Enzymes From Gut and Soil Bacteria Strip Toxic Dyes From Wastewater at Record Concentrations

Textile factories around the world discharge an estimated 280,000 tons of dye waste into the environment every year, turning rivers into toxic, colored streams that block sunlight, poison aquatic life, and carry potentially carcinogenic compounds up the food chain. Now, a research team at Vellore Institute of Technology in India has shown that two bacterial enzymes can strip some of the most stubborn synthetic dyes from water at concentrations far higher than most previous treatments could handle — and the treated water is dramatically less toxic to plants. The study, published in Frontiers in Chemical Engineering, combined computer simulations with laboratory experiments to identify and validate the enzymes, offering what the researchers describe as a sustainable solution for decolorizing, degrading, and detoxifying highly concentrated dye pollutants.

The enzymes at the heart of the work are azoreductases, extracted from two bacteria isolated from dye-contaminated samples: Pseudomonas aeruginosa VITKSS5 and Klebsiella pneumoniae VITKSS6. Azoreductases are flavin-dependent enzymes that catalyze the reductive cleavage of the azo bond — the –N=N– linkage that gives azo dyes their color and their chemical stability. Azo dyes make up roughly 60 to 70 percent of all commercial dyes, and their stability is precisely what makes them persistent environmental pollutants. When an azoreductase breaks the azo bond, the dye loses its chromophore and the color disappears. But the enzymes’ usefulness has traditionally been thought to end there, since they are named for their azo-bond activity and most studies have tested them only on azo dyes at low to moderate concentrations, generally below 800 milligrams per liter.

To find out whether these enzymes could do more, the team turned to molecular docking, a computational technique that predicts how strongly a small molecule binds to a protein’s active site. Using AutoDock Vina, the researchers docked ten reactive textile dyes — collected from dyeing units in Tiruppur, Tamil Nadu — against two reference crystallographic azoreductase structures: AzoP from P. aeruginosa (Protein Data Bank entry 2v9c) and AzoK from K. pneumoniae (entry 7n2w). The dye structures were retrieved from PubChem as canonical SMILES strings, converted into three-dimensional models with UCSF Chimera, energy-minimized, and fitted with Gasteiger charges before docking. The experimentally resolved flavin mononucleotide cofactor was retained in each receptor structure, since it is essential for the enzyme’s catalytic cycle. Redocking the co-crystallized ligands validated the protocol, yielding root-mean-square-deviation values of 0.333 angstroms for AzoP and 0.037 angstroms for AzoK — excellent agreement between predicted and experimental poses.

The docking results were striking. Four dyes emerged with the most favorable binding energies: Acid Red 131 (AR131), Reactive Red 195 (RR195), Reactive Blue 19 (RB19), and Reactive Blue 160 (RB160). AzoK showed its strongest affinity for RB19 at −11.2 kilocalories per mole, followed by RR195 at −10.9 and AR131 at −10.8. AzoP preferred RR195 at −11.3, AR131 at −11.1, and RB160 at −10.9. These values are comparable to or better than those reported for other azoreductases in recent docking studies, such as Pseudomonas putida azoreductase, which scored −10.7 kilocalories per mole for Reactive Red 120. The interaction maps revealed dense networks of conventional hydrogen bonds, C–H bonds, π–alkyl, π–sulfur, and π–π T-shaped interactions with residues such as ARG B:12, ASN B:99, HIS B:149, TRP B:124, and TYR A:131. Hydrogen bonding, the researchers note, plays a key role in substrate alignment and in facilitating transitions between azo and hydrazone forms, ensuring efficient bond breakdown.

What makes these results particularly notable is that RB19 is an anthraquinone dye, not an azo dye. Anthraquinone dyes like Remazol Brilliant Blue R are prized in the textile industry for their structural stability and resistance to degradation, making their wastewater notoriously difficult to treat. Reports of azoreductases directly decolorizing RB19 are extremely limited, and the authors state that, based on the available literature, this study appears to be the first to demonstrate azoreductase-mediated decolorization of RB19 at a high concentration of 1,000 milligrams per liter. The researchers suggest the observed activity may involve NADH-dependent reduction of the dye’s quinone moiety rather than classical azo-bond cleavage, pointing to a broader substrate specificity that some bacterial azoreductases are known to exhibit toward quinone compounds, including anthraquinone, naphthoquinone, and benzoquinone derivatives.

To validate the computational predictions, the team extracted crude azoreductases from both bacterial strains, partially purified them through ammonium sulphate precipitation, dialysis, and gel filtration chromatography on Sephadex G-50, and analyzed the fractions by SDS-PAGE. Prominent protein bands appeared between approximately 26 and 50 kilodaltons in K. pneumoniae VITKSS6 and around 22 to 30 kilodaltons in P. aeruginosa VITKSS5 — sizes consistent with azoreductases reported from other bacteria, such as the 30-kilodalton enzyme from Pseudomonas sp. KF46. Enzyme activity assays showed that AzoK-VITKSS6 reached 18.46 units per milliliter against the azo dye Methyl Red in the presence of the NADH cofactor, while both enzymes showed identical activity of 7.273 units per milliliter against the anthraquinone dye RBBR. Enzymes produced under dye-free conditions generally showed higher activity than those produced with dye supplementation.

