Structure–Activity Studies of Brassinosteroids and the Search for Novel Analogues and Mimetics with Improved Bioactivity

A number of novel brassinosteroid analogues were synthesized and subjected to the rice leaf lamina inclination bioassay. Modified B-ring analogues included lactam, thiolactone, cyclic ether, ketone, hydroxyl, and exocyclic methylene derivatives of brassinolide. Those derivatives containing polar functional groups retained considerable bioactivity, whereas the exocyclic methylene compounds were devoid of activity. Analogues containing normal alkyl and cycloalkyl substituents at C-24 (in place of the isopropyl group of brassinolide) showed an inverse relationship between activity and chain length or ring size, respectively. The corresponding cyclopropyl and cyclobutyl derivatives were significantly more active than brassinolide and appear to be the most potent brassinosteroids reported to date. When synergized with the auxin indole-3-acetic acid (IAA), their bioactivity can be further enhanced by 1–2 orders of magnitude. The cyclopropyl derivative, when coapplied with the auxin naphthaleneacetic acid, gave a significant increase in yield of wheat in a field trial. Certain 25- and 26-hydroxy derivatives are known metabolites of brassinosteroids. All of the C-25 stereoisomers of 25-hydroxy, 26-hydroxy, and 25,26-dihydroxy derivatives of brassinolide were prepared and shown to be much less active than brassinolide. This indicates that they are likely metabolic deactivation products of the parent phytohormone. A series of methyl ethers of brassinolide was synthesized to block deactivation by glucosylation of the free hydroxyl groups. The most significant finding was that the compound where three of the four hydroxyl groups (at C-3, C-22, and C-23) had been converted to methyl ethers retained substantial bioactivity. This type of modification could, in theory, allow brassinolide or 24-epibrassinolide to resist deactivation and thus offer greater persistence in field applications. A series of nonsteroidal mimetics of brassinolide was designed and synthesized. Two of the mimetics showed significant bioactivity and one had bioactivity comparable to brassinolide, but only when formulated and coapplied with IAA. They thus represent the first nonsteroidal analogues possessing brassinosteroid activity.     

Configuration at C-25 in 5 beta-cholestane-3 alpha,7 alpha,12 alpha,25,26-pentol excreted by patients with cerebrotendinous xanthomatosis: circular dichroism and 13C-NMR studies

The configuration at C-25 in 5 beta-cholestane-3 alpha, 7 alpha, 12 alpha,25,26-pentol isolated from the bile and feces of patients with cerebrotendinous xanthomtosis (CTX) was determined from the lanthanide-induced circular dichroism (CD) Cotton effects and 13C-NMR measurements. Under anhydrous conditions, CD spectra of 5 beta-cholestane-3 alpha, 7 alpha, 12 alpha,25,26-pentol in the presence of Eu(fod)3 exhibited a large induced negative Cotton effect at 320 nm. On the basis of the empirical rule (primary-tertiary-alpha-diols) in which R compounds have positive Cotton effects and S compounds have negative Cotton effects at 320 nm, it was concluded that 25,26-pentol has the 1,2,glycol structure with C-25 having the S-configuration. This assignment was based upon comparison with model compounds, 25(R and S),26-dihydroxy cholesterols and 25(R and S),26-dihydroxy cholecalciferols whose single-crystal X-ray structure and 13C-NMR studies have been performed. It is suggested that these data may be helpful to clarify the stereospecificity of the hydroxylation of the terminal methyl group of the cholesterol side chain in CTX.     

Synthesis and bioactivity of C-2 and C-3 methyl ether derivatives of brassinolide

