Total Syntheses and Cytotoxicity of (R)- and (S)-Boehmeriasin A

Abstract

Both enantiomers of boehmeriasin A were synthesized in seven steps each, using a chiral pool approach. Key steps in the synthesis are a one-flask, two-step protocol to generate the quinolizine core and a C-H functionalization reaction between tetrahydroquinolizinones and an aryltrifluoroborate. The natural product (R)-boehmeriasin A demonstrated potent cytotoxicity against several cancer cell lines, whereas the unnatural (+)-(S)-isomer was significantly less potent.

Boehmeriasin A and B were recently isolated from the aqueous ethanolic extract of Boehmeria siamensis via bioassay-guided fractionation (Figure 1).¹ The alkaloids were evaluated against a panel of cancer cell lines that included leukemia and cancers of the lung, colon, breast, prostate, and kidney. Boehmeriasin A was found to be more potent than paclitaxel and boehmeriasin B in most cell lines evaluated with GI₅₀ values ranging from 0.80 to 265 nM. In addition, boehmeriasin A potently inhibits the proliferation of the breast cancer cell line MDA-MB-231 through G1 cell cycle arrest and differentiation induction by altering the expression levels of several genes involved with cell proliferation, cell cycle regulation, and apoptosis.² The synthesis of this natural product was first reported in racemic form leaving the absolute stereochemistry of the natural product initially unknown.³

Figure 1. Structures of boehmeriasin A and B.

While our work on the enantiospecific synthesis of (R)- and (S)-boehmeriasin A was ongoing, an asymmetic synthesis of (R)-boehmeriasin A was reported.⁴ The synthesis of the natural product was achieved in 13 steps and involved the SAMP-hydrazone method to introduce asymmetry, ring-closing metathesis to form the piperidine ring system, an aldol reaction to prepare the quinolizidine core structure, and a radical reaction to furnish the phenanthrene moiety.

We had become interested in this natural product because of the promising anticancer activity reported and our interest in this class of compounds.⁵ We now report the total synthesis of both boehmeriasin A enantiomers and data concerning their in vitro anticancer activity. In a retrosynthetic sense (Scheme 1), boehmeriasin A (1) was to be derived from an intramolecular bi-aryl coupling of intermediate 2 which itself was to be accessed from a palladium-mediated cross-coupling reaction. The coupling partner 3 for the cross-coupling was to be obtained by conversion of cyclic enaminone 4 to the triflate 3. This advanced intermediate 4 was envisioned to arise from a novel, palladium(II)-catalyzed C-H functionalization utilizing organotrifluoroborates.⁶ The coupling partner for this transformation would be obtained from Weinreb amide 6 through methods developed in our laboratories.⁷

Scheme 1. Retrosynthesis for boehmeriasin A.

The commercially available acids 7 were converted into the corresponding Weinreb amides 6 under standard conditions and subsequently treated with ethynylmagnesium bromide to afford the corresponding ynones 8 in excellent yields (Scheme 2).^8,9 These intermediates were subjected to a one-flask, two-step protocol for the cyclization of Boc-ynones to enaminones to afford the desired enaminones 5 in good yields.⁷

Scheme 2. Synthesis of α-arylated enaminones 4.

With rapid access to the desired enaminones established, our recently reported palladium(II)-catalyzed C-H functionalization using enaminones and potassium organotrifluoroborates was employed.⁶ This method represents an efficient means of accessing these products, as previous methods required an initial prefunctionalization of the enaminone to the appropriate α-halogenated derivative followed by a Suzuki coupling.¹⁰ This protocol eliminates this requirement and thus allows for a more streamlined approach to access these compounds. Utilizing potassium 3,4-dimethoxyphenyltrifluoroborate (9), prepared according to a known procedure,¹¹ in the Pd(II)-catalyzed reactions with enaminones 5 furnished the desired arylated products 4 in good yields (Scheme 2).

With a route to the α-arylated intermediates established, the final synthetic sequence was undertaken as shown in Scheme 3. Enaminones 4 were treated with L-Selectride and the resultant enolates were trapped with Comins' reagent (N-(5-chloropyridin-2-yl)-1,1,1-trifluoro-N-(trifluoromethylsulfonyl)methanesulfonamide) to arrive at the desired triflates 3 in good yields.¹² A Negishi cross-coupling was then employed to furnish the desired intermediates 2 in near quantitative yields.¹³ The synthesis was completed utilizing an oxidative bi-aryl ring closure mediated by VOF₃ to afford (R)- and (S)-boehmeriasin A in good yields.¹⁴ The final products were crystallized from CH₂Cl₂/MeOH and the crystal structure for the (R)-antipode was determined.¹⁵ It is important to note that little racemization occurred throughout the course of the synthesis to afford (R)- and (S)-boehmeriasin A in 97.5:2.5 er and 98:2 er as determined by chiral HPLC.¹⁶ In addition, the specific rotation of the (R)-enantiomer gave a matching sign and magnitude with those found in the literature (lit. −80, MeOH, c = 0.10; obs. −86, MeOH, c = 0.10).^1,4

Scheme 3. Completion of the synthesis of (R)- and (S)-1.

(R)- and (S)-Boehmeriasin A and synthetic intermediates were subjected to in vitro cytotoxicity assays to confirm the reported biological activity and establish an initial SAR for boehmeriasin A in breast (MCF7), drug-resistant ovarian (NCI-ADR-RES), and colon (COLO-205) cancer cell lines (Table 1). Arylated enaminone 3, and seco-boehmeriasin A (2) were devoid of any cytotoxic activity in the cell lines evaluated, which indicates that a full phenanthrene ring system is required for potent cytotoxic activity. This is in accord with other studies of this class of natural products and their analogues.¹⁷ Furthermore, (−)-boehmeriasin A ((−)-1·HCl) was more potent than its antipode, (+)-1·HCl, in all of the cell lines evaluated, indicating that the (R)-configuration is essential for potent cytotoxic activity. Most significantly, the natural product showed activity in the drug resistant cancer cell line, NCI-ADR-RES, where paclitaxel is inactive.

Table 1.

Cytotoxicity evaluation of (−)-(R)- and (+)-(S)-boehmeriasin A in comparison to paclitaxel.

Compound	IC50 (nM)
	COLO-205	MCF-7	NCI-ADR-RES
paclitaxela	3.31	1.62	>6400
(−)-(R)-1·HClb	4.18	43.4	36.7
(+)-(S)-1·HCla	103	92.7	434

COLO-205 = human colorectal adenocarcinoma (GI₅₀ = 0.80 nM reported for (-)-1, see ref. 1); MCF-7 = human breast carcinoma (GI₅₀ = 13 nM reported for (-)-1, see ref. 1); NCI-ADR-RES = drug-resistant human ovarian adenocarcinoma;

^aaverage of six assays each;

^baverage of three assays each.

In summary, the total syntheses of (−)-(R)- and (+)-(S)-boehmeriasin A were accomplished in seven steps from commercially available material with an overall yield of 33% and the absolute stereochemistry of the natural product was verified to be of the (R)-configuration. The synthesis showcases the utility of the enaminone chemistry and the palladium(II)-mediated C-H functionalization developed in our laboratories. When evaluated for cytotoxic activity, (−)-(R)-boehmeriasin A demonstrated potent cytotoxicity in several cancer cell lines including a drug-resistant cancer cell line where paclitaxel is inactive. The (S)-enantiomer was significantly less potent. (R)-Boehmeriasin A will serve as a lead compound for further development and studies in this regard are currently underway in our laboratories.