High-performance countercurrent chromatography isolation of flavonoids from a cytotoxic extract of Andira parviflora (Ducke) waste wood from the Amazon region.

Andira parviflora Ducke (Fabaceae), known as "sucupira vermelha", is mainly found in dryland forests of the Amazon region. Due to its large size and dense wood, the timber industry widely uses the species. Nevertheless, there are few studies about the chemical composition of its residues from industrial wood processing. Some pterocarpans and isoflavone derivatives have already been reported in the literature. In this work, we report the isolation of phenolic compounds from the heartwood waste of A. parviflora, using high-performance countercurrent chromatography (HPCCC) associated with high-performance liquid chromatography coupled to ultraviolet spectroscopy (HPLC-UV), applied to the ethyl acetate fraction of its ethanol extract, which showed potential cytotoxic activity. The cytotoxic potential was evaluated against tumor cells astrocytoma, human colon carcinoma, prostate carcinoma, and promyelocytic leukemia cell lines. The purification of 4 g of the ethyl acetate fraction was carried out by reversed-phase HPCCC using a semipreparative 136 mL column and a non-linear step-gradient elution two biphasic solvent systems: hexane-ethyl acetate-methanol-water 1:3:1:3 v/v (step 1) and 1:2:1:2 v/v (step 2), with the aqueous phase of the both systems used as mobile phase. The HPLC-UV purification of fractions 4 and 17 afforded the two dihydrochalcone isomers of coatline A (compounds 1 and 2) from fraction 4, as well as isoliquiritigenin, (3), medicarpin, (4), and biochanin A (5). Compounds 1 and 2 are unprecedented in the species whereas compound 3 is being described for the first time in the genus. The use of preparative HPCCC prior to HPLC purification enabled the separation of three different classes of flavonoids from A. parviflora and simplified the samples before HPLC purification. Moreover, it reduced the overall cost of the process by minimizing solvent consumption, shortening processing time and enhancing sample recovery.