The following six novel methyl ether derivatives of brassinolide were prepared and their brassinosteroid activity was measured by means of the rice leaf lamina inclination bioassay: 2-O-methylbrassinolide, 3-O-methylbrassinolide, 2,22,23-tri-O-methylbrassinolide, 3,22,23-tri-O-methylbrassinolide, 2-O-methyl-25-methoxybrassinolide and 3-O-methyl-25-methoxybrassinolide. Brassinolide was used as a standard for comparison. All six compounds were also tested in the presence of 1000 ng of indole-3-acetic acid (IAA), an auxin that synergizes the effects of brassinosteroids. The 2-O-methyl- and 3-O-methylbrassinolide derivatives showed weak activity at high doses, which was enhanced by IAA, especially in the case of the 3-O-methyl derivative. Similarly, the 2,22,23-tri-O-methyl- and 3,22,23-tri-O-methyl derivatives displayed weak bioactivity on their own, but significantly stronger activity when applied with IAA. The 3-O-methyl and 3,22,23-tri-O-methyl analogues plus IAA were comparable in bioacivity to brassinolide alone, but were less active than brassinolide plus IAA. In each case, O-methylation at C-2 resulted in a greater loss of activity than O-methylation at C-3 under the same conditions. The relatively strong activity of 3,22,23-tri-O-methylbrassinolide in the presence of IAA is especially noteworthy as it indicates that free hydroxyl groups at C-3, C-22, and C-23 are not essential for bioactivity. Finally, 2-O-methyl- and 3-O-methyl-25-methoxybrassinolide were essentially inactive alone, and showed only a modest increase in bioactivity when coapplied with IAA.     

C-25 epimeric 26-haloponasterone A: synthesis, absolute configuration and moulting activity

A convenient synthesis of inokosterone has been accomplished. Inokosterone exists as two C-25 epimers, which could be separated from each other through their diacetonide derivatives. The absolute configuration of these compounds was determined. Two C-25 epimers of 26-chloroponasterone A were synthesized from the respective C-25 epimeric inokosterone. Two epimeric 26-bromo and 26-iodoponasterone A compounds were also synthesized. Moulting activity of these compounds was evaluated using the Musca bioassay, and it was found that the (25S)-26-halo analogues were more active than the corresponding (25R)-26-halo analogues. Among the 25S series, an increase in activity with an increase in size of the halogen atom was observed, indicating that the steric factor was more important than the electronic factor in binding of these ecdysteroid analogues to the receptor. On the other hand, a decrease in activity with an increase in size of the halogen atom was noted in the 25R series, suggesting that the steric factor was less important than the electronic factor. The results indicated that the configuration at C-25 and the substituent at C-26 have significant influences on the interaction of ecdysteroids with their receptor.     

Stereoselective synthesis of (24R and 24S) 5 beta-cholestane-3 alpha,7 alpha,12 alpha,24,25-pentols and (25R and 25S) 5 beta-cholestane-3 alpha,7 alpha,12 alpha,25,26-pentols using a modified osmium-catalyzed Sharpless asymmetric dihydroxylation process.

Described herein are the stereoselective syntheses of the (24R, 24S) and (25R, 25S) isomers of 5 beta-cholestane-3 alpha,7 alpha,12 alpha,24,25-pentols and 5 beta-cholestane-3 alpha,7 alpha,12 alpha,25,26-pentols by using a modified osmium-catalyzed Sharpless asymmetric dihydroxylation process. Also presented herein are the results of lanthanide-induced CD Cotton effect measurements and 1H- and 13C-nuclear magnetic resonance studies of (24R, 24S) and (25R, 25S)-5 beta-cholestanepentols and their derivatives. These compounds were required to study the biosynthesis of cholic acid from cholesterol.     

Brassinolide and [26, 28-2H6]brassinolide are differently demethylated by loss of C-26 and C-28, respectively, in Marchantia polymorpha

Metabolism of brassinolide in Marchantia polymorpha was investigated by use of in vivo suspension cultured cells. GC-MS analysis of metabolites derived from non-labelled brassinolide and [26, 28-2H6] brassinolide revealed that brassinolide was converted to 26-norbrassinolide while [26, 28-2H6]brassinolide to [26-2H3]28-norbrassinolide. It seems that Marchantia cells recognized [26, 28-2H6]brassinolide as a xenobiotic rather than brassinolide and deteriums attached to C-28 significantly affect demethylation reaction due to isotopic effect. Thus, demethylation of brassinolide in planta seems to proceed by loss of C-26 rather than C-28. The present finding is the first evidence for demethylation metabolism of brassinosteroids. The biological activity of 26-norbrassinolide was 10-fold reduced as shown by the rice lamina inclination test. However, because of its high biological activity, it remains difficult to conclude whether or not C-26 demethylation serves as an important deactivation process of brassinolide.     

