Tetracycline Water Soluble Formulations with Enhanced Antimicrobial Activity

The negligible water solubility of tetracycline (TC), a well-known antibiotic of clinical use, is the major disadvantage for its oral administration. With the aim to improve the water solubility of TC, the micelles of formulae SLS@TC and CTAB@TC (SLS = sodium lauryl sulphate and CTAB = cetrimonium bromide) were synthesized. The micelles SLS@TC and CTAB@TC were characterized by melting point (m.p.), thermogravimetric differential thermal analysis (TG-DTA), differential scanning calorimetry (DTG/DSC), attenuated total reflection spectroscopy (FT-IR-ATR), ultra-violet visible (UV/vis) spectroscopy, proton nucleus magnetic resonance (1H-NMR) spectroscopy, and the ultrasonically-induced biregringence technique. The antimicrobial activity of SLS@TC and CTAB@TC was evaluated, by means of minimum inhibitory concentration (MIC), minimum bactericidal concentration (MBC), and inhibition zone (IZ), against the Gram negative bacterial strains Pseudomonas aeruginosa (P. aeruginosa) and Escherichia coli (E. coli) and the Gram positive ones of the genus of Staphylococcus epidermidis (S. epidermidis) and Staphylococcus aureus (S. aureus). Generally, both micelles show better activity than that of TC against the microbial strains tested. Thus, the MIC value of CTAB@TC is 550-fold higher than that of free TC against S. epidermidis. Despite the stronger activity of CTAB@TC than SLS@TC against both Gram negative and Gram positive microbes, SLS@TC is classified as a bactericidal agent (in that it eliminates 99.9% of the microbes), in contrast to CTAB@TC, which is bacteriostatic one (inhibits, but does not kill the organisms). The toxicity of SLS@TC and CTAB@TC was evaluated against human corneal eukaryotic cells (HCECs). Moreover, SLS@TC and CTAB@TC exhibit low in vivo toxicity against Artemia salina, even at concentrations up to threefold higher than those of their MICmax. Therefore, SLS@TC and CTAB@TC can be candidates for the development of new antibiotics.


Introduction
The properties of tetracycline (TC) are considered as ideal for antibiotic drugs [1]. TC is used for the treatment of a number of infections caused by both aerobic and anaerobic Gram positive and negative bacteria [2]. Its clinical safety and acceptable tolerability is proven [1]. Moreover, TC offers a valuable and relatively safe therapeutic modality in the treatment or inhibition of corneal ulceration [3]. However, the antimicrobial activity of TC suffers with resistance developed by microbes [1]. Moreover, the activity of TC is limited by the first passage phenomenon, where a portion of the antibiotic is always ( 1 H-NMR), and the ultrasonically-induced birefringence technique. The minimum inhibitory concentration (MIC), minimum bactericidal concentration (MBC), and inhibition zone (IZ) of the micelles against P. aeruginosa, E. coli, S. epidermidis, and S. aureus were evaluated. The in vitro toxicity of the new formulations was determined against human corneal eukaryotic cells (HCECs). The in vivo toxicity was tested against Artemia salina. The SLS and CTAB were chosen for the solubilazation of TC because (i) both are used for drug delivery [30][31][32][33][34]; (ii) CTAB cation is an effective antiseptic agent against bacteria and fungi [35]; and (iii) SLS is an anionic body detergents-cleansers (shower gels and toothpastes) ingredient, without any harmful irritations [36], while United States Food and Drug Administration (FDA or USFDA) recognized it as a harmless food ingredient (21 CFR 172,822). Their usage is limited, however, by their irritating side effect, especially upon their administration to the eye. This disadvantage, on the other hand, is overcome by the high, nonirritant concentration (<0.1% w/w), which allows their use for the treatment or inhibition of corneal ulceration [36].

Determination of Critical Micelle Concentration (CMC)
Because of the low solubility of TC in aqueous media, the micelles SLS@TC and CTAB@TC are formed, using the SLS and CTAB as surfactants. In order to assert the formation of the SLS@TC and CTAB@TC, the determination of the CMC was performed via potentiometric titration at 36 • C [37] Figure S1). Above the CMC value, surfactant molecules self-agglomerate to form the micelles [27]. Upon the formation of the micelles, a turning point (change of slope) is observed in the graph [27]. The CMC values were determined as follows: a solution of SLS or CTAB (2 × 10 −1 M) in ddH 2 O was added to 20 mL ddH 2 Figure S1). These values lead to a content of 8.3% w/w TC in SLS@TC and 7.6% w/w TC in CTAB@TC. Both micelles SLS@TC and CTAB@TC are highly soluble in dd water and dimethyl sulfoxide (DMSO). To confirm whether TC interacts with SLS or CTAB in the solid state to give a composite material or a mixture, DSC studies were carried out. DSC diagrams of SLS@TC and CTAB@TC are shown in Figure S3. Four endothermic phase transitions at 111.1, 185.8, 229.7, and 259.3 • C are observed in the DSC diagram of SLS@TC, while the corresponding ones were observed at 112.3 and 257.6 • C for CTAB@TC. The phase of SLS at 95.8 • C is observed at 110.6 • C in the corresponding one of SLS@TC, suggesting the formation of a composite material ( Figure 1). Meanwhile, the transformation at 109.3 • C in the case of CTAB is also observed at 112.2 • C in CTAB@TC, indicating that a mixture of the micelle's components is obtained in this case.

