Ratiometric rapid distinction of two structurally similar fluoroquinolone antibiotics by a Tb/Eu hydrogel

Norfloxacin and ofloxacin are two frequently prescribed second-generation fluoroquinolone antibiotics with an identical 4-quinolone chromophore and hence, are difficult to distinguish by conventional methods (UV or fluorescence). We have designed a Tb³⁺/Eu³⁺/cholate cocktail that enabled us to differentiate these two drugs and rapidly measure their concentrations when present together. Additionally, a Tb³⁺-cholate gel-based paper sensor was developed to detect and quantify them in a single drug containing system with a limit of detection (LOD) well below 100 nM.

Rapid distinction of structurally similar fluoroquinolone antibiotics norfloxacin and ofloxacin, and their quantification in a mixture, were achieved using a supramolecular Tb/Eu gel cocktail.

Fluoroquinolone antibiotics belong to one of the most important classes of human and veterinary medicines owing to their broad-spectrum activity and effective oral absorption. Norfloxacin and ofloxacin (Fig. 1) are among the most prescribed second generation fluoroquinolone antibiotics^1,2 used for the treatment of respiratory and urinary tract infections, ocular and skin infections, pelvic inflammatory disease, gonococcal urethritis, infectious diarrhoea, etc. However, potential adverse reactions such as tendinitis, tendon rupture, arthralgia, myalgia, peripheral neuropathy, hematuria, and central nervous system effects due to their excess intake have been well documented in the literature.³ Extensive use and misuse of the antibiotics as veterinary medicines have led to their appearance in milk,^4,5 chicken⁶ and fish⁷ imposing adverse side effects for the consumer. Contamination of municipal wastewater, surface water and even groundwater results from their extensive usage, and from their excretion from human and animals unchanged.⁸ All these can cause the propagation of antibiotic-resistant micro-organisms⁹ leading to a disturbed ecosystem. Therefore, the development of a robust, rapid and straightforward analytical protocol for screening of Norfloxacin (NFLX) and Ofloxacin (OFLX) residues in biological fluids, sewage water, edible tissues and foodstuff is highly desirable for the control of their appropriate dosage, maintaining their pharmacokinetics and avoid environmental crisis.¹⁰ Various analytical methods for this purpose have been reported in the literature including spectrophotometry,¹¹ spectrofluorimetry,¹² voltammetry,^13,14 capillary electrophoresis,¹⁵ chemiluminescence,^16,17 and high performance liquid chromatography (HPLC).¹⁸ Several fluorescent probes have utilized molecular imprinting chemiluminescence,¹⁹ Neodymium-modified micellar media,²⁰ carbon dots etc.^21,22 Although fluorescence-based reporters provide greater intrinsic sensitivity, many of them are, however, incompatible with biological samples because of complications arising from background scattering and autofluorescence from endogenous components.²³ Trivalent lanthanide ions have long radiative lifetimes allowing background-free measurement using time-gating. They also have sharp, well-defined long-wavelength emission fingerprints²⁴ ideal for biological applications and can therefore serve as a strong sensing platform. They can be sensitized by proximate chromophores (‘‘antennae’’) and may act as attractive fluorogenic (turn on) and fluorolytic (turn-off) bio-probes. This group has extensively investigated photoluminescent lanthanide cholate supramolecular hydrogels^25–32 by utilizing their in-built hydrophobic pockets to encapsulate small organic ‘antennas’. The self-assembly brings the antenna in close proximity to the acceptor lanthanide leading to energy transfer to the lanthanide ion. This strategy thus eliminates the need for the synthesis of antenna-linker-ligand type of probes to achieve lanthanide sensitization. A variety of lanthanide-luminescence based fluoroquinolone-assays^17,33–38 have been reported in the literature. However, tedious synthesis of the probe, complex sample treatments, non-aqueous detection media, strict maintenance of pH have restricted the practical applicability of many of them.