The researchers designed ten hybrids, designated LS-1 to LS-10, by joining a theophylline scaffold to a 1,2,3-triazole unit and an acetamide group bearing different substituents on its terminal phenyl ring. The strategy combined the purine core of theophylline with a triazole ring intended to provide a stable linker capable of hydrogen-bonding interactions. The central synthesis used copper(I)-catalyzed azide–alkyne cycloaddition, commonly described as a click reaction, to form the triazole ring with high regioselectivity. Ultrasound irradiation at 25 kHz accelerated the final coupling in an equal-volume mixture of dimethylformamide, tert-butanol and water. The design drew on the established use of theophylline as a structurally rigid, synthetically accessible medicinal-chemistry scaffold and on the capacity of 1,2,3-triazoles to act as stable heteroaromatic linkers in bioactive molecules.

The synthetic route began with N-alkylation of theophylline using propargyl bromide in dimethylformamide, producing a terminal alkyne. In parallel, substituted anilines were acylated with chloroacetyl chloride to make α-chloroacetamides, which were then converted to the corresponding azides with sodium azide. The azide intermediates were obtained in yields of 80–90% before being coupled to the theophylline-derived alkyne. Copper sulfate pentahydrate and sodium ascorbate were used to generate the catalytic Cu(I) conditions. Each final product was isolated by filtration and recrystallized from ethanol, with the study reporting yields of 90–96% for the 10-member series. The click step was performed under sonochemical irradiation for 55 minutes, and the products were designated 8a–j during the chemical discussion before being identified as LS-1 through LS-10.

For a representative reaction, the researchers calculated an atom economy of 89.8% and an E-factor of 0.335, with the latter defined as the mass of waste relative to isolated product and excluding recoverable solvent. They compared the 55-minute sonochemical reaction with conventional triazole-synthesis methods described in a review, reporting that those methods required 8–24 hours and produced yields of 50–78%. These comparisons suggest that ultrasound shortened the reported reaction substantially, but the green-chemistry assessment was based on a model reaction and calculated metrics rather than a full life-cycle analysis. The synthesis also used a mixed solvent system containing dimethylformamide, tert-butanol and water, so the procedure should not be described as solvent-free. The authors presented the method as operationally simple, with straightforward work-up and reduced waste relative to the representative conventional methods.

The structures of all ten products were confirmed using Fourier-transform infrared spectroscopy, proton and carbon-13 nuclear magnetic resonance spectroscopy, and mass spectrometry. The infrared spectra showed amide N–H stretches between 3432 and 3742 inverse centimeters, carbonyl absorptions around 1705–1710 inverse centimeters, and bands attributed to triazole ring vibrations from 1644 to 1658 inverse centimeters. Halogenated derivatives also displayed additional absorptions in the 686–855 inverse-centimeter fingerprint region. In proton NMR, the triazole proton appeared near 8.06 to 8.52 parts per million, while the amide proton occurred farther downfield at 9.66 to 10.70 parts per million. Methylene bridge signals appeared at 5.21 to 5.96 parts per million, and the theophylline N-methyl groups were observed at 3.13 to 3.43 parts per million. Carbon-13 NMR and mass-spectral molecular-ion peaks were also consistent with the assigned molecular structures.

Antibacterial testing used agar-well diffusion, minimum inhibitory concentration and minimum bactericidal concentration assays against three Gram-positive organisms—Staphylococcus aureus, Bacillus cereus and Corynebacterium rubrum—and three Gram-negative organisms—Pseudomonas aeruginosa, Salmonella typhimurium and Escherichia coli. The experiments were performed in triplicate, with tetracycline as the reference antibiotic and dimethylsulfoxide as a negative control. In the agar-diffusion assay, 100 microliters of a 20 milligrams-per-milliliter solution of each compound was placed in wells in Mueller–Hinton agar seeded with 200 microliters of a 1 × 10^8 colony-forming-unit-per-milliliter inoculum. Plates were incubated at 37 degrees Celsius for 24 hours, and inhibition-zone diameters were recorded in millimeters. Among the hybrid compounds, LS-5 produced a 32-millimeter inhibition zone against S. aureus, while LS-10 reached 32.5 millimeters against P. aeruginosa. LS-6 and LS-9 each produced 32-millimeter zones against P. aeruginosa.

