Diastereoselective Spiroannulation of Phenolic Substrates: Advances Towards the Asymmetric Formation of the Manumycin m-C7N Core Skeleton

Abstract

The asymmetric syntheses of two new spirolactones prepared in optically pure form from L-3-nitrotyrosine are described. The key step, an oxidative spiroannulation, was carried out on the optically active phenols 11a and 11b and afforded the new spirolactones 5a and 5b in 85% and 83% yields, respectively, as mixtures (3:1 dr) of diastereomers. The major diastereomers from these mixtures could be isolated in optically pure form by trituration using acetone-hexanes as the solvent. Thus, the optically active spirolactones (+)-5a (+ 92.8^o, c=0.125 acetone) and (+)-5b (+112.0^o, c= 0.125 acetone) were obtained after four synthetic steps from L-3-nitrotyrosine in 41% and 43% yield, respectively.

Introduction

Our previous investigations of the diastereoselective spiroannulation of simple phenols have shown that it is possible to obtain facial selectivity in these reactions [1,2,3,4]. While only modest diastereoselectivities (~4:1 ratio) have been obtained to date, these results suggested that this method could be used in the design of diastereoselective syntheses of natural products. More accurately, we felt that spirocompounds of general structure 1 (Figure 1) could be used in the synthesis of the core carbon skeleton of natural products (or analogs) from three families: Gymnastatins (2), Aranorosins (3) and Manumycins (4). A representative example for each family is shown in Figure 1 (structures 2-4).

Figure 1.

Figure 1.

Assuming that the spirocompound 1 bears the proper functional groups at positions R₁ and R₂, it appears to be a good starting point for the syntheses of these natural products. While there are already numerous syntheses of members of the Manumycins [5,6,7,8,9,10,11,12,13], Aranorosins [14,15,16] and Gymnastatins [17], this new approach could be advantageous since it may allow for the synthesis of more than one of these structures from the same starting material. We have investigated the asymmetric synthesis of spirolactones 5 (Figure 2) bearing substituents that would mimic the general functionality found in most members of the Manumycin family (cf. structure 4). We now wish to report our preliminary results toward the asymmetric synthesis of compound 5, which we hope will become a precursor for the synthesis of analogs of the Manumycin family of compounds.

Results and Discussion

The diastereoselective formation of spirolactones 5a and 5b is described in Scheme 1. Commercially available L-3-nitrotyrosine 8 was chosen as the starting material for two reasons: (1) it provides the means to generate the amide function found in the Manumycins via its nitro function (see Figure 1), and (2), it bears an optically pure chiral center on the side chain para to the hydroxyl, as shown in phenols 11a/11b. Our previous studies of the diastereoselective spiroannulation of simple phenols have shown that for side chains similar to the one found in 11, it was possible to induce face selectivity and obtain a 4:1 diastereomeric ratio of spiroethers 6a-c and 7, which are analogous in structure to 5/12, as shown in Figure 2 [1,2,4]. We hoped that in the formation of spirolactones 5/12, a similar or greater ratio of diastereomers could be generated since the chiral center in the side chain of 11 is located at a position similar to that of the chiral center used in our previous studies.

Figure 2.

Figure 2.

Scheme 1.

Scheme 1.

Treatment of 8 with p-toluenesulfonyl chloride in tetrahydrofuran afforded a mixture of products consisting of the mono-N-protected phenol 9 and a product that we identified by ¹H-NMR as a diprotected product, ie. both the amino and hydroxyl functions in 8 had reacted with the sulfonyl chloride. Hydrolysis of the resulting mixture with 1M KOH at 80^oC afforded only 9 in 86% yield as ayellow solid after recrystallization from benzene. Since the sulfonate ester is easier to hydrolyze than the sulfonamide function, we were able to avoid the extra steps that protection/deprotection of the phenolic hydroxyl would have necessitated. Hydrogenation of the nitro function in 9 afforded the aniline derivative 10 as a white solid. The crude yield for 10 appeared to be quantitative (based on mass recovery), however, every attempt to purify this compound by column chromatography or recrystallization failed. In these cases, we obtained low yields (40-60%), but more importantly, these purification attempts at 10 generated a product that was less pure (based on ¹H-NMR) than the crude reaction product itself. Consequently, compound 10 was used without purification in the subsequent reaction and the identity of this compound was based primarily on its ¹H-NMR spectrum.

Treatment of 10 with either acetyl chloride or benzoyl chloride afforded 11a and 11b in 84% and 87% yields, respectively, after purification by column chromatography. The spiroannulation of 11a and 11b was carried out with phenyliodine(III) bistrifluoroacetate (PIFA) in acetone at 0^oC. In the reaction of 11a, a mixture of diastereomers 5a/12a was obtained in a 3:1 diastereomeric ratio. The ratio of major/minor diastereomers formed in the spiroannulation was determined by integration for the signals of H-9 (two doublets centered at 6.37 ppm) in the ¹H-NMR spectrum of the crude diastereomeric mixture. In the reaction of 11b, the two doublets for compounds 5b/12b (centered at 6.40 ppm) were overlapping slightly, making the determination of the diastereomeric ratio less accurate. However, it was obvious from the ¹H-NMR of the crude reaction mixture that one stereoisomer was produced in a larger amount than the other one. We estimated that for 5b/12b this ratio was between 3:1 and 4:1. Attempts to separate the major and minor stereoisomers from these mixtures met with limited success. Our first attempt was carried out by column chromatography of the 5a/12a mixture using 50% ethyl acetate-hexanes as eluant. We were able to separate the major diastereomer 5a from the mixture under these conditions but this fraction represented only 38% of the total isolated mass (or 32% yield from 11a). We were not successful in isolating the minor diastereomer from this mixture. We later found that both 5a and 5b could be isolated via trituration using an acetone/hexanes mixture. In both cases, we were able to isolate the major diastereomers (+)-5a and (+)-5b from the crude reaction mixtures. The isolated mass obtained for each compound represents a 57% yield from the corresponding optically active phenols 11a/11b, or a 41% total yield for (+)-5a and 43% for (+)-5b. Finally, while we were able to isolate the major diastereomer in these reactions, we have not yet identified the configuration of the newly formed chiral spirocenter in 5a or 5b. As a result, in Scheme 1 we have not drawn the stereo bonds of the spirocenters in the structures of 5a/b or 12a/b. However, both major diastereomers gave a positive optical rotation and are labeled accordingly in Scheme 1.

Conclusions

We have successfully prepared optically active spirocompounds (+)-5a and (+)-5b. While the diastereomeric ratio observed in the reaction leading to these compounds is modest at best, it suggests that the method could be useful in the asymmetric synthesis of natural products since we were able to obtain the major diastereomers in optically pure forms. We have now begun an investigation of the use of these compounds in the synthesis of the core skeleton of the Manumycins. Furthermore, we are attempting to identify the configuration of the new chiral spirocenters in (+)-5a and (+)-5b as well as ways to enhance the diastereomeric ratios observed in these reactions.