A chemical investigation of leaves collected from the African tree species Erythrina caffra has yielded a previously undescribed sterol glycoside alongside several known secondary metabolites, providing empirical insight into the plant’s chemical profile and baseline in vitro biological properties. The study, conducted by a team of researchers examining African medicinal flora, focused on separating and structurally characterizing the primary constituents of an ethanol leaf extract while profiling fractions using liquid chromatography coupled with tandem mass spectrometry. Beyond chemical isolation, the investigators conducted standardized laboratory assays to assess the total phenolic content, radical-scavenging capacities, direct cytotoxicity across three immortalized cell models, and in vitro inhibitory activity against the severe acute respiratory syndrome coronavirus 2 papain-like protease.

The plant material was gathered in May 2024 from the campus grounds of Sefako Makgatho Health Sciences University in Ga-Rankuwa, Pretoria North, within South Africa’s Gauteng Province. Following collection, the botanical specimen was formally authenticated at the South African National Biodiversity Institute by comparison against a verified reference specimen cataloged under voucher number F. Olawale 1. In traditional medicine across southeastern and southern Africa, the stem bark, roots, and leaves of Erythrina caffra have documented ethnobotanical applications for managing inflammatory conditions, superficial wounds, arthritis, earaches, and localized infections. Because previous investigations into the genus Erythrina had documented diverse secondary metabolites such as flavonoids, alkaloids, and terpenoids possessing bioactivity, the investigators sought to establish whether the foliar tissues of this species contained specialized metabolites that could substantiate these traditional uses under rigorous experimental evaluation.

To obtain sufficient material for chromatographic separation, the botanical samples were thoroughly dried and ground into 3.6 kilograms of fine powder, which subsequently underwent cold maceration three successive times for 24 hours each in 10 liters of ethanol at room temperature. Concentration of the resulting filtrate under reduced pressure with a rotary evaporator yielded 202 grams of dry crude extract. A 200-gram portion was resuspended in 500 milliliters of distilled water and partitioned against 1000 milliliters of ethyl acetate in a liquid-liquid extraction system repeated three times, producing 110 grams of an ethyl acetate-soluble fraction and leaving 86 grams of a polar aqueous residue designated as the methanol fraction.

Chromatographic fractionation of a 100-gram portion of the ethyl acetate phase over silica gel stationary phases using gradient mixtures of n-hexane and ethyl acetate yielded seven distinct sub-fractions, designated LA through LF. Subsequent purification of these sub-fractions through silica gel column chromatography resulted in the isolation of four pure chemical entities. Sub-fraction LB yielded compound 4 as 2.0 grams of a white amorphous powder, while its secondary chromatographic resolution gave sub-fraction LB3, yielding compound 2 as 230 milligrams of white amorphous powder. Fraction LE yielded 326.1 milligrams of compound 3 as a distinct red powder, and fraction LF yielded 32.0 milligrams of compound 1, which precipitated as a viscous gum.

Structural elucidation of the isolated constituents was carried out using high-resolution electrospray ionization mass spectrometry along with one-dimensional and two-dimensional nuclear magnetic resonance spectroscopy, encompassing proton, carbon-13, DEPT-135, homonuclear correlation spectroscopy, heteronuclear single quantum correlation, and heteronuclear multiple bond correlation experiments. Compound 1 was determined to be a previously unreported sterol derivative, named trivially as erythrinoside and systematically designated as stigmasta-5,22-diene 3-O-(6’-tridecanoate)-β-D-glucopyranoside. High-resolution mass spectrometry established its molecular formula as C48H82O7 through a sodium adduct ion at mass-to-charge ratio 793.5988, while spectral correlations confirmed the attachment of a β-D-glucopyranosyl unit at carbon-3 of the stigmasterol skeleton and an esterified thirteen-carbon tridecanoate fatty acyl chain at the 6’-position of the sugar moiety. The remaining isolated compounds were identified by spectroscopic comparison with published literature as stigmasterol (compound 2), the dietary carotenoid lutein (compound 3), and the long-chain fatty alcohol triacontanol (compound 4).

In parallel, the ethyl acetate and methanol fractions were analyzed by ultra-performance liquid chromatography coupled to quadrupole time-of-flight tandem mass spectrometry using a reversed-phase C18 column and an acidified water-methanol gradient. Qualitative profiling in both positive and negative electrospray ionization modes tentatively identified twenty-two secondary metabolites based on accurate mass measurements, elemental compositions generated in MassLynx, and database matching against LOTUS, ChemSpider, and PubChem, supported by diagnostic fragmentation patterns. The ethyl acetate fraction revealed sixteen putative compounds, including prenylated flavonoids and isoflavones such as 6-prenylgenistein, 6-prenylapigenin, erythrivarone A, carpachromene, and abyssinone V 4’-methyl ether. The polar methanol fraction yielded six metabolites, predominantly comprising C-glycosyl flavonoids such as isoschaftoside, luteolin-6-C-β-D-glucopyranoside, and vitexin, which exhibited characteristic cross-ring cleavage fragments during collision-induced dissociation.

Quantitative determination of the total phenolic content using the Folin-Ciocalteu colorimetric method against a gallic acid standard calibration curve demonstrated clear differences in polyphenol concentration across the extracts. The methanol fraction exhibited the highest phenolic density at 148.2 ± 2.02 milligrams of gallic acid equivalents per gram of dry sample, whereas the crude ethanol extract displayed a moderate level of 50.38 ± 0.17 milligrams per gram, and the ethyl acetate fraction showed a relatively low value of 17.83 ± 0.15 milligrams per gram. These quantitative results indicated that polar solvent partitioning concentrated the bulk of the foliar phenolic and flavonoid constituents into the aqueous-methanolic phase rather than the intermediate ethyl acetate phase, notwithstanding the chromatographic detection of numerous individual lipophilic flavonoids in the latter.