^34–37 We have found that when doped in micromolar concentrations in Tb-Ch (5/15 mM) or Eu-Ch (5/15 mM) gels, both Norfloxacin and Ofloxacin significantly enhanced the luminescence intensities. Based on this finding we explored cholate-hydrogel based detection and quantification of these two fluoroquinolones. Initial experiments were performed in Tb (5 mM) cholate (15 mM) hydrogel by doping it with increasing concentrations of norfloxacin (NFLX). The emission intensity of Tb³⁺ showed a linear relationship with concentration (Fig. S1 in ESI^†). Even 4 μM of NFLX resulted in 30-fold enhancement of sensitization of Tb³⁺ luminescence. Atomic force microscopy showed that the fibrous network of Tb³⁺-cholate gel remained unaltered (Fig. S2 in ESI^†) when NFLX was doped in micromolar concentrations, proving the robustness and stability of the sensing gel system. Fluorescence microscopic image of semi-dried NFLX doped Tb-cholate gel (Fig. S2c in ESI^†) showed uniform green luminescent gel fibres, indicating that the NFLX was distributed and embedded on the hydrophobic fibres, promoting the energy transfer process making the fibres green luminescent.Open in a separate windowFig. 1Structures of Norfloxacin and Ofloxacin.In order to simplify the sensing method, we chose a paper-based system. Such sensors^39–41 are affordable, biodegradable, sensitive, specific, user-friendly, rapid, robust, and deliverable to end-users. For this purpose, Tb-cholate (5/15 mM) gel doped with NFLX (4 μM) was prepared, sonicated for 5–6 s to reduce the viscosity, and then applied (20 μL) on a 3 mm paper disc (Fig. 2). SEM images showed dried flakes of gel on the cellulose fibre network (Fig. S3 in ESI^†). The thickness of the gel layer coated on the paper surface was about 8 μm, as determined by tilt-SEM (Fig. S3c in ESI^†). Time gated fluorescence (TRF) was measured on the coated discs using a plate reader. After 5 min of drying, a significantly higher Tb³⁺ luminescence output was observed, compared to that measured in wet gels (Fig. 3b, 8 vs. 7). This increase may be attributed to the efficient wicking of water molecules from the gel in the hydrophilic cellulose frameworks of paper which in turn helps reduce the OH-mediated luminescence quenching.⁴² Increased sensitization on gel-coated paper discs was observed for both NFLX and OFLX. We also confirmed that the extent of sensitization of Tb³⁺ remained constant as a function of time (Fig. S4 in ESI^†). The limit of detection (LOD) for NFLX and OFLX were measured based on Tb³⁺-luminescence output as a function of the drug concentration and were found to be 13.6 nM (Fig. S5a in ESI^†) and 67 nM (Fig. S5b in ESI^†), respectively. The paper-based method was used to quantify the NFLX spiked in milk samples (cow raw milk and a commercial homogenized toned milk) and in human blood serum. The LOD values measured were 68 nM, 100 nM and 100 nM, respectively (Fig. S12 and S13 in SI^†). The time-gated detection method eliminated any interference from the matrix (serum or milk). This method is therefore simple, rapid, sensitive, autofluorescence and background emission free, and may thus serve as a practical tool for analytical screening.Open in a separate windowFig. 2Design of paper-based sensors (paper disc cutting, placing them on plate groove, gel dropcasting).Open in a separate windowFig. 3Tb³⁺ -luminescence (λ_em 545 nm, λ_ex 330 nm) enhancement by NFLX (4 μM) (Ac, Ch, D: acetate solution, cholate matrix and coated paper disc, respectively).Distinguishing two structurally similar compounds of the same class is a fundamental objective of analytical research.Dennany et al. recently reported a [Ru(bpy)₃]²⁺ based ECL sensor⁴³ to distinguish Atropine and Scopolamine, two similar alkaloids. Encouraged by this report, we focused on the differentiation of norfloxacin and ofloxacin which have comparable 4-quinolone chromophores with similar functional groups attached (Fig. 1). With both drugs, the luminescence enhancement for Tb³⁺ was greater than that for Eu³⁺ (Fig. S8 & S9 in ESI^†). Therefore, single lanthanide (Tb³⁺ or Eu³⁺) derived gels cannot differentiate them. For this purpose, heterobimetallic ensemble strategy was explored by adopting the ratiometric detection technique.^44,45 In order to identify optimum conditions, a set of five cholate gels samples containing varying ratios of Tb³⁺ & Eu³⁺ (maintaining total concentration of 5 mM) was prepared. This set was doped with 1 μM NFLX. An identical set was prepared with 1 μM OFLX as the dopant. After drop casting the gels on paper discs, luminescence intensities were quantified on a plate reader (Fig. 4a and b). From these data, the sensitization bias I_Tb/I_Eu, as measured by I₅₄₅/I₆₁₇ (λ_ex 330 nm), was calculated and plotted against the concentration of Tb³⁺ in the mixed gel. Tb³⁺/Eu³⁺ cocktail with [Tb³⁺] ≥ 2.5 mM doped with NFLX showed sensitization bias (I₅₄₅/I₆₁₇) values greater than 1, whereas with OFLX the bias was always less than 1 (Fig. 4c). Based on these observations we chose Tb³⁺/Eu³⁺/cholate (4.5 mM/0.5 mM/15 mM) cocktail as the probe to differentiate these two drugs over a wider concentration