The decolorization experiments were conducted under microaerophilic, nutrient-limited conditions designed to resemble real textile effluent — static incubation at 37 degrees Celsius in minimal salt medium with only 0.5 percent glucose. Under these demanding conditions, the partially purified enzymes consistently outperformed whole bacterial cells and bacterial consortia. AzoK-VITKSS6 achieved 96.4 percent decolorization of AR131, 95.6 percent of RB19, and 92.1 percent of RR195 at 1,000 milligrams per liter, while AzoP-VITKSS5 reached 94.7 percent for AR131 and 93.4 percent for RR195 at the same concentration. Most remarkably, AzoP-VITKSS5 decolorized 88.3 percent of RB160 at 2,000 milligrams per liter — a dye load far exceeding what most previous studies have attempted. By comparison, the individual bacterial strains alone showed much lower and more variable efficiencies, generally below 55 percent except for RB19 and RR195, indicating that the azoreductase activity itself is the dominant driver of dye degradation.

Fourier transform infrared spectroscopy provided chemical evidence that the enzymes were not merely removing color but actually degrading the dye molecules. Spectra of enzyme-treated dye mixtures showed a marked reduction in N–H stretching at 3328 inverse centimeters, indicating degradation of secondary amines; decreased C–H stretching at 2948 and 2836 inverse centimeters, suggesting cleavage of alkane and aldehyde groups; and reduced C=H stretching at 1662 and O–H bending at 1444, indicating disruption of unsaturated bonds and oxidation of acidic functional groups. Changes in the 1450 to 1600 inverse centimeter region pointed directly to reduction of the azo –N=N– chromophore, while alterations in the fingerprint region suggested the formation of simpler hydrocarbons, ethers, and aromatic fragments. The enzyme-treated samples showed more pronounced functional group transformations than consortium-treated samples, indicating that degradation extends beyond chromophore cleavage to broader structural breakdown of the dye molecules.

Perhaps the most compelling evidence of detoxification came from phytotoxicity assays using two crop plants: Zea mays (maize, a monocot) and Vigna radiata (mung bean, a dicot). Untreated dyes severely inhibited seedling development — germination rates of V. radiata fell to between 10 and 30 percent with no root development at all, while Z. mays germination ranged from 30 to 70 percent. After enzymatic treatment, germination was restored to 85 to 95 percent, and root and shoot growth improved significantly. Root elongation of Z. mays exposed to enzyme-degraded AR131 increased by 20.7 percent, and shoot elongation of V. radiata exposed to enzyme-degraded RB19 increased by 43.9 percent. Overall, enzymatic treatment reduced phytotoxicity by 80.1 percent, from 79 percent in untreated dyes to 15.7 percent after treatment. The highest vigor index after treatment was observed in Z. mays with RB19 at 2023.5, compared with vigor indices as low as 120 for untreated RR195 in the same plant. Statistical analysis confirmed that enzyme-treated groups showed no significant differences from the tap-water control, indicating effective restoration of normal plant growth.

The authors emphasize that their enzymes achieved these results at dye concentrations substantially higher than most previously reported treatments, which typically operated at 100 to 500 milligrams per liter. The work supports several United Nations Sustainable Development Goals related to clean water, responsible consumption, and ecosystem protection. However, the researchers caution that field-scale validation has not yet been performed, and further testing with actual textile dyeing effluents is required before practical implementation. Future studies, they suggest, should focus on purification and molecular characterization of the azoreductases, followed by protein engineering and enzyme immobilization to improve stability and efficiency for industrial-scale effluent treatment. If those steps succeed, enzymes once thought to be limited to breaking a single chemical bond may become versatile biocatalysts capable of cleaning some of the most heavily polluted industrial wastewater on the planet.

Subject of Research: Azoreductase enzymes for bioremediation of azo and anthraquinone textile dyes

Article Title: Molecular docking and validation of azoreductases for bioremediation of high concentrations of azo and anthraquinone dyes- an integrated in silico-in vitro approach

Article References: Molecular docking and validation of azoreductases for bioremediation of high concentrations of azo and anthraquinone dyes- an integrated in silico-in vitro approach. (n.d.). https://doi.org/10.3389/fceng.2026.1955388

Image Credits: AI Generated

DOI: 10.3389/fceng.2026.1955388

Keywords: Molecular, docking, validation, azoreductases, bioremediation, high, concentrations, anthraquinone, dyes-, integrated, silico-in, vitro