The radical-scavenging capacities of the extracts, fractions, and isolated compounds were quantified using two mechanistic assays: the Oxygen Radical Absorbance Capacity assay, which operates through hydrogen atom transfer, and the 2,2’-azino-bis(3-ethylbenzothiazoline-6-sulfonic acid) assay, which functions via single electron transfer, both standardized against Trolox equivalents. In the electron-transfer assay, the crude extract showed substantial scavenging activity at 3.831 ± 0.29 micromoles of Trolox equivalents per milligram, correlating closely with measured phenolic contents. Among the isolated compounds evaluated in this system, only lutein demonstrated measurable activity, recording 0.5086 ± 0.05 micromoles of Trolox equivalents per milligram, while failing to exhibit activity in the fluorescence-based hydrogen atom transfer assay. In the oxygen radical absorbance assay, the polar methanol fraction exhibited the strongest overall antioxidant capacity at 5.52 ± 0.14 micromoles of Trolox equivalents per milligram, consistent with its enriched polyphenol and C-glycosyl flavonoid content, whereas the crude extract and ethyl acetate fraction displayed modest values near 0.6 micromoles per milligram.

In vitro cytotoxicity was evaluated against three non-tumorigenic cell cultures: human immortalized keratinocytes (HaCaT), normal human adult epidermal melanocytes (NHEM-Ad), and human embryonic kidney cells (HEK293). Cells were exposed to a uniform test concentration of 0.1 milligrams per milliliter (equivalent to 100 micrograms per milliliter) for 24 hours, followed by viability quantification using the fluorometric resazurin reduction assay and biomass density measurement through crystal violet staining. The crude extract, fractions, and compounds 1, 2, and 4 exhibited no significant cytotoxic effects against keratinocytes or melanocytes, preserving a benign profile in skin-derived models. In contrast, lutein tested at this elevated concentration of 100 micrograms per milliliter significantly diminished cell viability to 39.19 ± 9.27 percent in HaCaT, 61.39 ± 3.26 percent in NHEM-Ad, and 71.29 ± 7.82 percent in HEK293 cells, accompanied by corresponding reductions in adherent cell density to 24.35 ± 4.29, 35.96 ± 1.44, and 39.16 ± 6.17 percent, respectively.

The cytotoxicity experiments also revealed cell-type-specific vulnerabilities within the renal lineage that were absent in dermal models. When applied to HEK293 embryonic kidney cells, the crude extract, ethyl acetate fraction, and the newly isolated sterol glycoside erythrinoside induced measurable reductions in metabolic viability, lowering survival to 76.7 ± 0.83 percent, 70.5 ± 1.07 percent, and 85.21 ± 1.67 percent, respectively. The investigators noted that this heightened susceptibility in HEK293 cells aligns with the high mitochondrial metabolic turnover and xenobiotic-handling physiological function characteristic of renal cells, contrasting with the relative tolerance demonstrated by keratinocyte and melanocyte monolayers under identical in vitro exposure conditions.

Direct antiviral screening was performed using an in vitro enzymatic cleavage assay targeting recombinant SARS-CoV-2 papain-like protease with the fluorogenic substrate Z-RLRGG-AMC across a concentration range from 0.1 to 0.8 milligrams per milliliter. The reference chemical inhibitor GRL0617 exhibited potent enzymatic suppression with a half-maximal inhibitory concentration of 0.487 micrograms per milliliter, confirming test validity. In comparison, all evaluated botanical extracts and isolated compounds demonstrated markedly weaker inhibitory activity. The methanol fraction exhibited only limited suppression with a half-maximal inhibitory value below 100 micrograms per milliliter, while isolated lutein yielded an inhibitory concentration of 389.0 ± 0.35 micrograms per milliliter. The authors concluded that the leaf constituents of Erythrina caffra do not serve as potent direct inhibitors of the viral protease in vitro, emphasizing that future investigations must isolate individual active components from polar fractions and validate their antioxidant or cytoprotective mechanisms through controlled in vivo models before therapeutic relevance can be established.

Subject of Research: Phytochemical profiling and in vitro biological evaluation of Erythrina caffra leaf extracts and isolated constituents

Article Title: LC-MS/MS-based phytochemical profiling of Erythrina caffra leaves, isolation of a novel sterol glycoside, and evaluation of antioxidant, cytotoxic, and anti-sars-cov-2 activities

Article References: Tsakem, B., Serem, J. C., Hlophe, Y. N., Sikhakhane, N., Teponno, R. B., Sandjo, L. P., & Siwe Noundou, X. (2026). LC-MS/MS-based phytochemical profiling of Erythrina caffra leaves, isolation of a novel sterol glycoside, and evaluation of antioxidant, cytotoxic, and anti-sars-cov-2 activities. Applied Biological Chemistry, 69(1), Article 49. https://doi.org/10.1186/s13765-026-01126-w

Image Credits: AI Generated

DOI: 10.1186/s13765-026-01126-w

Keywords: Erythrina caffra, phytochemistry, erythrinoside, natural products, antioxidant activity, cytotoxicity, spectroscopy, metabolomics, LC-MS, MS-based, phytochemical, profiling