Synthesis and structure of 26 (or 27)-nor-5 beta-cholestane-3 alpha,7 alpha,12 alpha,24S,25 xi-pentol isolated from the urine and feces of a patient with sitosterolemia and xanthomatosis

The urine and feces of a patient with the rare inherited lipid storage disease, sitosterolemia and xanthomatosis, were analyzed. Substantial quantities of C26-bile alcohol, 26 (or 27)-nor-5 beta-cholestane-3 alpha,7 alpha,12 alpha,24S,25 xi-pentol along with 5 beta-cholestane-3 alpha,7 alpha,12 alpha,24-tetrol, 5 beta-cholestane-3 alpha,7 alpha,12 alpha,25-tetrol, 5 beta-cholestane-3 alpha,7 alpha,12 alpha,24R,25-pentol, and 5 beta-cholestane-3 alpha,7 alpha,12 alpha,25,26-pentol were found. The structure of the C26-bile alcohol was confirmed by direct comparison (gas-liquid chromatography-mass spectrometry and thin-layer chromatography) with a standard sample synthesized from cholic acid. The configurational assignment at C-24 was determined by lanthanide-induced circular dichroism Cotton effect measurements. The increased excretion of these C26- and C27-bile alcohols suggests an abnormality of bile acid biosynthesis in this disease.     

Convergent synthesis, chiral HPLC, and vitamin D receptor affinity of analogs of 1,25-dihydroxycholecalciferol

A series of analogs of 1,25-dihydroxycholecalciferol was obtained with an additional chiral center at the terminus of the aliphatic side chain (C-25). The analogs were obtained from (+)-(R)- and (-)-(S)-2-methylglycidols, by opening of the oxirane ring with the carbanions derived from vitamin D C23a,24- or C22-sulfones. The diastereomeric purity of the analogs was determined by high-performance liquid chromatography on a chiral stationary phase. The binding affinity of analogs for the calf thymus intracellular vitamin D receptor (VDR) was two orders of magnitude lower than that of the lead compound of this group, 24a,24b-dihomo-1,25-dihydroxycholecalciferol, and it was comparable to the affinity of analogs of 24-nor-1,25-dihydroxycholecalciferol. However, a twofold difference was observed for analogs diastereomeric at C-25 in their affinity for VDR. The diastereodifferentiation of the binding affinity was found to be specific for vitamin D vicinal 25,26-diols as it disappears for analogs where 26-hydroxyl, neighboring the C-25 chiral center, is replaced with methyl.     

Isolation and characterization of calmodulin-inactivating cholesterol hydroperoxides.

A series of cholesterol hydroperoxides has been prepared and tested as inactivators of calmodulin. Two previously undescribed compounds, tentatively identified as 20-(R)-25-dihydroperoxy-5-cholesten-3 beta-ol and its 20-(S) isomer inactivate calmodulin with 50% loss of activity at 5-10 microM. Cholesterol derivatives with a single hydroperoxy group at C-20 or C-25 are less effective, while 7 alpha-hydroperoxy-cholesterol and 25-hydroxy-cholesterol are inactive. The side-chain hydroperoxide compounds were isolated from a mixture shown earlier to suppress formation of fatty streaks in aortas of rabbits fed a diet supplemented with cholesterol.     

Synthesis and biological activity of 26-norbrassinolide, 26-norcastasterone and 26-nor-6-deoxocastasterone

26-Norbrassinolide, identified as a metabolite of brassinolide in cultured cells of the liverwort, Marchantia polymorpha, as well as 26-norcastasterone and 26-nor-6-deoxocastasterone were synthesized. Synthesis of these new brassinosteroids was conducted by employing the orthoester Claisen rearrangement and asymmetric dihydroxylation as key reactions. The modified rice lamina inclination test indicated that these three 26-norbrassinosteroids were less active than their corresponding C28 brassinosteroids. Growth-promoting activities were also examined by using the brassinosteroid-deficient, dwarf mutant lkb of garden pea (Pisum sativum L.). In this assay, 26-norbrassinolide was as effective as brassinolide and 26-norcastasterone was more effective than castasterone although 26-nor-6-deoxocastasterone was much less effective than 6-deoxocastasterone. Therefore, removal of C-26 of brassinosteroids does not necessarily reduce the biological activity. The role of C-26 removal in Marchantia cells remains unclear.     