Vibrational Spectroscopy
The IR spectra of TC, SLS, CTAB, SLS@TC, and CTAB@TC are shown in Figures S4 and S5. The IR spectrum of SLS@TC is mainly dominated by the vibrations of SLS. Thus, the ν(H-C) vibrations in the IR spectrum of SLS@TC are observed at 2918 and 2850 cm −1 . The vibration band at 1070 cm −1 in the spectrum of SLS@TC is attributed to ν asym of the S=O, while the vibration at 1403 cm −1 in the IR Antibiotics 2020, 9,845 4 of 23 spectrum of CTAB@TC is attributed to δ(N + (CH 3 ) 3 ). However, no characteristics bond vibrations such as to ν asym (C=O) or C=C (1637 and 1577 cm −1 , respectively) of TC are observed in the IR spectrum of the micelle owing to the low stoichiometry of TC/SLS or TC/CTAB in the micelles.

Vibrational Spectroscopy
The IR spectra of TC, SLS, CTAB, SLS@TC, and CTAB@TC are shown in Figures S4 and S5. The IR spectrum of SLS@TC is mainly dominated by the vibrations of SLS. Thus, the ν(H-C) vibrations in the IR spectrum of SLS@TC are observed at 2918 and 2850 cm −1 . The vibration band at 1070 cm −1 in the spectrum of SLS@TC is attributed to νasym of the S=O, while the vibration at 1403 cm −1 in the IR spectrum of CTAB@TC is attributed to δ(N + (CH3)3). However, no characteristics bond vibrations such as to νasym (C=O) or C=C (1637 and 1577 cm −1 , respectively) of TC are observed in the IR spectrum of the micelle owing to the low stoichiometry of TC/SLS or TC/CTAB in the micelles.

UV/vis Study
The content of TC in the micelles was determined by UV/vis spectroscopy. Thus, the UV/vis spectra of five standard solutions in DMSO of TC (1 × 10 −4 , 8 × 10 −5 , 6 × 10 −5 , 4 × 10 −5 , 2×10 −5 , 1 × 10 −5 M) were recorded ( Figure S6A). The linear curves of the absorbance at λmax = 369 nm of DMSO solutions of TC versus concentration are shown in Figure S6B. Figure 2 shows the UV/vis spectra of TC, SLS, CTAB, SLS@TC, and CTAB@TC. Given that both SLS and CTAB are UV inactive at 369 nm (Figure 2), any absorption in this wavelength is attributed to TC. Therefore, the concentrations of TC found in SLS@TC and

UV/vis Study
The content of TC in the micelles was determined by UV/vis spectroscopy. Thus, the UV/vis spectra of five standard solutions in DMSO of TC (1 × 10 −4 , 8 × 10 −5 , 6 × 10 −5 , 4 × 10 −5 , 2×10 −5 , 1 × 10 −5 M) were recorded ( Figure S6A). The linear curves of the absorbance at λ max = 369 nm of DMSO solutions of TC versus concentration are shown in Figure S6B. Figure 2 shows the UV/vis spectra of TC, SLS, CTAB, SLS@TC, and CTAB@TC.  Given that both SLS and CTAB are UV inactive at 369 nm ( Figure 2), any absorption in this wavelength is attributed to TC. Therefore, the concentrations of TC found in SLS@TC and CTAB@TC are 8.3% w/w (SLS@TC) and 7.9% w/w (CTAB@TC), respectively. These values are in agreement with those found from CMC, namely 8.3% w/w (SLS@TC) and 7.6% w/w (CTAB@TC), which correspond to the 17/1 ([SLS]/[TC]) and 16/1 ([CTAB]/[TC]) molar ratios, respectively. The MIC, MBC, biofilm elimination concentration (BEC), and so on for the micelles SLS@TC and CTAB@TC were determined with respect to TC using its concentrations in micelles, as these were derived by UV/vis spectroscopy. The concentrations of micelles in solutions, however, are far higher to their CMCs, which ensures their stability.

