Tryptophan-Derived Carbon Dots Detect Last-Resort Antibiotic in Cells and Water

Meropenem is one of medicine’s last lines of defense against drug-resistant bacterial infections, and knowing exactly how much of it is present in a patient’s blood, a pharmaceutical vial, or a hospital’s wastewater matters enormously for both treatment and environmental safety. A team of chemists and laboratory medicine researchers now reports a new fluorescent probe that can measure this carbapenem antibiotic with nanomolar sensitivity, using carbon dots engineered from two inexpensive starting materials: the amino acid L-tryptophan and thiourea. The work, published in Frontiers in Chemistry, describes nitrogen and sulfur co-doped carbon dots, or N,S-CDs, that light up blue under ultraviolet excitation and dim predictably when meropenem is present.

The synthesis itself is strikingly simple. The researchers dissolved 0.05 grams of L-tryptophan and 0.02 grams of thiourea in 50 milliliters of deionized water, sealed the solution in a Teflon-lined stainless-steel autoclave, and heated it at 180 degrees Celsius for six hours. After centrifugation, filtration through a 0.2-micrometer membrane, and 48 hours of dialysis, the resulting brown-yellow solution contained the finished carbon dots. The team systematically optimized the reaction conditions, testing hydrothermal temperatures from 160 to 190 degrees Celsius and reaction times from 5 to 8 hours. At 180 degrees and 6 hours, the particles remained tiny, spherical, and well dispersed; pushing the temperature to 190 degrees caused visible clustering into larger, irregular aggregates, which the authors attribute to aggregation-caused quenching that erodes the fluorescence output.

Characterization confirmed the material’s structure and composition. High-resolution transmission electron microscopy showed well-dispersed spherical nanoparticles with diameters ranging from 2.2 to 8.2 nanometers, an average of 4.56 plus or minus 1.11 nanometers, and the largest population centered at 4.8 nanometers. Powder X-ray diffraction revealed a single broad hump centered at 2-theta of 26.2 degrees, the signature of an amorphous, disordered graphitic carbon framework rather than a crystalline one. Fourier transform infrared spectroscopy identified a rich surface chemistry: hydroxyl and amine groups around 3,400 per centimeter, overlapping carbonyl and imine stretches between 1,600 and 1,700 per centimeter, carbon-nitrogen bands at 1,380 to 1,450 per centimeter, and fingerprint-region features confirming that sulfur from the thiourea had been successfully incorporated into the carbon framework.

The optical behavior is what makes the probe useful. Excited at 350 nanometers, the carbon dots emit a single, symmetric band centered at 457 nanometers. When meropenem is added, that emission fades in a systematic, concentration-dependent way. In ultraviolet-visible titrations, adding meropenem in 0.2-micromolar steps decreased the absorption bands at 230 and 350 nanometers while a new peak grew at 298 nanometers, corresponding to the antibiotic’s own intrinsic absorption. Crucially, a well-defined isosbestic point appeared at 280 nanometers, evidence of a direct, specific ground-state interaction between the probe and the drug rather than random, nonspecific contact. In fluorescence titrations, each 30-nanomolar addition of meropenem produced a progressive drop in intensity across a linear range extending from zero to 390 nanomolar, with a correlation coefficient of 0.990 and a calculated limit of detection of 6.8 nanomolar.

To understand why the fluorescence fades, the team turned to Stern-Volmer analysis, binding constant calculations, and time-resolved fluorescence measurements. The Stern-Volmer plot curved upward at higher meropenem concentrations, indicating that both static and dynamic quenching mechanisms operate simultaneously rather than a single pathway. From the initial linear region, the Stern-Volmer quenching constant came out at 3.97 times ten to the sixth per molar, and a modified double-logarithmic analysis yielded a binding constant of 9.07 times ten to the sixth per molar with a binding stoichiometry of 1.81, suggesting that more than one site on the heterogeneous dot surface accommodates each meropenem molecule. Time-resolved measurements showed the average fluorescence lifetime dropping from 6.45 to 5.68 nanoseconds in the presence of the drug. Because static quenching by ground-state complex formation leaves the lifetime of the remaining free fluorophore unchanged, while dynamic collisional quenching shortens it, the lifetime reduction points to a dynamic component layered on top of the static complexation already implied by the isosbestic point. The authors also note a possible partial inner-filter effect from the overlap between the dots’ excitation band and meropenem’s 298-nanometer absorption, which they evaluated qualitatively without quantitative correction.

