Chemists Discover a New Biflavonoid Hidden in the Wood of the Amazon’s Sacred Sumaúma Tree

Deep in the heart of the Amazon, one of the tallest and most culturally revered trees in Brazil has just given chemists a molecular surprise. Researchers working at the Federal University of Amazonas and the National Institute for Amazonian Research (INPA) in Manaus have isolated a previously unknown biflavonoid compound from the wood of Ceiba pentandra, the giant canopy tree known throughout the Amazon as sumaúma or sapopema. The discovery, published in Frontiers in Natural Products, adds a new entry to the chemical inventory of one of the world’s most iconic tropical trees and underscores how much molecular diversity still hides inside the timber of Amazonian species that scientists rarely study.

Ceiba pentandra is a pantropical giant belonging to the Malvaceae family, capable of towering over the forest canopy and holding deep cultural significance for riverine and Indigenous communities, who have traditionally regarded it as a sacred tree and even used its hollow trunk for acoustic communication. Botanically, the species is already known as a chemical treasure house. Previous phytochemical investigations of its leaves, bark, seeds, and flowers have revealed fatty acids, steroids, triterpenes, sesquiterpenes, xanthones, flavonoids, flavanolignans, and phenolic acids. Yet despite this rich record from other plant parts, the wood itself — the tissue that makes up the bulk of the tree — has remained almost entirely unexplored chemically, particularly for native Amazonian specimens.

The research team set out to fill that gap. Their raw material came from the trunk of a native C. pentandra specimen growing on the INPA campus in Manaus, a tree that had fallen naturally after termite infestation destroyed its taproots. The researchers pulverized 1.4 kilograms of wood and extracted it sequentially, first with hexane for seven days and then with methanol for another seven days, both at room temperature. It was the methanol extract, weighing 12 grams, that became the focus of the investigation, because polar solvents like methanol tend to pull out the phenolic compounds — including flavonoids — that chemists were most interested in finding.

Separating that crude extract into individual molecules required a painstaking sequence of chromatographic steps. The methanol extract was loaded onto a silica gel column and eluted with solvents of increasing polarity, from hexane through ethyl acetate to methanol, generating 41 fractions. Three of those fractions — numbers 26, 29, and 36 — were then subdivided further using additional silica gel columns, Sephadex LH-20 size-exclusion chromatography, and C18 solid-phase extraction. From this laborious funnel of purification, the team eventually obtained seven individual flavonoids, including six known molecules and one compound that did not match anything in the literature.

The six known compounds were identified as liquiritigenin, garbanzol, 5-deoxykaempferol, 3-methoxy-7,4′-dihydroxyflavone, 7,4′-dihydroxyflavone, and guibourtacacidine. For three of them, the finding carried extra weight: liquiritigenin, garbanzol, and guibourtacacidine had never before been reported in the genus Ceiba, while 5-deoxykaempferol, 3-methoxy-7,4′-dihydroxyflavone, and 7,4′-dihydroxyflavone were new records for the entire Malvaceae family. Each identification was confirmed by comparing nuclear magnetic resonance (NMR) spectra with published data, with the assignments cross-checked using two-dimensional NMR experiments — specifically heteronuclear single quantum coherence (HSQC) and heteronuclear multiple bond correlation (HMBC) — that map which hydrogen atoms are attached to which carbons and which nuclei influence each other across multiple chemical bonds.

The star of the study, however, was compound 7, obtained as an amorphous colorless solid in an amount of just 2.3 milligrams. High-resolution electrospray ionization mass spectrometry (HRESI-MS) established its molecular formula as C31H28O9, with the deprotonated molecular ion appearing at mass-to-charge ratio 543.1671, matching the calculated value for C31H27O9 to within 1.9 parts per million. Tandem mass spectrometry revealed characteristic fragment ions at m/z 271.0853 and 243.0903, hinting that the molecule might be built from two flavonoid units — a suspicion the NMR data would soon confirm in detail.