1 H NMR Study
The 1 H NMR spectra in DMSO-d 6 of TC, SLS, CTAB, SLS@TC, and CTAB@TC (8-6 ppm) are shown in Figure 3. The resonances signals at 7.56-7.52 (t) ppm in the spectrum of TC are assigned to the amide H[NH 2− ] group (Scheme 1), while the signals at 7.13-7.10 (d) ppm and 6.92-6.90 (d) ppm are attributed to the aromatic protons HC arom 2-4 of TC (Scheme 1) [38]. The presence of these signals in the micelles spectra SLS@TC (7.52 (br), 7.12-7.10 (d), and 6.91-6.87 (d) ppm) and CTAB@TC (7.56-7.53 (t), 7.14-7.12 (d), and 6.93-6.91 (d) ppm) confirms the encapsulation of the antibiotic in the micelles. , and so on for the micelles SLS@TC and CTAB@TC were determined with respect to TC using its concentrations in micelles, as these were derived by UV/vis spectroscopy. The concentrations of micelles in solutions, however, are far higher to their CMCs, which ensures their stability.

Transient Acoustically-Induced Birefringence Measurements
For measuring the transient acoustically-induced birefringence, we used an instrument depicted in Figure 4. Its components have been described in detail elsewhere [39]. A solution of the sample was loaded in a cylindrically-shaped and thermostated non-birefringent glass cell with 1 cm optical path-length. A square electric pulse was delivered to a piezoelectric element, which further transmitted the ultrasonic wave into the liquid in perpendicular direction with respect to the light propagation. A 5 mW linearly polarized He-Ne laser (Melles-Griot, λ = 632.8 nm) working at TEM00 was used as the probe light source. In order to perform birefringence measurements, the cell was

