The compound, named 5-O-desmethyl kinanthraquinone B, or 5DKQB, was isolated from a laboratory strain in which one of the two O-methyltransferase genes in the kinanthraquinone biosynthetic gene cluster had been disrupted. The work, published as a Brief Report in The Journal of Antibiotics, builds on the group’s earlier identification of the kinanthraquinone biosynthetic gene cluster in the actinomycete Actinacidiphila reveromycinica SN-593 and their previous isolation and structural elucidation of new analogues of this anthraquinone.

Because O-methylation is a common tailoring reaction that can modulate the pharmacokinetic and pharmacodynamic properties of bioactive compounds, the team hypothesised that O-methylation, in addition to the side-chain carboxamide group they had previously shown to be essential for antimalarial activity, contributes to the biological profile of KQ analogues. Targeted gene disruption offered a direct way to test that idea.

The experiment was carried out in a heterologous host: Streptomyces lividans TK23 carrying a construct in which the kiqT1 gene had been deleted. Ultra-performance liquid chromatography coupled with mass spectrometry of the culture broth after 48 hours of cultivation revealed the accumulation of a compound with a molecular weight of 368, exactly 14 Da lower than that of kinanthraquinone B, whose molecular weight is 382. That 14-Da difference is the signature of a single demethylation, pointing to a demethylated derivative of KQB.

To obtain enough material for full characterisation, the researchers scaled up fermentation. The engineered strain was grown in 7 litres of Tryptic Soy Broth for 48 hours, after which the culture broth was mixed with an equal volume of acetone and filtered to remove bacterial cells. The filtrate was concentrated under reduced pressure to strip out the acetone, and the remaining aqueous solution was chromatographed on an HP20 column eluted stepwise with water-methanol mixtures of increasing organic content. The 40:60 fraction was then applied to an ODS column and eluted with a further stepwise water-methanol gradient, and the target fraction was polished by two rounds of preparative HPLC on PEGASIL ODS SP100 columns, first with a linear gradient of 10 to 100 percent acetonitrile in 0.05 percent aqueous TFA and then isocratically with 50 percent acetonitrile in 0.025 percent aqueous TFA. The yield was modest: 1.3 milligrams of a reddish-brown amorphous powder, designated 5DKQB.

Structural elucidation rested on high-resolution electrospray ionisation mass spectrometry and nuclear magnetic resonance spectroscopy, with the data subsequently compared against the spectroscopic profile of kinanthraquinone B. HR-ESI-MS gave a found mass of m/z 369.0974 for the protonated molecule, matching the calculated value for the formula C20H16O7 and confirming the 14-Da deficit relative to KQB. The UV-Vis spectrum showed absorption maxima at 240, 264 and 459 nanometres, nearly identical to those of the known analogue KQD, while the infrared spectrum displayed characteristic carbonyl absorptions at 1674 and 1633 wavenumbers, indicating that the new compound possesses the same anthraquinone skeleton as the other KQ analogues.

The decisive structural question was where the missing methyl group had been. Comparison of the carbon-13 NMR data of 5DKQB with those of KQB showed that all carbon signals except those assigned to C-5 and C-6 were essentially identical, meaning the two compounds share the same carbon skeleton and the difference is localised to one position on the aromatic ring. A proton-proton COSY correlation between H-7 and H-8, together with an ortho coupling constant of 7.8 hertz, established the relationship between those two aromatic protons. HMBC correlations from H-7 to C-5 and C-8a, and from H-8 to C-6 and C-10a, mapped the substitution pattern, and the HMBC correlation from the 6-methoxy protons to C-6, reinforced by a ROESY correlation between the 6-methoxy group and H-7, unambiguously placed the remaining methoxy group at C-6. That left C-5 bearing a free hydroxyl group, the position that carries a methoxy substituent in kinanthraquinone B itself.

The side chain was confirmed to be identical to that of KQB through a set of HMBC correlations linking the vinylic proton H-1′ and the methyl protons H3-5′ to the side-chain carbons, and an HMBC correlation from H-1′ to C-3 of the anthraquinone core established the point of attachment. The configuration of the C-1′ double bond was assigned as Z on the basis of the carbon-13 chemical shift of the vinylic methyl carbon at 24.7 parts per million, consistent with the reported criterion that a vinylic methyl resonance above 20 parts per million indicates a Z configuration, and this assignment was further supported by a ROESY correlation between H-1′ and H3-5′. Taken together, the spectroscopic data established 5DKQB as a new derivative of kinanthraquinone.

With the structure secured and purity confirmed by UPLC, the team evaluated the compound’s biological activities against a panel of test organisms and cells: Staphylococcus aureus 209, Escherichia coli HO141, the filamentous fungus Aspergillus fumigatus Af293, the yeast Candida albicans SC5314, the rice blast fungus Pyricularia oryzae kita-1, the malaria parasite Plasmodium falciparum 3D7 and human HL-60 leukaemia cells. Activity was observed against three of these targets. 5DKQB inhibited S. aureus with an IC50 of 3.7 micrograms per millilitre, P. falciparum with an IC50 of 5.5 micrograms per millilitre, and HL-60 cells with an IC50 of 11 micrograms per millilitre.

The comparison across the series is where the chemistry becomes informative. Setting the IC50 values of 5DKQB against those of kinanthraquinone B and kinanthraquinone itself showed that O-methylation at the C-5 hydroxyl group reduces antibacterial activity and cytotoxicity while having little effect on antimalarial activity. In other words, the demethylated analogue is the more potent antibacterial and cytotoxic compound of the pair, whereas the antimalarial profile is essentially unchanged by the presence or absence of that methyl group. The authors note that this finding further supports their previous conclusion that the side-chain carboxamide group is essential for antimalarial activity, and they conclude that C-5 O-methylation contributes to the antibacterial activity and cytotoxicity of KQ analogues while leaving antimalarial activity largely untouched. The work illustrates how targeted disruption of a single tailoring enzyme within a biosynthetic gene cluster can generate a structural variant that sharpens the structure-activity picture, and the researchers suggest that further studies of KQ analogues produced through targeted modification of biosynthetic genes may provide additional information on their structure-activity relationships.

Subject of Research: Isolation and characterization of the new kinanthraquinone analogue 5-O-desmethyl kinanthraquinone B from an O-methyltransferase gene disruptant

Article Title: Isolation and characterization of 5-O-desmethyl kinanthraquinone B generated through O-methyltransferase gene disruption

Article References: Sakai, K., Futamura, Y., Nogawa, T., Aono, H., Osada, H., & Takahashi, S. (2026). Isolation and characterization of 5-O-desmethyl kinanthraquinone B generated through O-methyltransferase gene disruption. The Journal of Antibiotics. https://doi.org/10.1038/s41429-026-00960-w

Image Credits: AI Generated

DOI: 10.1038/s41429-026-00960-w

Keywords: natural products, anthraquinones, kinanthraquinone, O-methyltransferase, gene disruption, antibacterial activity, antimalarial activity, cytotoxicity, structure elucidation, biosynthetic gene cluster, Streptomyces lividans, drug discovery