The proton NMR spectrum told a striking story: not one flavonoid framework but two. The researchers could distinguish two separate ABX aromatic spin systems, the telltale three-proton pattern of a differently substituted aromatic ring A, and two AA’BB’ systems, the symmetric four-proton pattern characteristic of a para-substituted ring B. In addition, the signals of saturated heterocyclic rings pointed to two flavan-3,4-diol units — a class of flavonoid scaffold in which the central C ring carries hydroxyl groups at positions 3 and 4. For the first monomer, the coupling constants of the C-ring protons at 4.44, 4.30, and 5.11 parts per million indicated a 2,3-trans, 3,4-cis relative configuration; the second monomer showed the same stereochemical pattern at slightly shifted chemical shifts. A three-proton singlet at 3.45 parts per million revealed a methoxy group, which HMBC correlations pinned to the 4-position of the second monomer.

The decisive evidence for the dimeric structure came from long-range HMBC correlations. A key four-bond correlation linked the H-3 proton of the first monomer to carbon C-6 of the second monomer, supporting a C-3 to C-7 interflavonoid connection — in other words, the two flavonoid units are joined head-to-tail through a carbon-carbon bond between position 3 of one unit and position 7 of the other. Putting all the spectroscopic pieces together, the team assigned the new molecule as rel-(2,3-trans,3,4-cis)-guibourtacacidine-(3→7)-(2,3-trans,3,4-cis)-4-O-methylguibourtacacidine, a dimeric biflavonoid never described before. The relative stereochemistry was assigned from experimental NMR data, including coupling constants, while the absolute configurations were inferred from well-established flavonoid biosynthetic logic rather than measured directly.

That biosynthetic reasoning deserves a note of its own. The configuration at carbon 2 of flavonoids originates from the stereospecific cyclization of chalcones catalyzed by the enzyme chalcone isomerase, which produces exclusively (2S)-flavanones. This configuration is then preserved through subsequent enzymatic steps: flavanone 3-hydroxylase converts flavanones to dihydroflavonols without altering the stereochemistry at C-2, and dihydroflavonol 4-reductase then performs a stereoselective reduction to yield leucoanthocyanidins — the flavan-3,4-diols that make up both halves of the new compound. Because this enzymatic cascade is highly conserved across plants, the researchers could reasonably assign the likely configurations even without experimental determination of absolute stereochemistry.

The discovery also fits into a broader chemical and ecological picture. Condensed flavonoids such as proanthocyanidins — oligomers of flavan-3-ol units — have been reported elsewhere in the Malvaceae family, notably in the genus Adansonia, the baobabs. Compound 7, however, is not a proanthocyanidin; it is a true dimeric biflavonoid built from two flavan-3,4-diol units, which makes it structurally distinct from the better-known condensed tannins and highlights the unique chemistry of Ceiba. The researchers also noted an intriguing pattern across all seven compounds: the NMR signals revealed low oxygenation in the aromatic rings, with the ABX system in ring A and AA’BB’ pattern in ring B characterizing relatively sparsely hydroxylated flavonoids. This suggests a lower biosynthetic energy cost compared with the more heavily oxygenated flavonoids found in other vegetative parts of the tree, hinting that wood chemistry may follow different metabolic economics than leaves or bark.

Beyond its chemical novelty, the finding carries a conservation and bioprospecting message. Brazil hosts one of the largest tropical forests on Earth, yet phytochemical studies of Amazonian timber species remain scarce, with most research attention directed toward leaves, fruits, and bark. By demonstrating that even the wood of a well-known species can yield molecules new to science, the study reinforces the argument that Amazonian woody species are promising sources of structurally diverse phenolic metabolites. The work was supported by the Fundação de Amparo à Pesquisa do Estado do Amazonas (FAPEAM) and the Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq), and the researchers suggest that the sumaúma — already a cultural icon of the Amazon — may now also earn recognition as a chemical one, its massive trunk harboring molecular secrets that only careful phytochemical exploration can reveal.

Subject of Research: Isolation and structural elucidation of a new biflavonoid from the wood of the Amazonian tree Ceiba pentandra

Article Title: Isolation and identification of a new biflavonoid from the Amazonian tree Ceiba pentandra (L.) Gaertn. (sumaúma)

Article References: Isolation and identification of a new biflavonoid from the Amazonian tree Ceiba pentandra (L.) Gaertn. (sumaúma). (n.d.). https://doi.org/10.3389/fntpr.2026.1888758

Image Credits: AI Generated

DOI: 10.3389/fntpr.2026.1888758

Keywords: Ceiba pentandra, sumaúma, biflavonoid, flavonoids, Amazon, natural products, NMR spectroscopy, mass spectrometry, phytochemistry, Malvaceae, flavan-3,4-diol, wood chemistry