Transient Acoustically-Induced Birefringence Measurements
For measuring the transient acoustically-induced birefringence, we used an instrument depicted in Figure 4. Its components have been described in detail elsewhere [39]. A solution of the sample was loaded in a cylindrically-shaped and thermostated non-birefringent glass cell with 1 cm optical path-length. A square electric pulse was delivered to a piezoelectric element, which further transmitted the ultrasonic wave into the liquid in perpendicular direction with respect to the light propagation. A 5 mW linearly polarized He-Ne laser (Melles-Griot, λ = 632.8 nm) working at TEM00 was used as the probe light source. In order to perform birefringence measurements, the cell was placed between a polarizer oriented at π/4 and a quarter wave plate (λ/4-plate) in cross-polarization relative to the polarizer. The quarter wave plate was followed by an analyzer with this plane rotated by an angle α clockwise or counterclockwise relative to the λ/4-plate. The transmitted intensity of the laser beam was detected by a fast photodiode and monitored in a digital oscilloscope.  When birefringence is induced by ultrasound in the liquid sample, a change in the transmitted intensity ∆I δ (t) = I δ (t) − I α is observed. I δ (t) is the light intensity attributed to phase retardation δ and I α corresponds to the light intensity at an angle α when the acoustic field is off. The intensity change is related to the induced optical retardation δ(t) through the following equation: The induced birefringence is associated with the retardation through the following: where d is the optical path-length and λ is the laser wavelength. The applied ultrasonic pulse and the acoustically-induced birefringence are presented in Figure 4b. When the acoustic square pulse perturbation is applied, the measured light intensity exhibits two different relaxations after switching the field on and off. During the acoustic field-on regime, the rise of the birefringence is given by the following: Antibiotics 2020, 9, 845 7 of 23 while, after sudden termination of the field, the birefringence decay is as follows: where ∆n max corresponds to the steady state birefringence when the acoustic field is on, τ is the characteristic relaxation time for the rise and decay of the birefringence signal, and β is assigned to the width of the relaxation time distribution. A simple exponential function (β = 1) is inadequate to describe the experimental data, implying a distribution of the relaxation times. The smaller the value of β, the larger the distribution of the relaxation times. A representative fitting example of the normalized birefringence decay ∆n/ ∆n max is presented in Figure 5, where an acceptable goodness of fit is observed.
given by the following: while, after sudden termination of the field, the birefringence decay is as follows: where Δnmax corresponds to the steady state birefringence when the acoustic field is on, τ is the characteristic relaxation time for the rise and decay of the birefringence signal, and β is assigned to the width of the relaxation time distribution. A simple exponential function (β = 1) is inadequate to describe the experimental data, implying a distribution of the relaxation times. The smaller the value of β, the larger the distribution of the relaxation times. A representative fitting example of the normalized birefringence decay Δn/Δnmax is presented in Figure 5, where an acceptable goodness of fit is observed.  For all measurements, the liquid sample was kept in a sealed cuvette under isothermal conditions corresponding to 37 °C. The entire volume of the liquid was exposed uniformly to the applied acoustic field. As no flow is permitted in and out of the cell, possible directional flow or liquid sharing is missing.
The application of the acoustic pulse allows us to observe the molecular motion dynamics directly in the time-domain by monitoring the induced birefringence trace. In general, the acoustic field will induce several types of dynamic processes in isotropic liquids and solutions at equilibrium involving intermolecular interactions. As the acoustic field is switched on, molecules are forced to orient along the propagation of the ultrasound and the whole system becomes un-axial and birefringent. After switching off the acoustic field, the induced orientation will fade away through a diffusion process of the molecules. The rotation is strongly influenced by the polarization of the neighboring molecules through dipole-induced-dipole interactions (DID). The strength of these interactions is subject to the distance and relative position between molecules. After the system is not exposed to the acoustic field, the DID interactions will level off to a minimum free-energy state, For all measurements, the liquid sample was kept in a sealed cuvette under isothermal conditions corresponding to 37 • C. The entire volume of the liquid was exposed uniformly to the applied acoustic field. As no flow is permitted in and out of the cell, possible directional flow or liquid sharing is missing.
The application of the acoustic pulse allows us to observe the molecular motion dynamics directly in the time-domain by monitoring the induced birefringence trace. In general, the acoustic field will induce several types of dynamic processes in isotropic liquids and solutions at equilibrium involving intermolecular interactions. As the acoustic field is switched on, molecules are forced to orient along the propagation of the ultrasound and the whole system becomes un-axial and birefringent. After switching off the acoustic field, the induced orientation will fade away through a diffusion process of the molecules. The rotation is strongly influenced by the polarization of the neighboring molecules through dipole-induced-dipole interactions (DID). The strength of these interactions is subject to the distance and relative position between molecules. After the system is not exposed to the acoustic field, the DID interactions will level off to a minimum free-energy state, while molecules will continue to rotate because of their inertia, but under the restrictions posed by the neighboring molecules. These constraints induce molecular libration relaxation on a femtosecond timescale, which is ultrafast and beyond the limits of the ultrasonically-induced birefringence technique. Molecular liquids and their mixtures exhibit usually fast birefringence response (see, e.g., [39] and references therein). In the case of SLS@TC and CTAB@TC dilute solutions, a relatively slower response is observed because of the peculiar nature of the underlying dynamics, where a self-organization into larger supramolecular aggregates takes place, leading to a dynamically complex structure. Thus, both SLS@TC and CTAB@TC solutions cannot be considered as a dispersion of anisotropic particles in a liquid matrix with no polarization and steric effects.
The experimental results recorded in this work were adequately fitted by a stretched exponential function (Equation (4)), exhibiting analogous behavior with other complex liquids [40,41]. With switching off the acoustic field, the recorded birefringence trace decreases to zero as a result of thermal orientational fluctuations (Brownian rotation) of the particles. The corresponding relaxation time τis related to the hydrodynamic volume V h of the particle through the following equation [42]: where T is the absolute temperature, k B is the Boltzmann's constant, and η is the shear viscosity of the solvent. The relaxation times from the birefringence measurements can be estimated from the fitting procedure and used to calculate the hydrodynamic volume of the entities in the solutions. In Figure 6, the concentration dependence of the fitting parameter α is presented for both SLS@TC and CTAB@TC systems. It seems that the relaxation time distribution of CTAB@TC solutions is wider compared with the relaxation time distribution of SLS@TC solutions. Furthermore, for each system, the distribution width becomes narrow monotonically from dilute to dense solutions.
