1 Introduction

Tobramycin (TOB) is an aminoglycoside antibiotic produced by Streptomyces tenebrarius, with a broad spectrum activity against aerobic gram-negative bacteria, particularly pseudomonas aeruginosa, which makes it the antibiotic of choice in the treatment of pulmonary infections. The bactericidal activity of TOB is accomplished by inhibiting ribosomal function, leading to interruption in bacterial protein synthesis. TOB is used topically for treatment of eye infections, parenterally for treatment of serious bacterial infection, and also for local application in the oral cavity and stomach as part of selective decontamination of the digestive tract [1].

Due to lack of UV chromophore and the polar basic nature of its structure, it is difficult to develop a method for TOB determination. TOB has an official monograph in USP for HPLC/UV determination after derivatization with 2,4- dinitrofluorobenzene and tris (hydroxylmethyl) aminomethane reagent at 60 ± 2 °C for 50 ± 5 min [2]. Besides, several methods have been described in the literature to determine TOB; such as paper chromatography [3], colorimetry [4], spectrophotometry [5], gas chromatography [6], direct HPLC methods with UV [7, 8], refractive index [9], evaporative light scattering detection (ELSD) [10], mass spectrometry [11, 12], pulsed electrochemical [13] or pulsed amperometric detection [14,15,16,17,18]. The other HPLC methods involved UV detection after derivatization with 2,4- dinitrofluorobenzene [19,20,21], or 2,4,6-trinitrobenzenesulphonic acid reagent [22, 23] and fluorescence detection after derivatization with fluorescamine [1], fluorescein isothiocyanate [24], o-phthaldehyde [25,26,27,28,29], or 9-fluorenylmethyl chloroformate [30]. Anion exchange chromatography [16,17,18] and capillary electrophoresis [31,32,33] were also reported. These methods are either expensive, laborious, not sufficiently sensitive or time consuming. Consequently, the aim of this work is to develop a reliable HPLC derivatization method for highly sensitive determination of TOB using 2-naphthalenesulfonyl chloride as a derivatizing agent coupled with fluorescence detection. This method could enable detection of non-chromophoric compounds and pave the road for detection of other aliphatic drugs [34,35,36].

2 Experimental

2.1 Apparatus

All chromatographic separation was performed on an Agilent 1200 SL RRLC auto-sampler instrument containing a Bin pump SL, model G131213, an auto-sampler injector ALS SL, model G132913, and a fluorescence detector (FLD), model G1321A. The pH of the diluent solution was adjusted using HANNA pH-meter with double junction glass electrode, calibrated with standard buffers. Separation of the analyte was performed on Knauer-Eurospher 100–5 silica column- (300 mm × 4 mm i.d, 5 μm particle size). The column oven temperature was maintained at 40 °C. The samples were eluted with an isocratic mobile phase consisting of acetonitrile –1% phosphoric acid in water, (90:10, v/v), at a flow rate of 2 mL/min. The injection volume was 20 μL. Fluorescence detection was achieved with excitation at 300 nm and emission at 380 nm.

2.2 Materials

TOB reference standard was kindly provided by Orchidia Company, with a purity of 99.4% for the drug, based on the company analysis certificate. Acetonitrile and methanol (HPLC grade) were purchased from J.T.Baker, USA and Fischer Chemical, UK, respectively. 2-Naphthalenesulfonyl chloride 99% and triethylamine were from Sigma-Aldrich (USA). Deionized water was obtained from a Milli-Q water purification system (Millipore, USA).

2.3 Reagents and solutions

Stock standard 100 µg/mL solution of TOB was prepared in distilled water. Further dilutions were prepared in the same solvent to obtain working standard solutions. All solutions were stored at 2–8 °C and brought to room temperature before used. NSCl solution must be freshly prepared on the day of derivatization by dissolving 140 mg in 400 mL of acetonitrile. This amount is sufficient for 80 samples.

2.4 Reaction procedures and calibration curve

To 5 mL of TOB solution in distilled water placed in a screw-capped (10 mL) reaction vial, and mixed with 5 mL of 0.35 mg/mL NSCl in acetonitrile. Then 20 µL of triethylamine (TEA) was added, after that the vial was tightly closed and heated in a thermostatically controlled water bath at 80 ºC for 40 min. After cooling, the solution was evaporated under vacuum and the residue was re-dissolved in 0.5 mL acetonitrile. 20 µL of this solution was injected into HPLC/FL system. For blank preparation, 5 mL distilled water was placed instead of drug solution and subjected to the same procedure. Fluorescence intensities were measured immediately at λem. 380 nm, (λex. 300 nm) and calibration curve was constructed by plotting the emission intensity versus the final concentration of TOB. Figure 1 summarizes the procedures applied for dosage form analysis.

