Chemoenzymatic Synthesis of Original Stilbene Dimers Possessing Wnt Inhibition Activity in Triple-Negative Breast Cancer Cells Using the Enzymatic Secretome of Botrytis cinerea Pers

Abstract

The Wnt signaling pathway controls multiple events during embryonic development of multicellular animals and is carcinogenic when aberrantly activated in adults. Breast cancer and triple-negative breast cancer (TNBC) in particular depend upon Wnt pathway overactivation. Despite this importance, no Wnt pathway-targeting drugs are currently available, which necessitates novel approaches to search for therapeutically relevant compounds targeting this oncogenic pathway. Stilbene analogs represent an under-explored field of therapeutic natural products research. In the present work, a library of complex stilbene derivatives was obtained through biotransformation of a mixture of resveratrol and pterostilbene using the enzymatic secretome of Botrytis cinerea. To improve the chemodiversity, the reactions were performed using i-PrOH, n-BuOH, i-BuOH, EtOH, or MeOH as cosolvents. Using this strategy, a series of 73 unusual derivatives was generated distributed among 6 scaffolds; 55 derivatives represent novel compounds. The structure of each compound isolated was determined by nuclear magnetic resonance and high-resolution mass spectrometry. The inhibitory activity of the isolated compounds against the oncogenic Wnt pathway was comprehensively quantified and correlated with their capacity to inhibit the growth of the cancer cells, leading to insights into structure-activity relationships of the derivatives. Finally, we have dissected mechanistic details of the stilbene derivatives activity within the pathway.

Keywords: Botrytis cinerea; Wnt inhibition; chemoenzymatic synthesis; dry load introduction; enzymatic secretome; high-resolution semi-preparative HPLC; stilbene dimers; triple negative breast cancer cells.

FIGURE 1

(A) Biotransformation reactions of the mixture of resveratrol and pterostilbene with the secretome of Botrytis cinerea in the presence of 50% of different organic solvents (i-PrOH, n-BuOH, i-BuOH, MeOH, EtOH). (B) UHPLC-MS profile of the different reactions at the analytical scale using an Acquity C₁₈ BEH column 50 × 2.1 mm i.d., 1.7 µm. The retention times of the starting compounds resveratrol and pterostilbene are indicated by the dashed line on the reaction with i-PrOH. The molecular ion species for all the main compounds detected are indicated with their corresponding m/z values.

FIGURE 2

(A) Scale-up conditions of the biotransformation reaction of resveratrol and pterostilbene with the secretome of Botrytis cinerea with 50% of i-PrOH, 2% of enzymatic secretome and 48% of water (total volume 200–400 ml). (B) Metabolite profile of the reaction after 24 h using UHPLC-PDA with optimized chromatographic conditions. (C) Analytical HPLC-PDA chromatogram after gradient transfer of the UHPLC-optimized conditions. (D) Semi-preparative HPLC-PDA chromatogram after gradient transfer of the UHPLC-optimized conditions.

FIGURE 3

Compounds isolated from the biotransformation reactions of the mixture of resveratrol and pterostilbene with the secretome of Botrytis cinerea in the presence of 50% of different organic solvents (i-PrOH, n-BuOH, i-BuOH, EtOH, or MeOH).

FIGURE 4

Mechanism of reaction explaining the generation of the 5 main scaffolds (trans-δ-viniferin, acyclic dimer, leachianol, pallidol, restrytisol A/B) of stilbene dimers obtained by the biotransformation reaction of resveratrol and pterostilbene using the enzymatic secretome of Botrytis cinerea in the presence of alcohol as cosolvents.

FIGURE 5

Overview of the typical ¹H-NMR shifts of the different scaffolds generated. (A) Five compounds representing the five main scaffolds obtained by the biotransformations reactions. (B) The typical ¹H-NMR spectra of each scaffold is highlighted in blue for the trans double-bound; orange for the 1,3,4-trisubstituted aromatic cycle; red for a para-disubstituted system; purple for the 1,3,5-trisubstituted aromatic cycle; green for the two meta coupled proton; and yellow for the methine protons.

FIGURE 6

Key ROESY correlations observed for (A) leachianol F series (B) leachianol G series and (C) for the different isomer (compound 32).

FIGURE 7

Library chemical similarity and pharmacophore models. (A) Similarity matrix for all six series. A value of 1 corresponds to the identity. (B) Pharmacophore model for specific actives overlapped on the four compounds with an SI > 10. (C) Pharmacophore model for non-specific actives overlapped on 62. The red and green spheres represent respectively an H-bond acceptor and a hydrophobic region. The orange circle represents the ideal 3D positioning of aromatic rings.

FIGURE 8

Relations between derivative scaffolds and anti-Wnt activity. (A) Distribution between the selectivity index (SI) and anti-Wnt potency of the investigated compounds. The compounds selected for further biochemical analysis are highlighted as stars (B) Correlation between anti-Wnt IC₅₀ and IC₅₀ of long-term (3 days) inhibition of the cell growth measured by MTT assay (C) Correlation between anti-Wnt IC₅₀ and IC₅₀ of acute cytotoxicity.

FIGURE 9

Representative compounds from the different structural groups of the derivatives (A) inhibit expression of the basal levels of Wnt target genes Axin2 (B) and c-Myc (C) in the TNBC cell lines BT-20 (left) and MDA-MB-468 (right). Quantification of the Western blots is presented on (D). Statistical significance was assessed by two-way ANOVA followed by multiple comparisons with DMSO-treated values, p values are shown as **p < 0.01, ***p < 0.001.