dynamically complex structure. Thus, both SLS@TC and CTAB@TC solutions cannot be considered as a dispersion of anisotropic particles in a liquid matrix with no polarization and steric effects. The experimental results recorded in this work were adequately fitted by a stretched exponential function (Equation (4)), exhibiting analogous behavior with other complex liquids [40,41]. With switching off the acoustic field, the recorded birefringence trace decreases to zero as a result of thermal orientational fluctuations (Brownian rotation) of the particles. The corresponding relaxation time τ is related to the hydrodynamic volume Vh of the particle through the following equation [42]: where T is the absolute temperature, kB is the Boltzmann's constant, and η is the shear viscosity of the solvent. The relaxation times from the birefringence measurements can be estimated from the fitting procedure and used to calculate the hydrodynamic volume of the entities in the solutions. In Figure 6, the concentration dependence of the fitting parameter α is presented for both SLS@TC and CTAB@TC systems. It seems that the relaxation time distribution of CTAB@TC solutions is wider compared with the relaxation time distribution of SLS@TC solutions. Furthermore, for each system, the distribution width becomes narrow monotonically from dilute to dense solutions.  The hydrodynamic volumes, as obtained from Equation (5), were used to confirm the incorporation of tetracycline in SLS and CTAB micelles and to quantitatively follow the volume change. The relative micelle volume changes are shown in Figure 7. The results clearly indicate the incorporation of TC in the parental micelles for both surfactants. We also calculated the theoretically predicted volume change between the initial and final complex micelle. We considered the diameter of the initial SLS micelle as approximately two times the size of the SLS (19.5 Å) and the diameter of the initial CTAB micelle as two times the size of the CTAB (24.2 Å), respectively. The diameter of the The hydrodynamic volumes, as obtained from Equation (5), were used to confirm the incorporation of tetracycline in SLS and CTAB micelles and to quantitatively follow the volume change. The relative micelle volume changes are shown in Figure 7. The results clearly indicate the incorporation of TC in the parental micelles for both surfactants. We also calculated the theoretically predicted volume change between the initial and final complex micelle. We considered the diameter of the initial SLS micelle as approximately two times the size of the SLS (19.5 Å) and the diameter of the initial CTAB micelle as two times the size of the CTAB (24.2 Å), respectively. The diameter of the SLS@TC and CTAB@TC, after the incorporation of TC in the parental micelle, is calculated as the sum of the parental micelle diameter plus the mean diameter of the TC molecule (45.11 Å). The calculation of the mean TC diameter was carried out in vacuum (without any interactions), with the use of the Gaussian 09 series of programs [43]. The theoretically calculated volume changes for SLS@TC and CTAB@TC micelles are in agreement with the experimental data, confirming the proposed underlying mechanism. use of the Gaussian 09 series of programs [43]. The theoretically calculated volume changes for SLS@TC and CTAB@TC micelles are in agreement with the experimental data, confirming the proposed underlying mechanism. Additional information concerning the length distribution of the micelles can be obtained from the birefringence decay measurements. In the context of the Watson-Jennings approach, the length distribution is assumed to follow a log-normal behavior and is given by the following [44]: The above equation is a two-parameter equation, where dm is the median and σ corresponds to the breadth of the size distribution. Both parameters can be estimated as the solution of the following system of equations: The DWJ factor is the initial logarithmic derivative, while IWJ represents the area under the birefringence trace. Following the above methodology, we estimated the particle size distributions for both systems and the results are presented in Figure 8. In both systems, the size distribution is shifted and widened with increasing solution concentration, with the CTAB@TC system exhibiting the most intense changes. The different behavior in terms of particle aggregation arises from the different head group properties, which in turn influence the extent of the hydrophobic or hydrophilic tail interactions. Both SLS and CTAB form complexes that involve an arrangement of Additional information concerning the length distribution of the micelles can be obtained from the birefringence decay measurements. In the context of the Watson-Jennings approach, the length distribution is assumed to follow a log-normal behavior and is given by the following [44]: The above equation is a two-parameter equation, where d m is the median and σ corresponds to the breadth of the size distribution. Both parameters can be estimated as the solution of the following system of equations: The D WJ factor is the initial logarithmic derivative, while I WJ represents the area under the birefringence trace. Following the above methodology, we estimated the particle size distributions for both systems and the results are presented in Figure 8. In both systems, the size distribution is shifted and widened with increasing solution concentration, with the CTAB@TC system exhibiting the most intense changes. The different behavior in terms of particle aggregation arises from the different head group properties, which in turn influence the extent of the hydrophobic or hydrophilic tail interactions. Both SLS and CTAB form complexes that involve an arrangement of electrostatic and hydrophobic interactions. The shift in the distribution functions is related to the characteristic particle size obtained from the birefringence relaxation due to rotational diffusion. The relative variance of the distribution for each case can be evaluated from the polydispersity index (PI) considering the standard deviation and the mean size of the distribution, and the results are presented in Figure 9. The lower the PI values, the lower the dispersion and, below 0.04, the system can be considered as monodispersed. It seems that both systems are polydispersed, with CTAB@TC more dispersed than the SLS@TC. The size dispersion behaves in an analogous manner with relaxation time dispersion, as observed in Figure 6. The analysis of the ultrasonically-induced birefringence relaxation permits to comprehend the distribution of the hydrodynamic diameter or volume of the micelles formed. Furthermore, on the basis of the concentration-dependent birefringence measurements, it is possible to distinguish between simple and TC incorporated micelles and obtain a more quantitative description of the mechanism efficiency. The relaxation time and hydrodynamic diameter distribution highlight the complexity of the interactions involved in the intermolecular assemblies present in surfactant-based solutions. The relative variance of the distribution for each case can be evaluated from the polydispersity index (PI) considering the standard deviation and the mean size of the distribution, and the results are presented in Figure 9. The lower the PI values, the lower the dispersion and, below 0.04, the system can be considered as monodispersed. It seems that both systems are polydispersed, with CTAB@TC more dispersed than the SLS@TC. The size dispersion behaves in an analogous manner with relaxation time dispersion, as observed in Figure 6. The analysis of the ultrasonically-induced birefringence relaxation permits to comprehend the distribution of the hydrodynamic diameter or volume of the micelles formed. Furthermore, on the basis of the concentration-dependent birefringence measurements, it is possible to distinguish between simple and TC incorporated micelles and obtain a more quantitative description of the mechanism efficiency. The relaxation time and hydrodynamic diameter distribution highlight the complexity of the interactions involved in the intermolecular assemblies present in surfactant-based solutions.

