatocyte water was not limiting. Furthermore, when sodium dehydrocholate augmented canalicular secretion, biliary glucose excretion increased proportionally suggesting that glucose entry into bile was not impeded. When estimates of hepatic glucose secretion were compared with biliary glucose excretion, the latter increased progressively when estimated secretion rates exceeded 50 micrograms/min or when phlorizin was given. Finally, during bile stop-flow experiments, [3-14C]O-MG and [14C]glucose were selectively removed from bile compared with [3H]mannitol. The findings suggest that glucose and 3-O-MG are reabsorbed from bile after entry at the hepatocyte, accounting for their low bile-plasma ratio. The biliary glucose transport process may be described by Michaelis-Menten kinetics and is analogous to recently defined kinetics for renal tubular reabsorption of glucose. These studies provide evidence that certain products of bile secretion may undergo a "biliohepatic" circulation.

Regio- and stereoselective reductions of dehydrocholic acid.

Dehydrocholic acid (DHCA), an unnatural bile acid, is manufactured by oxidation of cholic acid. Its biotransformation by two basidiomycetes (Trametes hirsuta and Collybia velutipes) is reported. These mycelia showed different affinities for the substrate and selectivities of attack: T. hirsuta in particular regio- and stereoselectively reduced the 3-keto group to yield 3 alpha-hydroxy-7,12-diketo-5 beta-cholan-24-oic acid (7,12-diketolithocolic acid) as the main product. A number of different chemical reductions were carried out on DHCA; among them hydrogenation with Raney Nickel in water under high-intensity ultrasound proved highly regio- and stereoselective, yielding 7,12-diketolithocolic acid exclusively. (1)H and (13)C resonances were assigned in details thanks to a series of 1D and 2D NMR runs including DEPT, NOESY, H-H COSY, gHSQC and gHMBC.

Enzymes involved in the formation of 3 beta, 7 beta-dihydroxy-12-oxo-5 beta-cholanic acid from dehydrocholic acid by Ruminococcus sp. obtained from human intestine.

Ruminococcus sp. PO1-3 from human intestinal flora reduced dehydrocholic acid to 3 beta-hydroxy-7,12-dioxo-5 beta-cholanic acid by means of the enzyme 3 beta-hydroxysteroid dehydrogenase (Akao, T., Akao, T., Hattori, M., Namba, T. and Kobashi, K. (1986) J. Biochem. (Tokyo) 99, 1425-1431). This bacterium and its crude extract gave rise to another product, showing a lower RF value on TLC, from dehydrocholic acid. The product was identified as 3 beta, 7 beta-dihydroxy-12-oxo-5 beta-cholanic acid. The crude extract reduced 7-ketolithocholic acid and its methyl ester, but not 6-ketolithocholic acid and 12-ketochenodeoxycholic acid, in the presence of NADPH, and oxidized ursodeoxycholic acid and beta-muricholic acid, but not cholic acid, chenodeoxycholic acid, deoxycholic acid and hydrocholic acid, in the presence of NADP+. Therefore, besides 3 beta-hydroxysteroid dehydrogenase, 7 beta-hydroxysteroid dehydrogenase was shown to be present in this bacterium. The two dehydrogenases were clearly separated from each other by butyl-Toyopearl 650 M column chromatography. From dehydrocholic acid, 7 beta-hydroxy-3,12-dioxo-5 beta-cholanic acid was produced by 7 beta-hydroxysteroid dehydrogenase and 3 beta, 7 beta-dihydroxy-12-oxo-5 beta-cholanic acid was produced by combination of two enzymes, 7 beta- and 3 beta-hydroxysteroid dehydrogenase.

Dynamic mechanistic modeling of the multienzymatic one-pot reduction of dehydrocholic acid to 12-keto ursodeoxycholic acid with competing substrates and cofactors.

Ursodeoxycholic acid (UDCA) is a bile acid which is used as pharmaceutical for the treatment of several diseases, such as cholesterol gallstones, primary sclerosing cholangitis or primary biliary cirrhosis. A potential chemoenzymatic synthesis route of UDCA comprises the two-step reduction of dehydrocholic acid to 12-keto-ursodeoxycholic acid (12-keto-UDCA), which can be conducted in a multienzymatic one-pot process using 3α-hydroxysteroid dehydrogenase (3α-HSDH), 7ß-hydroxysteroid dehydrogenase (7ß-HSDH), and glucose dehydrogenase (GDH) with glucose as cosubstrate for the regeneration of cofactor. Here, we present a dynamic mechanistic model of this one-pot reduction which involves three enzymes, four different bile acids, and two different cofactors, each with different oxidation states. In addition, every enzyme faces two competing substrates, whereas each bile acid and cofactor is formed or converted by two different enzymes. First, the kinetic mechanisms of both HSDH were identified to follow an ordered bi-bi mechanism with EBQ-type uncompetitive substrate inhibition. Rate equations were then derived for this mechanism and for mechanisms describing competing substrates. After the estimation of the model parameters of each enzyme independently by progress curve analyses, the full process model of a simple batch-process was established by coupling rate equations and mass balances. Validation experiments of the one-pot multienzymatic batch process revealed high prediction accuracy of the process model and a model analysis offered important insight to the identification of optimum reaction conditions.