range (Fig. S15 in ESI^†). Another set of measurements were made at low μM concentrations of the drugs (Fig. S14 in ESI^†). The Terbium sensitization bias (I_Tb/I_Eu) was calculated and plotted against the concentration of FLXs and found to be greater than 1 for NFLX but less than 1 for OFLX (Fig. S14c in ESI^†). In the Tb³⁺/Eu³⁺ mixed cholate hydrogel matrix three concurrent energy transfer processes can occur–energy transfer from sensitizer (i) to Tb³⁺, (ii) to Eu³⁺, and (iii) energy transfer from Tb³⁺ to Eu³⁺. While one of the emission bands of Tb³⁺ has overlap with the most intense 617 nm peak of Eu³⁺, the Eu³⁺ emission peak at 690 nm is solely contributed by Eu³⁺. The 690 nm emission intensity in 15 μM NFLX doped mixed gels was measured and found to linearly increase with the Tb³⁺ concentration (even though Eu³⁺ concentration is decreasing) indicating Tb³⁺– Eu³⁺ energy transfer (Fig. S19 in ESI^†). Significant reduction of Tb³⁺-luminescence lifetime from 1.53 ms in NFLX doped Tb³⁺-Ch (4.5 mM/15 mM) gel to 0.55 ms in NFLX doped Tb³⁺/Eu³⁺-Ch (4.5 mM/0.5 mM/15 mM) gel provided additional evidence. In the Tb³⁺/Eu³⁺/cholate (4.5 mM/0.5 mM/15 mM) cocktail, the three concurrent energy transfer processes are optimized to show Tb3+/Eu3+ emission intensity ratio exactly reversed for the two drugs (Fig. S14, S16 in ESI^†). The optimized cocktail was subsequently used to determine the ratio of the two drugs in a mixture. The drug mixtures (total concentration 20 μM) were doped in the cocktail and the photophysical behavior was investigated. The excitation spectra corresponding to Eu³⁺ 617 nm emission showed a unique pattern depending on the ratio of NFLX and OFLX in the mixture. The intensity (Fig. 5a) at 280 nm reflects NFLX excitation, whereas the intensity at 300 nm reflects OFLX excitation. Based on these findings, the Eu³⁺ emission intensity at 617 nm from the drug mixtures at two different excitation wavelengths 280 nm and 300 nm were evaluated. Ratiometric response I₆₁₇ (λ_ex 280)/I₆₁₇(λ_ex 300) was found to linearly increase with the ratio of NFLX concentration to the total concentrations of both the drugs in the mixture (Fig. 5c). To validate the need for using the cocktail in this study, the excitation spectra corresponding to Eu³⁺ at 617 nm in (NFLX + OFLX) doped Eu³⁺-Ch gel (no Tb) was recorded in which no such NFLX & OFLX ratio-dependent pattern was observed (Fig. S17 in ESI^†). For further exploration of the cocktail, Terbium emission sensitization bias was calculated and found to be directly proportional to the ratio of NFLX concentration to the total concentrations of both the drugs in the mixture (Fig. 6b). This protocol can thus be used as a platform to determine the ratio and individual concentration of the two drugs when present in a mixture.Open in a separate windowFig. 4Emission intensities from Tb³⁺ at 545 nm and Eu³⁺ at 617 nm (λex 330 nm of 1 μM (a) NFLX and (b) OFLX doped mixed cholate {[Tb³⁺] + [Eu³⁺] = 5 mM} gel coated discs for varying ratios of Tb³⁺& Eu³⁺ (c) Terbium sensitization bias I₅₄₅/I₆₁₇ (λ_ex 330 nm) as a function of Tb³⁺ concentration.Open in a separate windowFig. 5(a) Excitation spectra for Eu³⁺ emission (λ_em 617 nm) with respect to the ratio of two drugs (b) Intensity of Eu³⁺ emission at 617 nm for two excitation wavelengths (280 nm, 300 nm) in Tb³⁺/Eu³⁺/cholate (4.5 mM/0.5 mM/15 mM) gel coated discs with increasing concentrations of NFLX in a mixture of OFLX & NFLX, (c) intensity ratio plotted against the concentration of NFLX in the mixture.Open in a separate windowFig. 6(a) Tb³⁺ emission at 545 nm and Eu³⁺ emission 617 nm (b) Tb³⁺ sensitization bias in Tb³⁺/Eu³⁺/cholate (4.5 mM/0.5 mM/15 mM) gel coated discs with the ratio of NFLX concentration to the total drug concentration in the mixture (λ_ex 330 nm).In conclusion, a hetero bimetallic ensemble Tb³⁺/Eu³⁺/cholate was designed for the differentiation of structurally similar NFLX and OFLX. Till date, only physical separation by HPLC is known to quantify norfloxacin and ofloxacin concentrations in a mixture. Additionally, a paper-based sensor has also been developed for the detection and quantification of them in single drug containing system with a lower nanomolar range. The method allowed autofluorescence free detection with biological samples without the need of deproteinization and sample processing. Nonetheless, our preliminary investigations employed to utilize ratiometric bis-lanthanide strategy will serve as the foundation for the fundamental need to differentiate structurally similar molecules. Further studies to utilize this supramolecular paper-based methodology in real diagnostic fields is underway in our lab.