Attenuated Total Reflection Measurements
The infrared spectra of the solutions in attenuated total reflection (ATR) mode were recorded by employing an Alpha spectrometer (Bruker) in the 500-4000 cm −1 spectral region with a Deuterated Triglycine Sulfate (DTGS) detector. The infrared radiation is transferred to the ATR crystal made of ZnSe with an angle of incidence set at 45° and all spectra were recorded with a 2 cm −1 spectral resolution. The ATR crystal surface was completely covered by the liquid sample, thus permitting quantitative estimations from the spectra. The ATR crystal was cleaned with isopropanol after each spectrum acquisition to ensure that no impurities from previous measurements were retained on the crystal's surface.

Vibrational Modes-Short-Range Structure
The spectral fingerprints of the DMSO solvent are expected to dominate the spectra of all solutions in such low concentrations. Indeed, the main absorption observed in Figure 10 is ascribed to the S-O symmetric (1018 cm −1 ) and antisymmetric (1042 cm −1 ) stretching vibration of DMSO dimers [45][46][47]. Nevertheless, the spectral variations observed in Figure 10 are attributed to changes in the local environment due to surfactant self-association and to the replacement of the dipole-dipole interactions between solvent molecules (DMSO) with hydrogen-bonded "new" species with DMSO dimers. This interplay is observed as an isosbestic point in the concentration-dependent spectra of both SLS@TC and CTAB@TC systems presented in Figure 10.

Attenuated Total Reflection Measurements
The infrared spectra of the solutions in attenuated total reflection (ATR) mode were recorded by employing an Alpha spectrometer (Bruker) in the 500-4000 cm −1 spectral region with a Deuterated Triglycine Sulfate (DTGS) detector. The infrared radiation is transferred to the ATR crystal made of ZnSe with an angle of incidence set at 45 • and all spectra were recorded with a 2 cm −1 spectral resolution. The ATR crystal surface was completely covered by the liquid sample, thus permitting quantitative estimations from the spectra. The ATR crystal was cleaned with isopropanol after each spectrum acquisition to ensure that no impurities from previous measurements were retained on the crystal's surface.

Vibrational Modes-Short-Range Structure
The spectral fingerprints of the DMSO solvent are expected to dominate the spectra of all solutions in such low concentrations. Indeed, the main absorption observed in Figure 10 is ascribed to the S-O symmetric (1018 cm −1 ) and antisymmetric (1042 cm −1 ) stretching vibration of DMSO dimers [45][46][47]. Nevertheless, the spectral variations observed in Figure 10 are attributed to changes in the local environment due to surfactant self-association and to the replacement of the dipole-dipole interactions between solvent molecules (DMSO) with hydrogen-bonded "new" species with DMSO dimers. This interplay is observed as an isosbestic point in the concentration-dependent spectra of both SLS@TC and CTAB@TC systems presented in Figure 10.
The spectrum of SLS reveals a set of antisymmetric and symmetric stretching vibrations of -CH 2and -CH 3 groups of the hydrocarbon tail in the 3000-2800 cm −1 range [48,49]. The bending modes of these groups are observed at~1461 and~1377 cm −1 , respectively. The two peaks located at 1237 and 1200 cm −1 are assigned to antisymmetric S-O stretching, while the bands at 1060 and 974 cm −1 are assigned to symmetric S-O stretching vibration of the -OSO 3 − species [48,49]. The spectrum of the isolated sulfate SO 4 2− species with tetrahedral symmetry (T d ) exhibits a totally symmetric S-O stretching ν 1 mode at 980 cm −1 , which is IR inactive, and a triply-degenerate antisymmetric ν 3 band located at 1102 cm −1 [50]. In the case of the SLS system, the sulfonate head-group reveals lower tetrahedral symmetry with the C 3V point group instead of T d . Moderate distortion of the tetrahedral units results in a splitting of the antisymmetric stretching mode into a double-degenerate and a non-degenerate band observed at high-and low-frequencies, respectively [51,52]. Additional distortion causes further reduction of the tetrahedral symmetry to the C 2V point group with a subsequent splitting of the ν 3 band into three counterparts in the frequency range from 1250 to 1050 cm −1 . The correlation diagram presented in Table 1 shows the band splitting into a large number of bands after moderate and severe distortions of the tetrahedral units with initial T d symmetry. The spectrum of SLS reveals a set of antisymmetric and symmetric stretching vibrations of -CH2-and -CH3 groups of the hydrocarbon tail in the 3000-2800 cm −1 range [48,49]. The bending modes of these groups are observed at ~1461 and ~1377 cm −1 , respectively. The two peaks located at 1237 and 1200 cm −1 are assigned to antisymmetric S-O stretching, while the bands at 1060 and 974 cm −1 are assigned to symmetric S-O stretching vibration of the -OSO3 − species [48,49]. The spectrum of the isolated sulfate SO4 2− species with tetrahedral symmetry (Td) exhibits a totally symmetric S-O stretching ν1 mode at 980 cm −1 , which is IR inactive, and a triply-degenerate antisymmetric ν3 band located at 1102 cm −1 [50]. In the case of the SLS system, the sulfonate head-group reveals lower tetrahedral symmetry with the C3V point group instead of Td. Moderate distortion of the tetrahedral units results in a splitting of the antisymmetric stretching mode into a double-degenerate and a non-degenerate band observed at high-and low-frequencies, respectively [51,52]. Additional distortion causes further reduction of the tetrahedral symmetry to the C2V point group with a subsequent splitting of the ν3 band into three counterparts in the frequency range from 1250 to 1050 cm −1 . The correlation diagram presented in Table 1 shows the band splitting into a large number of bands after moderate and severe distortions of the tetrahedral units with initial Td symmetry.