Identification of 27-nor-5 beta-cholestane-3 alpha,7 alpha,12 alpha,24 xi, 25 xi,26-hexol and partial characterization of the bile alcohol profile in urine

The nature of the bile alcohols present in urine of an infant with neonatal cholestasis has been investigated. Urine was extracted with Sep-Pak C18 cartridges and a glucuronide fraction was isolated by ion exchange chromatography on Lipidex-DEAP. Following enzymatic hydrolysis and purification on Lipidex-DEAP, the bile alcohols were isolated by high performance liquid chromatography. Fourteen compounds were studied by a combination of microchemical reactions and capillary column gas-liquid chromatography-mass spectrometry. Both C26 and C27 bile alcohols were present. Among the former, three additional isomers of the previously identified 27-nor-5 beta-cholestane-3 alpha,7 alpha,12 alpha,24 xi,25 xi-pentol were detected. A new C26 bile alcohol, 27-nor-5 beta-cholestane-3 alpha,7 alpha,12 alpha,24 xi,25 xi,26 -hexol, was identified, and a 27-norcholestane-pentolone with hydroxyl groups at C-24 and C-25 and a keto group in the ring system was partially characterized. The C27 bile alcohols consisted of cholestanepentols, -tetrolones, and -pentolones. 5 beta-Cholestane-3 alpha,7 alpha,12 alpha,25,26-pentol (5 beta-bufol), one of its isomers and an isomer of cholestane-3,7,12,24,26-pentol were present. Two cholestanetetrolones and two cholestanepentolones having the keto group in the ring system were partially characterized. The hydroxyl groups in the side chain of the tetrolones were at C-24,26 and C-25,26, respectively, whereas the pentolones had hydroxyl groups at C-24,25 and C-25,26, respectively. The excretion of glucuronidated bile alcohols in urine is suggested to reflect an alternative metabolism of intermediates in the normal biosynthesis of bile acids.     

Synthesis and biological activity of (22E,25R)- and (22,25S)-22-dehydro- 1 alpha,25-dihydroxy-26-methylvitamin D3

Both 25-epimers of (22E)-22-dehydro-1 alpha,25-dihydroxy-26-methylvitamin D3 [22-dehydro-26-methyl-1,25-(OH)2D3] were synthesized. The biological activity of these compounds was tested in binding affinity to chick intestinal receptor protein of 1 alpha,25-dihydroxy-vitamin D3 [1,25-(OH)2D3] and in stimulating for intestinal calcium transport and bone calcium mobilization with vitamin D-deficient rats. The relative potency of (25R)- and (25S)-22-dehydro-26-homo-1,25-(OH)2D3 and 1,25-(OH)2D3 in competing for the intestinal cytosolic binding was 1.7:1.5:1. A similar order of activity was observed on intestinal calcium transport and bone calcium mobilization. In the ability for stimulation of intestinal calcium transport, (25R)- and (25S)-22-dehydro-26-methyl-1,25-(OH)2D3 were about 3.6 and 2.1 times as active as 1,25-(OH)2D3, respectively. In bone calcium mobilization tests, (25R)- and (25S)-22-dehydro-26-methyl-1,25-(OH)2D3 were estimated to be 2.2 and 1.6 times as potent as 1,25-(OH)2D3, respectively.     

