Regioselective Copper-Mediated Organometallic Ring Openings of Acylnitroso Cycloadducts and the Development of 5-Lipoxygenase Inhibitors

Acylnitroso cycloadducts have proven to be valuable intermediates in the syntheses of a plethora of biologically active molecules. The utility of relatively simple intermediates arises from the multitude of divergent pathways which can be accessed. For many years the Miller lab has been at the forefront in the discovery of methodologies by which the acylnitroso cycloadduct core may be modified. Recently, organometallic reagents were shown to open bicyclic acylnitroso cycloadducts and, more interestingly, the prospect of highly regioselective openings was raised. This transformation was employed in the synthesis of a compound with the expectation of inhibitory activity against 5-lipoxygenase, an important mediator of inflammation intimately involved in a number of disease states including asthma and cancer. In fact, the compound proved to be an extremely potent inhibitor (IC50 51 Ì_åáM).

The focus of the research described in this work was to expand upon these chemical and biological discoveries. The first issue addressed was the optimization of the copper-mediated organometallic ring opening reaction. Conditions which led to exclusive γ-addition were discovered and employed to produce sufficient material for derivatization of the core structure. Interestingly certain aspects of the optimized reaction ran counter to the commonly employed conditions for γ-selective copper-catalyzed allylic addition. In addition, with respect to biological activity, three sites on the lead compound were chosen for derivatization.

The three sites, deemed zones 1, 2, and 3, were the iron binding group, the diaryl ether, and the cyclopentene core, respectively. Zone 1 was addressed first and synthesis of a number of derivatives with varying affinity for metal binding as well as pendant groups in a range of sizes was accomplished. Additionally, a transcarbamoylation reaction was discovered which allowed the t-butyl N-hydroxycarbamates to be converted to methyl N-hydroxycarbamates in one high yielding step.

Exploration of zone 2 was initially envisioned to proceed through an aryl halide intermediate which, via palladium-catalyzed coupling, would provide the desired substituted diaryl ethers. The problematic synthesis of the aryl halide intermediate coupled with difficulties associated with the palladium-catalyzed reaction ultimately led to abandonment of this route. An alternative route employing copper-mediated coupling of phenols and arylboronic acids was designed, executed effectively, and, after modification of the protecting group strategy, provided a set of analogues suitable for the exploration of zone 2.

Synthetic work on zone 3 derivatives was limited and although met with some success was not expanded upon. Allylic oxidations of the lead structure were attempted with numerous established reagents and conditions and in all cases failed to provide a useful intermediate. Hydroboration of the cyclopentene core successfully provided ketones after oxidation. Further exploration of this route was limited and, while unsuccessful, opportunity still exists for derivatization via this route. A final route employing the aminocyclopentenol from N-O bond reduction of the acylnitroso cycloadduct was unsuccessfully explored.

Biological evaluation of the analogues was undertaken first through the implementation of an in-house biochemical assay. The instability of the 5-lipoxygenase enzyme along with potential interference of the inhibitors with the developing reagent necessitated an alternative evaluation procedure. Through a collaboration several zone 1 analogues were tested in a whole cell assay and, while specific IC50s were not determined, very interesting results were obtained. The assay revealed a subset of the compounds to be inhibitors of enzyme translocation, a mode of action not previously known and, potentially, extremely important for better understanding of the enzyme and inhibitor development. Additionally, the lead compound was tested in vivo in an established colon cancer model and showed very encouraging anti-tumorogenic properties.