Progress toward the total synthesis of 9β-hydroxyvertine: Construction of an advanced quinolizidine intermediate

Abstract

Progress toward the total synthesis of 9β-hydroxyvertine 2 is described. Our approach involves a Petasis borono-Mannich reaction to assemble a key intermediate A, securing the correct configuration at H-9 and H-10 in the final targeted molecule. Subsequent formation of the quinolizidine framework B allowed the synthesis of an advanced intermediate containing all but the lactone moiety of 2. Our unsuccessful attempts at introducing this lactone feature are also described.

Introduction

Since the discovery of seven alkaloids from the extracts of Decodon verticillatus (family Lythraceae) by Ferris in 1962 many other Lythraceae alkaloids have been isolated, especially from Heimia salicifolia, a wild flowering plant in the Lythraceae family [[1], [2], [3], [4]]. This plant has been used in traditional medicine as an antipyretic, an emetic, a laxative, a diuretic and as an anti-inflammatory agent [1,3,4]. Several Lythraceae alkaloids were found to exhibit remarkable biological activities, including vertine 1 (Fig. 1), first isolated by Ferris [2] and more recently in 2008 by Khan and coworkers [5], from the leaves of H. salicifolia, who demonstrated it had anti-inflammatory and antimalarial activities. The latter author also established the configuration of vertine at the three stereogenic centres and the chiral axis (as shown in Fig. 1) by X-ray diffraction analysis [5]. This alkaloid, and others in this family, comprise a quinolizidine ring skeleton with an axially chiral biphenyl substituent at C-4 which forms part of a 12-membered ring lactone system. The related alkaloid, 9β-hydroxyvertine 2, having a β-hydroxy group at the C-9 position, was also isolated in this study, however no information concerning its biological activity was reported. Other members of this alkaloid family include decinine (10-epi-13,14-dihydrovertine), lythrine (10-epi-vertine) and lyfoline 3 [2,6]. A 2018 study reported the isolation on 13 new related alkaloids along with vertine, lythrine and lyfoline 3 (Fig. 1) and three related biphenyl ether quinolizidine lactone alkaloids [7].

Because of their challenging ring structures and interesting biological properties, Lythraceae alkaloids have become attractive synthetic target molecules. Vertine 1 was first synthesized by Kündig and co-workers in 2010, as the racemate and then in 2012 as the (+)-enantiomer using the ring closing metathesis (RCM) and the Suzuki–Miyaura coupling reactions as key steps to establish the macrocyclic ring (by formation of the C-13–C-14 (Z)-olefinic bond) and the biaryl axis (forming the C-1′–C-1″ bond), respectively [8,9]. In 2015, She and co-workers reported the synthesis of the related alkaloid dihydrolyfoline (13,14-dihydro-3), using a hydrolase-catalyzed Mannich reaction to form the quinolizidine ring and a biogenetic, enzymatic oxidative intramolecular coupling reaction of phenols to construct the macrocyclic ring and the biphenyl axis [10]. Much earlier approach to these types of structures, using non-enzymatic oxidative coupling methods, proved unsuccessful [11,12]. However, Yang and coworkers reported the successful synthesis of decinine using a VOF₃-mediated intramolecular oxidative coupling reaction of aryl ethers [13]. To date the total synthesis of 9β-hydroxyvertine 2 has not been reported. Therefore, this challenging new quinolizidine core structure led us to investigate its synthesis.