Synthesis and biological activity of 25,26-dihydroxycholecalciferol

25,26-Dihydroxycholecalciferol (25,26-dihydroxyvitamin D₃), a metabolite of vitamin D3 preferentially active on intestine has been synthesized. This compound was prepared by converting 3β-hydroxy-27-norcholest-5-en-25-one to 25,26-epoxy-5-cholesten-3β-o1 and base-catalyzed hydrolysis of the latter to 5-cholestene-3β,25,26-triol; allylic bromination of the corresponding triacetate, and dehydrobromination gave the required 5,7-diene which yielded the vitamin derivative upon photolysis (Figure 3). The synthetic product shows the same activity pattern as the natural metabolite: at dose levels of 0.25 μg, the compound stimulates intestinal calcium transport, but has no effect on bone calcium mobilization in rats maintained on a vitamin D-deficient, low calcium didt. Higher doses (2.5 μg) elicit a more pronounced intestinal calcium transport response, but also have no significant effect on the bone mobilization system. The compound exhibits no biologial activity in nephrectomized animals.     

24,25-Epoxysterol metabolism in cultured mammalian cells and repression of 3-hydroxy-3-methylglutaryl-CoA reductase

In view of the potential importance of 24,25-epoxysterols as intracellular regulators of 3-hydroxy-3-methylglutaryl-CoA reductase, the C-24 epimers of 24,25-oxidolanosterol and 24,25-epoxycholesterol were tested for their biological activity and metabolism in cell cultures. All four compounds produced repression of the reductase in cultured mouse fibroblasts (L cells), and both 24(S)- and 24(R),25-epoxycholesterol exhibited high affinity binding to the cytosolic oxysterol-binding protein. However, binding of the epimeric 24,25-oxidolanosterols was not detected. 24(S),25-Epoxycholesterol was not rapidly metabolized in either L cells or Chinese hamster lung (Dede) cells. 24(S),25-Oxidolanosterol was rapidly converted to 24(S),25-epoxycholesterol in both cell lines. 24(R),25-Oxidolanosterol was converted to 24(R)-hydroxycholesterol in Dede cells, but was converted instead to 24(R),25-epoxycholesterol in L cells, which lack sterol delta 24-reductase activity. Although 24(S),25-oxidolanosterol does not appear to accumulate in these cell cultures, it was found in human liver in about one-fifth the amount of 24(S),25-epoxycholesterol. 24(R),25-Epoxycholesterol was also converted to 24(R)-hydroxycholesterol in Dede cells, but not in L cells. Triparanol inhibited the reduction of the 24(R),25-epoxides in Dede cells, consistent with the idea that this reaction is catalyzed by the delta 24-reductase. 24(R)-Hydroxycholesterol and its 24(S) epimer exhibited affinity for the binding protein and repressed 3-hydroxy-3-methylglutaryl-CoA reductase.     

Synthesis, biologic activity, and protein binding characteristics of a new vitamin D analog, 22-hydroxyvitamin D3

We synthesized a novel vitamin D analog, 22-hydroxyvitamin D3 9 and tested its biologic activity (and antivitamin properties) in vivo in vitamin D-deficient rats, and in vitro in the chick embryonic duodenum. We examined its ability to bind to the sterol carrier protein, vitamin D binding protein and the chick intestinal cytosol receptor for 1,25-dihydroxyvitamin D3. The new vitamin 9 was synthesized from 3 beta-hydroxy-22,23-dinorcholenic acid 1 in 12 steps. The vitamin 9 displayed no vitamin D agonist activity in the intestine or in bone in vivo and did not block the activity of vitamin D3 or 25-hydroxyvitamin D3. It was a weak vitamin D3 agonist in the chick embryonal duodenum in vitro. It did not antagonize the activity of 1,25-dihydroxyvitamin D3. Vitamin 9 bound to the chick intestinal cytosol receptor with low affinity. 22-Hydroxyvitamin D3 and various vitamin D sterols were bound to vitamin D binding protein in the following order: 25-hydroxyvitamin D3. (24R)-24,25-dihydroxyvitamin D3, and (25S)-25,26-dihydroxyvitamin D3 greater than 22-hydroxyvitamin D3 greater than 11 alpha-hydroxyvitamin D3 greater than 1,25-dihydroxyvitamin D3 greater than vitamin D3. We conclude that the introduction of a hydroxyl group at C-22 in the side chain of the vitamin D3 molecule decreases its biological activity.     