Fig. 1
figure 1

Procedures for pre-column derivatization and detection of TOB using HPLC/FL

Full size image

3 Results and discussion

Fluorescence detectors are very specific and selective among the others optical detectors. Typically, fluorescence sensitivity is 10–100 times higher than that of the UV detector for strong UV absorbing materials [37]. This is normally used as an advantage in the measurement of specific fluorescent species in samples. Only some compounds have a natural fluorescence. For example; conjugated double bond structures in combination with aliphatic or alicyclic, and compounds having several fused rings in their structure, show fluorescent characteristics. TOB has no natural fluorescence, but the naturally fluorescent reagent; NSCl was successfully used to react with TOB in presence of trimethylamine as a reaction catalyst and HCl scavenger to create derivatives that possess a fluorescing ligand. NSCl reacted by nucleophilic substitution with the terminal amine group to form an amide [38]. To the best of our knowledge, only two publications could employ NSCl in derivatization of primary amine [38, 39] for UV detection. However, NSCl could also be potentially used as a fluorophoric agent for sensitive fluorescence detection.

In this work, TOB reacted with NSCl after heating at 80 ºC for 40 min to give fluorescent derivative. The proposed reaction mechanism could be nucleophilic substitution with the hydroxyl or the amino groups. Triethylamine (TEA) was used as a catalyst with an assumption of being firstly reacted with NSCl, according to the mechanism scheme shown in Fig. 2. The reaction mechanism was investigated by NMR, IR and mass spectrometry. The number of signals in the NMR spectrum indicated the number of nonequivalent protons in a molecule. 1H NMR spectrometric analysis of TOB was obtained in deuterated DMSO as TOB was insoluble in deuterated chloroform CDCl3. Deuterated DMSO (dimethyl sulfoxide-d6) is an isotopologue of dimethyl sulfoxide with chemical formula ((CD3)2S = O), in which the hydrogen atoms (H) are replaced with their isotope deuterium (D). TOB structure had 20 types of nonequivalent protons, five primary amines and five hydroxyl groups. These protons are known to be involved in hydrogen bonding and are subsequently observed over a wide range of expected chemical shifts δ 0.5–5.5 ppm (Fig. 3a). This could be attributed to deshielding that occurs due to electron withdrawing effect of O and N. Since hydrogen bonds are dynamic in nature (keep forming and breaking), a wide range of chemical shifts was observed. In 1H NMR spectrum of derivatization product; extensive hydrogen bonding caused the spectrum to appear very broad and therefore of limited diagnostic value. However, the appearance of the characteristic sulfonamide peaks in the downfield at 7–9 ppm indicated that the reaction underwent between one of the primary amino groups and the sulfonate group of NSCl (Fig. 3b and c).

Fig. 2
figure 2

The suggested mechanism for the derivatization reaction

Full size image
Fig. 3
figure 3

1H NMR spectrum of tobramycin (a), 2-naphthalenesulfonyl chloride (b) and the derivatization product (c)

Full size image

IR Spectroscopy was employed to further study the reaction mechanism, as shown in Fig. 4. Amines and alcohols have N–H and O–H stretches which show up in the region 3300–3500 cm−1. The amine stretches tend to be sharper than O–H stretch which is distinguished by its broadness. Both sulfonamide and sulfonated ester could be predicted to result from the reaction of NSCl with TOB. Nevertheless, IR spectrum of the derivatization product exclude the later as no evidence for ester formation due to lack of signals in the region of 1650–1810 cm−1. Meanwhile, confirming sulfonamide formation by the presence of a signal of C-N in the region of 1030–1230 cm−1. A second evidence is distinguished by the disappearance of the forked NH2 stretch and predomination of broad O–H stretch at 3300–3500 cm−1.

Fig. 4
figure 4

IR spectra of tobramycin (a) and the derivatization product (b)

Full size image

Looking for confirming evidence, mass spectrometry was also employed; the presence of m/z 191.227, 211.681 and 309.421 peaks in the LC–MS spectra of the derivatization product confirmed the proposed assumption concerning the intermediates of the reaction; as all of them were supposed to be found in traces in LC–MS. Meanwhile, the detection of m/z 657.737, which is corresponding to the main product of the derivatization reaction, gave satisfactory evidence. Figure 5 shows the LC–MS spectra of the derivatization products.

Fig. 5
figure 5

The LC–MS spectra of the derivatization products

Full size image

3.1 Optimization of derivatization conditions

Increasing sensitivity and selectivity of the developed method have been taken into consideration by optimizing derivatization reaction conditions, including the concentration of the derivatizing agent, the type of catalyst, the temperature, the reaction time and the solvent. Different concentrations of NSCl were tested (0.25, 0.3, 0.35, 0.4, 0.45, 0.5 mg/mL). The most intense peak was obtained at a concentration of 0.35 mg/mL (Fig. 6a), which was chosen as the optimum concentration. Reaction temperatures for the derivatization procedure was investigated at 25 ºC, 50 ºC, 60 ºC, 70 ºC, 80 ºC and 90 ºC. The results indicated that temperature 80 ºC was sufficient to give satisfactory results (Fig. 6b).