Selectivity is the make-or-break property for any sensor intended for real biological fluids, and here the probe performed notably well. The researchers challenged it with inorganic cations including copper, iron, lead, mercury, cadmium, calcium, magnesium, and zinc; anions including chloride, bromide, nitrate, nitrite, sulfate, acetate, and thiocyanate; amino acids such as glycine, cysteine, lysine, and arginine; sugars including glucose and lactose; and a panel of clinically co-administered antibiotics spanning levofloxacin, tetracycline, doxycycline, ampicillin, ciprofloxacin, chloramphenicol, gentamicin, and vancomycin. None of these interfered with the meropenem signal. Most strikingly, three closely related carbapenems, doripenem, imipenem, and ertapenem, which share meropenem’s beta-lactam core and thioether-containing side chain, produced no appreciable quenching on their own and did not disturb the meropenem response when mixed with it. The interferents were tested at 0.5 millimolar against 0.4 micromolar meropenem, a 1,250-fold excess, and the response held.

Environmental conditions also proved manageable. The probe’s fluorescence was pH-dependent, peaking at pH 7 in the absence of the drug, and the largest difference between quenched and unquenched signals also occurred at pH 7, which suits work in physiological matrices such as serum and urine. The quenching response reached a stable plateau within about 5 minutes of meropenem addition and remained steady for up to 60 minutes, and the response stayed distinct across temperatures from 15 to 40 degrees Celsius, with room temperature chosen for subsequent experiments.

Before attempting measurements inside living cells, the team assessed biocompatibility using an MTT assay on PC3 cells across concentrations from zero to 10 milligrams per milliliter of carbon dots, with five replicates per dose. Viability stayed above 95 percent up to 0.5 milligrams per milliliter and declined in a dose-dependent manner at higher doses, reaching approximately 81 percent at the highest tested concentration of 10 milligrams per milliliter, still above the 80 percent threshold commonly accepted for biocompatibility in cell-based experiments. Confocal laser scanning microscopy then showed that PC3 cells incubated with the dots displayed intense blue fluorescence localized mainly in the cytoplasm. Treating those labeled cells with meropenem dimmed the intracellular signal in a concentration-dependent fashion: 1 micromolar of the drug produced a clear decrease, and 3 micromolar darkened the cells further, mirroring the solution-phase quenching behavior and demonstrating that the sensing chemistry survives inside a living cell.

The practical test came in five real sample matrices: a commercial meropenem injection sold as Kefen, human serum, human urine, tap water, and pond water. Samples were spiked with meropenem at 20, 50, and 100 nanomolar, all within the linear range, and recoveries across every matrix and every spiking level fell between 98 and 102 percent. Serum and urine were diluted 20-fold and filtered or centrifuged before analysis, and all measurements were performed in five replicates. The fluorimetric readings were cross-validated against a reference HPLC method using an Agilent Zorbax Eclipse Plus C18 column with acetonitrile and phosphate buffer as the mobile phase, and statistical comparison by Student’s t-test and F-test found no significant difference between the two methods at the 95 percent confidence level.

The authors position this as the first reported carbon-dot probe for meropenem, filling what they describe as a gap in a literature that already includes carbon-dot sensors for ciprofloxacin and tetracycline. Compared with previously published spectrofluorimetric methods for meropenem, including a gold nanoparticle-enhanced europium chelate approach with a 2.2-nanomolar detection limit and an NBD-Cl derivatization method at 5.6 nanomolar, the new probe’s 6.8-nanomolar limit is in the same sensitivity class while offering a broader linear range than some alternatives. The motivation is substantial: meropenem is a last-resort drug for severe infections such as sepsis, pneumonia, and meningitis, suboptimal dosing risks treatment failure and resistance, and excessive concentrations carry dose-dependent toxicity including seizures and renal effects in critically ill patients. Nearly 90 percent of the administered drug is excreted unchanged, so hospital effluents carry high concentrations into sewage systems, where ambient bacteria exposed to residues can develop carbapenem resistance. The authors caution that future work must include structural validation of the dot-drug interaction, testing in complex wastewater matrices and additional cell models, and development toward a portable point-of-care platform before the probe can move from the bench toward routine clinical and environmental monitoring.

Subject of Research: A fluorescent carbon-dot nanoprobe for detecting the antibiotic meropenem in biological and environmental samples.

Article Title: N,S-doped carbon dots derived from tryptophan for sensitive and selective detection of meropenem in live cells, human urine, and water samples

Article References: Hu, X., Zou, Y., Thangamani, K., Kayalvizhi, R., Anandhan, A., Damodaran, T., Rajalakshmi, K., Zhu, D., Muthusamy, S., & Chen, Y. (2026). N,S-doped carbon dots derived from tryptophan for sensitive and selective detection of meropenem in live cells, human urine, and water samples. Frontiers in Chemistry, 14, 1924756. https://doi.org/10.3389/fchem.2026.1924756

Image Credits: AI Generated

DOI: 10.3389/fchem.2026.1924756

Keywords: meropenem, carbon dots, fluorescent sensor, antibiotic detection, therapeutic drug monitoring, carbapenem resistance, water analysis, nanotechnology, analytical chemistry, PC3 cells, HPLC validation, tryptophan