ν4c(A)
An interesting spectral observation in the SLS@TC system is that the antisymmetric stretching band of SLS at ~1208 cm −1 , assigned to the charged ionic sulfonate head OSO3 − , is shifted in frequency above a concentration of 295 μg/mL. This finding is attributed to the gradual hydrogen bond formation between one or two sulfonate oxygen atoms with the tetracycline's functional groups, indicating the transition from SLS-micelle to complex SLS@TC micelle. The non-degenerate A1 antisymmetric stretching vibration is parallel to the direction of the SLS molecule, and thus is sensitive to the presence of tetracycline molecule, while the doubly-degenerate E vibration is perpendicular to the direction of the SLS molecule and is prone to lateral interaction between An interesting spectral observation in the SLS@TC system is that the antisymmetric stretching band of SLS at~1208 cm −1 , assigned to the charged ionic sulfonate head OSO 3 − , is shifted in frequency above a concentration of 295 µg/mL. This finding is attributed to the gradual hydrogen bond formation between one or two sulfonate oxygen atoms with the tetracycline's functional groups, indicating the transition from SLS-micelle to complex SLS@TC micelle. The non-degenerate A 1 antisymmetric stretching vibration is parallel to the direction of the SLS molecule, and thus is sensitive to the presence of tetracycline molecule, while the doubly-degenerate E vibration is perpendicular to the direction of the SLS molecule and is prone to lateral interaction between surfactant-surfactant entities [52]. Relative intensity changes and new bands are observed in the ATR spectra of solutions with increasing solution concentration in addition to the bands attributed to the main functional groups of the surfactant (SLS, CTAB), tetracycline, and the solvent (DMSO). These spectral observations are denoted with dotted lines in Figure 11 for SLS@TC and CTAB@TC solutions for selected concentrations. The spectrum of DMSO solvent is also shown in the same spectral region for comparison. These observations account for the above described interactions related to complex micelle formation.

Effect of SLS@TC, CTAB@TC, TC, SLS, and CTAB on the Growth of Microbial Strains
The antimicrobial potency of SLS@TC, CTAB@TC, TC, SLS, and CTAB was tested against Gram negative (P. aeruginosa and E. coli) and Gram positive (S. epidermidis and S. aureus) bacterial strains by means of MIC (Table 2, Figures S7-S10). MIC is defined as the lowest concentration needed for the inhibition of the bacterial growth [21][22][23][24][25][26][27][28]. The TC is suspended in water for the experiments. The ranges of MIC values of SLS@TC and CTAB@TC are 3.7-27.2 μΜ and 0.07-5.8 μΜ, respectively. The biological effects of the micelles SLS@TC and CTAB@TC against the strains tested were determined with respect to TC using its concentrations in the micelles, derived by UV/vis spectroscopy, that is, 8.3% w/w TC in SLS@TC and 7.9% w/w TC in CTAB@TC. Thus, the concentrations of micelles in solutions remain higher their CMCs, reserving their stability, but lower than the MIC values of the surfactants themselves.   (Table 2, Figures S7-S10). MIC is defined as the lowest concentration needed for the inhibition of the bacterial growth [21][22][23][24][25][26][27][28]. The TC is suspended in water for the experiments. The ranges of MIC values of SLS@TC and CTAB@TC are 3.7-27.2 µM and 0.07-5.8 µM, respectively. The biological effects of the micelles SLS@TC and CTAB@TC against the strains tested were determined with respect to TC using its concentrations in the micelles, derived by UV/vis spectroscopy, that is, 8.3% w/w TC in SLS@TC and 7.9% w/w TC in CTAB@TC. Thus, the concentrations of micelles in solutions remain higher their CMCs, reserving their stability, but lower than the MIC values of the surfactants themselves. SLS@TC and CTAB@TC show a higher antibacterial effect than TC towards the Gram negative and Gram positive bacteria. This is attributed to their water solubility, which enhances their bioavailability, in contrast to the water insoluble TC. Both micelles exhibit the highest activity towards the bacterial strains of S. epidermidis and S. aureus, in contrast to TC. SLS@TC and CTAB@TC exhibit up to 6-fold and up to 540-fold stronger activity than TC towards S. epidermidis, respectively. The MIC values of SLS and CTAB range from 0.4 to >250 µM and they exhibit lower activity than the micelles of tetracycline. However, CTAB exhibits higher activity than SLS against all tested strains, which is also associated with the higher activity of CTAB@TC than SLS@TC. Especially, CTAB@TC is up to 89-fold more active than SLS@TC against S. epidermidis.