Fig. 6
figure 6

Factors affecting reaction yield including NSCl concentration (a), temperature (°C), time (min) and mole fraction (d)

Full size image

In order to optimize the reaction conditions, different heating durations (10, 20, 40, 60, 80 min) were examined, after 40 min, a maximum yield was reached and stayed stable, thus 40 min was chosen as the optimum heating duration (Fig. 6c). As for the reaction catalyst, TEA, triethanolamine, pyridine, and sodium bicarbonate buffer pH 9.5 were investigated, among which only TEA gave satisfactory results, as other catalysts produced many peaks in the chromatogram, interfering with the derivatization product peak.

3.2 Determination of the stoichiometry of the reaction using Job's method

The stoichiometry of the reaction was studied using Job’s method of continuous variation [40, 41]. Complementary proportions of the two reactants (TOB and NSCl) were mixed in 10 mL screw capped test tubes, and 20 µL of TEA was added in each tube, then each tube was tightly closed and heated in a thermostatically controlled water bath at 80 ºC for 40 min. After cooling, the solution was evaporated under vacuum and the residue was re-dissolved in 0.5 mL acetonitrile. 20 µL of this solution was injected into HPLC/FL system. The fluorescence intensity was then plotted against mole fraction of TOB (Fig. 6d). The plot indicated 1:1 molar ratio for TOB and NSCl.

3.3 Method validation

The validity of the method was tested regarding linearity, specificity, accuracy and precision according to the ICH guidelines [42]. Under the specified experimental conditions, linear regression analysis of the data gave the following equation:

Y = 1.171 X – 4.8169 (r = 0.9998).

Where Y is the peak area of derivatization product of TOB, X is the concentration of TOB in ng/mL, r is the correlation coefficient. The proposed method was able to determine different concentrations of TOB with mean percentage recoveries of 99.9 ± 1.41 as shown in Table 1. System suitability was assessed by calculating capacity factor (k′), selectivity (α), resolution (Rs), tailing factor (T), and theoretical plate (N), where the system was found to be suitable (Table 2). The specificity of the proposed method was proven by injection of blank subjected to the same procedures except drug addition. The peaks produced by the blank injection do not interfere with the TOB derivatization product peak (Fig. 7).

Table 1 Accuracy of the proposed HPLC method
Full size table
Table 2 System suitability parameters by the proposed HPLC method
Full size table
Fig. 7
figure 7

Chromatograms obtained by the proposed HPLC/FL method, a blank, b derivatization product of TOB (1000 ng/mL)

Full size image

The precision was evaluated by analysis of three concentrations; using three determinations per concentration on three consecutive days. The relative standard deviation did not exceed 1.48% (Table 3). The relationship between the peak areas and the amounts of drug was linear; the linear correlation coefficients were 0.9998 for TOB (Table 4). Robustness of the proposed method was assessed by making small changes such as variation of the pH of the mobile phase by ± 0.2; this minor change did not have significant effect on peak area of the method. The derivatization product was stable for at least 24 h in daylight and at room temperature.

Table 3 Precision of TOB by the proposed HPLC method
Full size table
Table 4 Validation parameters of the proposed HPLC method for the determination of TOB after derivatization
Full size table

The proposed derivatization method was compared to an official USP method for the determination of TOB in Tobrin® ophthalmic solution [43] and statistical comparison was made using student's t-test and variance ratio F-test. Statistical comparison revealed no statistically significant difference between the performances of both proposed and official methods (Table 5). However, the developed method was more sensitive for determination of TOB in aqueous solutions. The method is also more sensitive than other reported HPLC methods for TOB analysis in pharmaceuticals and biological fluids, as shown in Table 6. This method opened the gate for more applications in non-chromophoric drug analysis, using NSCl as a derivatizing agent.

Table 5 Statistical comparison between the proposed derivatization method and the official method for determination of tobramycin in Tobrin® ophthalmic solution
Full size table
Table 6 reported HPLC methods for determination of tobramycin in pharmaceuticals and biological fluids
Full size table

To explore the possibility of applying this derivatization reaction to other drugs, the interaction of NSCl with various other nitrogenous compounds was tested. The contour plots obtained by a spectrofluorometer before and after derivatization of selected analytes are shown in Table S1. These results indicate that NSCl could be applied for other amine compounds under similar experimental conditions, especially for non-chromophoric drugs.

4 Conclusion

A novel, highly sensitive, selective and fully validated HPLC/FL method was developed for analysis of TOB after a simple one-step pre-column derivatization. The suggested mechanism for the proposed reaction between TOB and 2-naphthalenesulfonyl chloride was confirmed by further investigation using IR, proton NMR and LC–MS in order to elucidate the structure of the derivatization product. Favorable advantages of the proposed HPLC/FL method are twofold: a significant increase in sensitivity, as well as simplicity of sample preparation. The method is applicable for routine analysis.