3 beta-Hydroxysteroid dehydrogenase of Ruminococcus sp. from human intestinal bacteria.

Ruminococcus sp. PO1-3 obtained from human intestinal flora is able to reduce dehydrocholate as well as 3-ketoglycyrrhetinate. From this bacterium dehydrocholate- and 3-ketoglycyrrhetinate-reducing activities were purified one thousand-fold together with 3-ketocholanate-reducing and 3-beta-hydroxyglycyrrhetinate (glycyrrhetic acid) oxidizing activities by means of Matrex Red A, Sephadex G-200 and Octyl-Sepharose column chromatography. The purified enzyme catalyzed the reduction of dehydrocholic acid to 3 beta-hydroxy-7,12-diketocholanic acid and of 3-ketocholanic acid to 3 beta-hydroxycholanic acid. Studies on substrate specificity revealed that the enzyme had absolute specificity for the beta-configuration of a hydroxyl group at the 3 position of bile acid and steroids having no double bond in the A/B ring. This enzyme was neither beta-hydroxysteroid dehydrogenase [EC 1.1.1.51] nor 3 beta-hydroxy-delta 5-steroid dehydrogenase [EC 1.1.1.145], but a novel type of enzyme, defined as 3 beta-hydroxysteroid dehydrogenase.

Identification and characterization of dehydrocholic acid reductase system in the cytosol of human red blood cells.

We conducted in vivo and in vitro studies of the reductive metabolism of the cholagogue, dehydrocholic acid (DHCA). Immediately after the intravenous administration of 1 g of DHCA in normal subjects (n = 6), the concentration of the reductive metabolite, 3 alpha-hydroxy-7,12-dioxo-cholanoic acid (unconjugated form), increased sharply in the systemic circulation, rising to 95.8 microM 10 min after administration. The results of in vitro experiments with DHCA and whole blood showed that 3 alpha-hydroxy-7,12-dioxocholanoic acid were produced from DHCA. In vitro experiments using DHCA and the red blood cell fraction, and DHCA and the red blood cell cytoplasmic fraction gave similar results to those described above with whole blood. However, a reductive metabolite was not formed by the incubation of DHCA and the red blood cell membrane fraction. These findings indicated that, contrary to the conventional theory that intravenously administered DHCA is subjected to reductive metabolism only in the liver, reduction also occurs in the systemic circulation, and the mechanism for this reductive metabolism is present in the cytoplasmic fraction of red blood cells. Further investigation to characterize this reductive metabolic system revealed an optimum temperature of 37 degrees C, an optimum pH of 7.4, a Km value of 2.0 x 10(-3) M, and inactivation by heat treatment (70 degrees C for 2 min).

A novel enzyme system for the reduction of 3-oxo bile acids in human red blood cells.

7 alpha,12 alpha-Dihydroxy-3-oxo- and 3,7,12-trioxo-5 beta-cholanoic acids labeled with 18O atoms were incubated with human red blood cells, and the biotransformation products were separated and characterized by gas chromatography-mass spectrometry as the pentafluorobenzyl ester-trimethylsilyl and -dimethylethylsilyl ether derivatives with the negative ion chemical ionization mode. The reduced products, 3 beta,7 alpha,12 alpha-trihydroxy-5 beta-cholanoic acid for the former, and 3 alpha-hydroxylated dioxo bile acid together with 3 beta-hydroxylated 7,12-dioxo-5 beta-cholanoic acid for the latter, were identified as metabolites. When 3-oxo bile acid was incubated with human blood denatured at 70 degrees C for 2 min, no metabolites were formed. The enzymic reduction activity has been localized in the red blood cell fraction.

Effect of dehydrocholic, chenodeoxycholic, and taurocholic acids on the excretion of bilirubin.