Minimum inhibitory concentration measurements identified the same general pattern. LS-5, LS-6, LS-9 and LS-10 reached the lowest reported value, 46.9 micrograms per milliliter, against S. aureus or P. aeruginosa, whereas most other compounds had values between 93.75 and 750 micrograms per milliliter. The minimum bactericidal concentrations for the leading compounds were 375 micrograms per milliliter, higher than their inhibitory concentrations and consistent with predominantly bacteriostatic activity in the reported assays. The researchers characterized these values as moderate relative to standard antibiotics, and noted that large agar inhibition zones can partly reflect favorable diffusion through the medium rather than intrinsic potency alone. Activity was also strain-dependent, with reduced or absent effects against E. coli and C. rubrum for some derivatives. The assays therefore measured laboratory activity against a defined bacterial panel, not treatment outcomes in infected people.

Analysis of the substitution pattern indicated that the common theophylline–triazole–acetamide core contributed measurable activity, but that modifications to the terminal aromatic ring altered potency. The unsubstituted analogue LS-1 produced a 30-millimeter zone against S. aureus with an MIC of 375 micrograms per milliliter and a 26-millimeter zone against P. aeruginosa with an MIC of 750 micrograms per milliliter. The 2,4-difluoro compound LS-5 and the 3-chloro-4-fluoro compound LS-6 were among the most active, as were the brominated LS-9 and LS-10. LS-8, which carried a para-methoxy substituent, produced a 31.5-millimeter zone against S. aureus and an MIC of 93.75 micrograms per milliliter. These comparisons support an association between halogen substitution and stronger activity in the tested panel. They do not establish that halogenation is universally beneficial, because activity also varied by bacterial strain and substitution position.

To investigate possible target interactions, the team docked the compounds into two DNA gyrase B structures: PDB 3G75 from S. aureus and PDB 5L3J from E. coli. GyrB is an ATP-dependent component of bacterial DNA gyrase, and the selected structures contained co-crystallized inhibitors within the ATP-binding region. All ten compounds had predicted affinities from −6.8 to −9.0 kilocalories per mole for the S. aureus model and −7.2 to −8.2 kilocalories per mole for the E. coli model. LS-4 had the best score against 3G75 at −9.0 kilocalories per mole, compared with −8.0 for the reference ligand novobiocin. LS-6 scored −8.2 kilocalories per mole against 5L3J, slightly better than novobiocin’s −8.1 in that model. Redocking produced root-mean-square deviations of 1.36 and 1.42 angstroms for the two receptors. Docking scores and interaction patterns provide a computational rationale for study, but they do not demonstrate enzyme inhibition in cells or animals.

Drug-likeness predictions generated with SwissADME suggested that the compounds had broadly acceptable physicochemical properties. Their topological polar surface areas ranged from 121.63 to 167.45 square angstroms, predicted octanol–water partition coefficients ranged from 1.95 to 3.24, and calculated log solubility values ranged from −2.37 to −3.59. The models predicted six to eight hydrogen-bond acceptors and one donor per molecule, high gastrointestinal absorption for most compounds, limited blood–brain barrier penetration, and possible P-glycoprotein substrate behavior. A bioavailability score of 0.55 was reported for all derivatives, while Ames predictions classified them as non-mutagenic. These are model outputs rather than measurements from pharmacokinetic, animal or human studies. Density functional theory calculations at the B3LYP/6-31G level assessed the compounds’ electronic structures in the gas phase. HOMO energies ranged from −6.307 to −5.677 electronvolts and LUMO energies from −3.042 to −0.732 electronvolts, producing HOMO–LUMO gaps of 3.265 to 4.956 electronvolts. LS-7 had the smallest calculated gap and the greatest predicted electronic reactivity within the series, while LS-1 and LS-2 had the largest gaps. The authors did not include continuum solvation models that could better represent a biological environment. Taken together, the synthesis, bacterial assays, docking and in silico profiling identify the hybrid series as early chemical leads, with the reported evidence limited to laboratory antibacterial testing and computational analysis.

Subject of Research: Synthesis and computational evaluation of theophylline–triazole acetamide hybrids as antimicrobial chemical leads

Article Title: Green ultrasound assisted synthesis of theophylline triazole acetamide hybrids and their antimicrobial and computational evaluation

Article References: Khoyanee, A., Shah, P., Vyas, A., Rakholiya, K., Kaneria, M., & Khunt, R. (2026). Green ultrasound assisted synthesis of theophylline triazole acetamide hybrids and their antimicrobial and computational evaluation. Discover Chemistry, 3(1), Article 584. https://doi.org/10.1007/s44371-026-01039-z

Image Credits: AI Generated

DOI: 10.1007/s44371-026-01039-z

Keywords: theophylline hybrids, ultrasound synthesis, click chemistry, antibacterial screening, DNA gyrase B, molecular docking, drug-likeness, density functional theory, green chemistry, Green, ultrasound, assisted