Cytotoxic triterpenoid saponins from the fruits of Aesculus pavia L

Continued chemical investigation on the fruits of North American Aesculus pavia L. resulted in the isolation and identification of 13 polyhydroxyoleanene pentacyclic triterpenoid saponins, named aesculiosides IIe-IIk (1-7), and IIIa-IIIf (8-13), together with 18 known compounds: aesculiosides Ia-Ie (14-18), IIa-IId (19-22), IVa-IVc (23-25), 3-O-[beta-D-galactopyranosyl(1-->2)]-alpha-L-arabinofuranosyl(1-->3)-beta-D-glucuronopyranosyl-21,22-O-diangeloyl-3beta,15 alpha,16 alpha,21 beta,22 alpha,28-hexahydroxyolean-12-ene (26), 3-O-[beta-D-glucopyranosyl(1-->2)]-alpha-L-arabinofuranosyl(1-->3)-beta-D-glucuronopyranosyl-21,22-O-diangeloyl-3beta,16 alpha,21 beta,22 alpha,24 beta,28-hexahydroxyolean-12-ene (27), 3-O-[beta-D-galactopyranosyl(1-->2)]-alpha-L-arabinofuranosyl(1-->3)-beta-D-glucuronopyranosyl-21,22-O-diangeloyl-3beta,16 alpha,21 beta,22 alpha,28-pentahydroxyolean-12-ene (28), R(1)-barrigenol (29), scopolin (30), and 5-methoxyscopolin (31). The structures of these compounds were elucidated by spectroscopic and chemical analyses. Compounds 14-22 and 26-28 were tested in vitro for their activity against 59 cell lines from nine different human cancers including leukemia, non-small cell lung, colon, CNS, melanoma, ovarian, renal, prostate, and breast. It was found that compounds with two-acyl groups at C-21 and C-22 had cytotoxic activity for all cell lines tested with GI(50) 0.175-8.71 microM, while compounds without acyl groups at C-21 and C-22 had weak or no cytotoxic activity. These results suggest that the acyl groups at C-21 and C-22 are essential for their activity.     

Stereospecific formation of (24E)-3 alpha,7 alpha,12 alpha-trihydroxy-5 beta-cholest-24-en-26-oic acid and (24R,25S)-3 alpha,7 alpha,12 alpha,24-tetrahydroxy-5 beta-cholestan-26-oic acid from either (25R)- or (25S)-3 alpha,7 alpha,12 alpha-trihydroxy-5 beta-cholestan-26-oic acid by rat liver homogenate

Studies of the stereochemistry of the intermediates, 3 alpha,7 alpha,12 alpha-trihydroxy-5 beta-cholest-24-en-26-oic acid and 3 alpha,7 alpha,12 alpha,24-tetrahydroxy-5 beta-cholestan-26-oic acid, in the biosynthetic sequence between 3 alpha,7 alpha,12 alpha-trihydroxy-5 beta-cholestan-26-oic acid and cholic acid have been undertaken. (25R)- or (25S)-3 alpha,7 alpha, 12 alpha-Trihydroxy-5 beta-cholestan-26-oic acid was incubated with rat liver homogenates. The reaction products were converted to p-bromophenacyl ester derivatives and the esters were analyzed by high-performance liquid chromatography. By comparison with authentic samples of two (24E)- and (24Z)-isomers of the alpha, beta-unsaturated acid and of four isomers at C-24 and C-25 of the beta-hydroxy acid, (24E)-3 alpha,7 alpha, 12 alpha-trihydroxy-5 beta-cholestan-26-oic acid and (24R,25S)-3 alpha,7 alpha,12 alpha,24-tetrahydroxy-5 beta-cholestan-26-oic acid were found to be formed from either (25R)- or (25S)-3 alpha,7 alpha, 12 alpha-trihydroxy-5 beta-cholestan-26-oic acid. No formation of the (24Z)-isomer of the trihydroxycholestenoic acid or the other three isomers of the tetrahydroxycholestanoic acid was detected. The findings are discussed in relation to the assumed pathway for side chain cleavage in cholic acid biosynthesis.     