The effects of IV bile acid infusion (at approx 20% of normal excretion rate) on the biliary excretion of 3-alpha-hydroxy bile acids and bilirubin were investigated in ponies prepared surgically with chronic external biliary fistulas. Endogenous bile acid excretion (approx 45 mumol/min) decreased to the hepatic synthesis rate (approx 1.5 mumol/min) during the initial 4 to 5 hours of bile drainage. In type 1 studies, both chenodeoxycholic and taurocholic acid infusion (8 to 9 mumol/min) increased bilirubin excretion by 58% to 82% following 5 hours of biliary diversion. During type 2 studies, 3-hour IV infusions (10.5 mumol/mon) of dehydrocholic acid, 4 hours following biliary diversion, increased bile flow by 45% to 62% and excretion of 3-alpha-hydroxy bile acid by 34% to 36% above preinfusion (hepatic synthesis) levels. Bilirubin excretion was not significantly changed during those increases in bile flow and bile acid excretion. Immediately after dehydrocholic acid infusion, taurocholic acid infusion (8.1 mumol/min) greatly increased bilirubin excretion for 1 hour (a reversal of hepatic storage identical to that found during type 1 studies), prolonged excretion (mg/2 hours) being two to three times that caused by dehydrocholic acid infusion. Bilirubin excretion appeared to correlate with the micelle-forming capacity of endogenous bile acids as opposed to the nonmicelle-forming characteristic of synthestic dehydrocholic acid.

The secretory characteristics of dehydrocholate in the dog: comparison with the natural bile salts.

1. During dehydrocholate administration in the taurine replete dog, the maximum excretory rate of total bile salt (almost entirely dehydrocholate derivative, mostly conjugated) was 3-84 +/- 0-53 (S.D.) mumole/min. kg body wt. (eleven experiments). This was much less than the excretory maximum previously obtained for taurocholate (8-64 +/- 1-31 (S.D.) mumole/min. kg total cholate, mostly conjugated). 2. The superimposition of taurocholate infusion did not cause any significant change in the 'dehydrocholate' maximum but taurocholate itself was excreted into bile at no more than about half its normal maximum. When taurocholate maximum excretion was established first, it was reduced by dehydrocholate administration. In both types of experiment the joint bile salt excretory maximum was of the same order as that of taurocholate alone, provided taurocholate made up at least 40-50% of the total bile salt. 3. When taurocholate administration was stopped, the maximum excretory rate of 'dehydrocholate' rose to values up to 63% above the initially determined excretory maximum; the enhanced 'dehydrocholate' excretory maximum, when calculated for optimal conditions, approached that of actively conjugated vholate, even though the effective 'dehydrocholate' concentration in bile was ten to twenty times the critical micellar concentration of taurocholate. This suggests that the effective bile salt concentration in bile is not an important determinant of the secretory performance of a bile salt. 4. To explain findings (2) and (3) it is necessary to postulate that taurocholate has both a facilitatory and an inhibitory action on 'dehydrocholate' excretion. The facilitatory action, which persists after taurocholate has left the animal, may consist either of an increase in the maximum rate at which modification of dehydrocholate takes place within the liver cell, or an increase in the number of functioning 'carriers' for 'dehydrocholate' transfer. The data suggest that the inhibitory effect is due to the competitive interaction that also appears to exist between the two bile salts. 5. The increase in bile flow rate per unit increase in 'dehydrocholate' excretion (15 ml./m-mole) was about twice that obtained for taurocholate. There was no significant formation of micellar aggregates during 'dehydrocholate' excretion, as judged from the total electrolyte concentration of bile and its osmalality. 6. During the excretion of 'dehydrocholate'-taurocholate mixtures (approximately 1:1) at submaximal rates the associated bile flow rate was not less than the sum of the separate components, thus suggesting that 'dehydrocholate' was not being incorporated in taurocholate mixed micelles.

Cholic acid binding to isolated rat liver plasma membranes.

Cholic acid binding to isolated rat liver plasma membranes was studied using a centrifugal filtration technique which allowed independent determination of free and membrane-bound cholic acid. Binding of cholic acid was very rapid and reversible. Scatchard analysis revealed at least three binding sites with high, medium and low affinity. The high affinity binding a) displayed saturability and isotope replacement, b) was not present in rat liver mitochondria and red blood cell ghosts and c) was temperature dependent. This binding has a very low capacity with a dissociation constant in the physiological range of plasma cholic acid concentration and has an affinity for other common bile acids. Cholic acid binding to the high affinity binding site was not inhibited by estrone, beta-estradiol or cholesterol. These results would suggest that the high affinity binding site represents a specific binding site for cholic acid and may also be specific for other common bile acids. This binding was not dependent on Na and was inhibited by bromosulfophthalein. Cholic acid binding to the high affinity site has some features in common with cholic acid uptake by isolated rat hepatocytes, and this would suggest that the high affinity binding site could be the postulated carrier for hepatic uptake of cholic acid.