Hydroxylamines: From Synthetic Intermediates to Synthetic Targets

Synergy between the teaching and research activities of a University should be a source of new ideas, each informing the other. A classroom discussion gave rise to our concept of using hydroxylamines as a form of “tethered nitrogen” for alkaloid synthesis. The “tether” temporarily connects a nucleophilic nitrogen atom to the substrate, rendering an intermolecular reaction intramolecular, thus providing stereo- and regiochemical control for C-N bond formation. In the context of the synthesis of 1,3-amino alcohols, this necessitated the synthesis of isoxazolidines. This concept led to the exploration of methods for the synthesis of these heterocycles beyond the well-established 1,3-dipolar cycloadditions. The first two methods developed based upon this concept were palladium catalyzed cyclocarbonylation and silver catalyzed allene cyclization. These new methods were applied to two syntheses of sedamine and the synthesis of three Nuphar alkaloids. Intramolecular aza-Michael addition, combined with the use of cross-metathesis to generate the substrates, gave access to both isoxazolidines and tetrahydro-1,2-oxazines. This new method made possible a synthesis of monomorine with high stereochemical control. The extension to more complex alkaloids required the incorporation of additional chemistry. Combining allene cyclization with iminium ion chemistry allowed the extension to the more complex piperidine alkaloids porantheridine and sedinine. In these cases, successful stereocontrol relied on the ability to predict the conformation of the intermediates and the trajectory of the nucleophilic attack, which may be sterically or stereoelectronically controlled. Inspection of our collection of methods revealed that we had good methods for cis-3,5-disubstituted isoxazolidines and trans-3,6-disubstituted tetrahydro-1,2-oxazines. Inspired by earlier work involving iminium ions, methods were developed to provide the complementary trans-isoxazolidines and cis-oxazines, using an allylation reaction. This, combined with our interest in applications of hydroformylation, led to the synthesis of 5-hydroxysedamine. Venturing away from piperidines, this method was successfully applied to the synthesis of the β-amino acid antibiotic negamycin. Use of N,O-heterocycles as intermediates naturally sparked an interest in N,O-heterocycles as synthetic targets, as there are many natural products that contain the hydroxylamine moiety, often incorporated into an isoxazolidine or 1,2-oxazine ring. As the presence of the hydroxylamine moiety cannot be fully demonstrated using the usual spectroscopic methods, total synthesis can be employed to bridge this logical gap. A synthesis of raistrickindole A, utilizing an intramolecular Mitsunobu reaction of a hydroxamic acid to form a 1,2-oxazine, confirmed the structure of this diketopiperazine natural product. A synthesis of preisomide, using an aza-Michael addition to form the required 1,2-oxazine, also confirmed the structure of this natural product.