The synthesis and biological activity of 25,26-dihydroxy-cholecalciferol, a polar metabolite of vitamin D3

The synthesis of 25,26-dihydroxycholecalciferol, a biologically active metabolite of cholecalciferol (vitamin D₃) is described. 3β-Hydroxy-27-nor-5-cholesten-25-one was converted in three steps to 5,7-cholestadiene-3β,25 (RS), 26-triol. The latter compound was irradiated with ultraviolet light to give 25 (RS), 26-dihydroxyprecholecalciferol; this compound underwent thermal isomerisation to yield 25 (RS), 26-dihydroxycholecalciferol. The structure of the final product was confirmed by ultra-violet spectroscopy, mass spectroscopy and by periodate degradation to the known 25-oxo-27-nor-cholecalciferol. 25 (RS), 26-Dihydroxycholecalciferol was able to stimulate the intestinal absorption of calcium but had little or no effect on the healing of rickets.     

NMR assignment of the absolute configuration of C-25 in furostanol steroidal saponins

The chemical shifts of the geminal proton resonances of H(2)-26 (δa and δb) are a widely used predictor of C-25 stereochemistry in furostanol steroidal saponins, being in general more resolved in 25S than 25R compounds. Unexpectedly, we found that application of this empirical rule in different solvents led to conflicting assignments of stereochemistry. An experimental survey revealed that, while the chemical shifts of H(2)-26 exhibit a dependence on C-25 configuration, it is less pronounced in methanol-d4 than pyridine-d5 solvent, and thus the general rule derived for pyridine-d5 fails when NMR spectra are acquired in methanol-d4. We propose a modified empirical method for the direct assignment of C-25 stereochemistry in furostanol saponins in methanol-d4 (Δ(ab)=0.45-0.48 ppm for 25S; Δ(ab)=0.33-0.35 ppm for 25R), and provide several detailed examples. In addition, the absolute configuration of compound 8, a steroidal saponin isolated in previous work from Ruscus colchicus, is corrected from 25R to 25S stereochemistry.     

Dual metabolic pathway of 25-hydroxyvitamin D3 catalyzed by human CYP24.

Human 25-hydroxyvitamin D3 (25(OH)D3) 24-hydroxylase (CYP24) cDNA was expressed in Escherichia coli, and its enzymatic and spectral properties were revealed. The reconstituted system containing the membrane fraction prepared from recombinant E. coli cells, adrenodoxin and adrenodoxin reductase was examined for the metabolism of 25(OH)D3, 1alpha,25(OH)2D3 and their related compounds. Human CYP24 demonstrated a remarkable metabolism consisting of both C-23 and C-24 hydroxylation pathways towards both 25(OH)D3 and 1alpha,25(OH)2D3, whereas rat CYP24 showed almost no C-23 hydroxylation pathway [Sakaki, T. Sawada, N. Nonaka, Y. Ohyama, Y. & Inouye, K. (1999) Eur. J. Biochem. 262, 43-48]. HPLC analysis and mass spectrometric analysis revealed that human CYP24 catalyzed all the steps of the C-23 hydroxylation pathway from 25(OH)D3 via 23S, 25(OH)2D3, 23S,25,26(OH)3D3 and 25(OH)D3-26,23-lactol to 25(OH)D3-26, 23-lactone in addition to the C-24 hydroxylation pathway from 25(OH)D3 via 24R,25(OH)2D3, 24-oxo-25(OH)D3, 24-oxo-23S,25(OH)2D3 to 24,25,26,27-tetranor-23(OH)D3. On 1alpha,25(OH)2D3 metabolism, similar results were observed. These results strongly suggest that the single enzyme human CYP24 is greatly responsible for the metabolism of both 25(OH)D3 and 1alpha,25(OH)2D3. We also succeeded in the coexpression of CYP24, adrenodoxin and NADPH-adrenodoxin reductase in E. coli. Addition of 25(OH)D3 to the recombinant E. coli cell culture yielded most of the metabolites in both the C-23 and C-24 hydroxylation pathways. Thus, the E. coli expression system for human CYP24 appears quite useful in predicting the metabolism of vitamin D analogs used as drugs.