Study On The Synthesis Of A Natural Antidepressant Flavonoid Glycoside And Its Analogues

Quercetin 3-O-β-D-apiofuranosyl-(1→2)-[α-L-rhamnopyranosyl-(1→6)]-β-D- glucopyranoside (1) is a flavonol O-glycoside isolated from the seeds of glandless cotton (a variant of cotton plant resulting from selective breeding) in our laboratory. It showed notable antidepressant effects in forced swimming experiments on mice. It also exhibited the ability to promote the proliferation of hippocampal cells in chronically stressed mice in vivo and cultured hippocampal neural progenitor cells in vitro. Moreover, its intact form could be detected in brain dialysate samples from rats after oral administration. It seems to be a potential antidepressant. Since the content of 1 in original plant is limited, it is difficult to isolate it in large scale. In order to provide sufficient material for further pharmacological studies, it is necessary to develop an effective synthetic method for preparing this compound.The structure of compound 1 is characterized by the unique 2,6-branched trisaccharide moiety: aβ-D-glucopyranose attached to the 3-hydroxyl group of the aglycone, a rareβ-D-apiofuranose and anα-L-rhamnopyranose as the terminal sugars linked to the 2-hydroxyl and 6-hydroxyl groups of the glucose, respectively. The aglycone and the three monosaccharides have multiple hydroxyl groups of similar chemical reactivity. Moreover, two conformational isomers (α- orβ-) can be formed in glycosylation reactions. Thus, the most challenging problem in the synthesis of 1 is how to create glycosidic bonds with desired conformation specifically at the 3-OH of quercetin and the 2-OH and 6-OH of the glucose fragment. Another difficulty in the synthesis of 1 arises from the rare apiose that is commercially unavailable and should be especially prepared.Creation of a flavonol 3-O-glycosyl linkage is commonly carried out by reaction of the aglycone with glycosyl bromides under phase-transfer catalysis (PTC) or the action of silver salts. These methods are not applicable when the glycosyl donor is a bulky trisaccharide. Thus, a stepwise glycosylation approach will be a favorable choice for the synthesis of compound 1. In view of this, the centralβ-D-glucopyranose should be linked to the 3-OH of quercetin first, and suitably protected to avoid tedious protection-deprotection steps in the process of constructing the 2,6-branched trisaccharide moiety. Such steps often lead to low yield of the target product. The free hydroxyl groups on a D-glucopyranose moiety were known to have a reactivity order of 6-OH>> 3-OH >2-OH >4-OH. Previous studies gave examples of selectively glycosylating the 6-OH in a glucose moiety in the presence of an unprotected 2-OH or glycosylating the 2-OH without protecting the 4-OH. These results suggested that it might be possible to sequentially glycosylate these three hydroxyl groups by just protecting the 3-OH. By this means, the protective manipulation in construction of the branched trisaccharide moiety will be substantially simplified.Based on the above analysis, four building blocks were designed by disconnecting the glycosidic bonds of compound 1 and synthesized separately. Connecting the four building blocks by stepwise glycosylation afforded the target compound. The synthetic work was carried out as follows: (1) Commercially available rutin was converted into the suitably protected aglycone, i.e., 7,3′,4′-tri-O-benzylquercetin 5, by benzylation and subsequent acid hydrolysis in excellent yield. (2) D-glucose was converted into 2,4,6-tri-O-acetyl-3-O-benzoyl-α-D-glucopyranosyl bromide 6 in 5 steps. (3) The apiofuranosyl donor 2,3,5-tri-O-benzoyl-α/β-D-apiofuranosyl bromide 2 was synthesized from D-mannose in 9 steps. Benzoyl group was applied as protective group for 5-OH on the apiose moiety to ensure the D-configuration. (4) Under the PTC conditions, glycosylation of quercetin derivative 5 with the glucopyranosylα-bromide 6 led to the desired 3-O-β-D-glycoside 11 in 90% purified yield. The coupling of 6 and 5 occured at C-3 of 5, since the free hydroxyl at C-5 is less reactive in this reaction because of the chelating effect of the carbonyl at C-4. Compound 20 was benzylated with BnBr and K2CO3 in DMF to protect the quercetin 5-hydroxyl group, and then deacetylated with 5% acetyl chloride to give the key intermediate 4 which had three unprotected hydroxyls at C-2, C-4, C-6 of the glucosyl moiety. (5) Selectively glycosylating the hydroxyl at C-6 of 4 with the glycosyl donors 2, 3, 4- tri-O-benzoyl -α-L-rhamnopyranosyl bromide 3a and 2,3,4-tri-O-benzoyl-α-L-rhamnopyranosyl- trichloroacetimidate 3b under two different reaction conditions provided the quercetin disaccharide derivative 22 in close yield. (6) Subsequent glycosylation on 2-OH of 22 with the D-apiofuranosyl donor 2 under the catalysis of silver triflate afforded the quercetin trisaccharide derivative 23 in a moderate isolated yield. (7) Finally, removal of the benzoyl and benzyl protecting groups afforded the target product 1 in high yield. All the spectroscopic data recorded for 1 were in accordance with those of its natural counterpart.As a part of our effort to understand the antidepressant structure-activity relationship of compound 1, we designed a series of analogues of 1 in this dissertation by replacing theα-L-rhamnopyranose with D-mannose and replacing theβ-D-apiofuranose with D-ribofuranose, D-xylofuranose or L-arabinofuranose, respectively. Two compounds were successfully synthesized (32, 33). Selective glycosylation reaction of 22 with different glycosyl donors catalyzed by silver triflate was investigated. In addition, a 7,3′,4′-trihydroxyethyl derivative of compound 1 was synthesized by reaction of 1 with 2-chloroethanol.In summary, compound 1, a flavonol glycoside carrying a unique 2,6-branched trisaccharide, was synthesized in a total of 26 steps using rutin, D-glucose, L-rhamnose and D-mannose as starting materials. The longest linear sequence requires 12 steps (from D-gluscose) and an overall yield of 11% is reached. In addition, three analogues of compound 1 (32, 33, A2) have also been synthesized. These results made it possible to prepare sufficient materials for further studies on their pharmacological activities and structure-activity relationship. The method used in the construction of the trisaccharide subunit was concise and could be used in the synthesis of other glycoconjugates bearing similar 2,6-branched trisaccharides.