Determination of the Inhibition Zone (IZ) through the Agar Disk-Diffusion Method
The agar disk-diffusion method was used in order to survey the sensitivity of P. aeruginosa, E.coli, S. epidermidis, and S. aureus towards SLS@TC, CTAB@TC, TC, SLS, and CTAB at the concentration of 1 mM with regards to TC [21][22][23][24][25][26][27][28]. Therefore, the solutions of 1 mM SLS@TC, CTAB@TC, and TC contain the same amount of TC (1 mM), which makes their biological effect comparable. The diameter of inhibition zones of bacterial growth was measured after 20 h ( Table 2, Figures 12 and 13). Given that the solutions of SLS@TC and CTAB@TC contain TC in the same concentration, their IZs towards the tested bacteria lie between 31.8-23.8 and 13.7-34.0 mm, respectively, which are longer than the corresponding ones of TC itself. Although CTAB@TC forms IZs comparable with TC, SLS@TC creates significantly greater zones. As both SLS@TC and CTAB@CT are water soluble, this effect should be attributed to the formation of the composite material, confirming the conclusion of MBC (see above). The MBC value is defined as the lowest concentration of an antibacterial agent that can eliminate 99.9% of the bacterial inoculum [21][22][23][24][25][26][27][28]. The MBC values of SLS@TC, CTAB@TC, TC, SLS, and CTAB were determined against P. aeruginosa, E. coli, S. epidermidis and S. aureus ( Figures  S11-S14). The MBC values of SLS@TC and CTAB@TC range from 9.1-116.0 and 0.5-95.0 μΜ, towards all the bacteria, respectively ( Table 2). The MBC/MIC ratio classifies antibacterial agents into bactericidal or bacteriostatic ones. Thus, (i) bactericidal agents exhibit MBC/MIC lower than 2, while (ii) when this ratio is higher than 4, the agent is classified as bacteriostatic. Therefore, both micelles are bacteriostatic agents towards P. aueroginosa, E. coli, S. epidermidis and S. aureus. However, in the case of S. epidermidis, the SLS@TC is a bactericidal agent against (MBC/MIC = 1.9 ( Table 2)).

Determination of the Inhibition Zone (IZ) through the Agar Disk-Diffusion Method
The agar disk-diffusion method was used in order to survey the sensitivity of P. aeruginosa, E.coli, S. epidermidis, and S. aureus towards SLS@TC, CTAB@TC, TC, SLS, and CTAB at the concentration of 1 mM with regards to TC [21][22][23][24][25][26][27][28]. Therefore, the solutions of 1 mM SLS@TC, CTAB@TC, and TC contain the same amount of TC (1 mM), which makes their biological effect comparable. The diameter of inhibition zones of bacterial growth was measured after 20 h (Table 2, Figures 12 and 13). Given that the solutions of SLS@TC and CTAB@TC contain TC in the same concentration, their IZs towards the tested bacteria lie between 31.8-23.8 and 13.7-34.0 mm, respectively, which are longer than the corresponding ones of TC itself. Although CTAB@TC forms IZs comparable with TC, SLS@TC creates significantly greater zones. As both SLS@TC and CTAB@CT are water soluble, this effect should be attributed to the formation of the composite material, confirming the conclusion of MBC (see above).   Given that the CMCs of the micelles of TC with SLS and CTAB are 1:17 (SLS@TC) and 1:16 (CTAB@TC), the concentration of the surfactants in the case of IZ's measurements could be 17 and 16 mM, respectively, as the corresponding one of TC is 1 mM. However, micelles are considered as single entities, which build up as a result of hydrophobicity of the surfactants at higher than their CM concentrations. They can also encircle small molecules such as TC in one entity. Thus, SLS@TC and CTAB@TC are single entities and their concentrations are equal to the corresponding one of TC (1 mM), which is the limiting ingredient in micelles. Consequently, the same concentration (1 mM) was used for SLS or CTAB during the IZ's measurement. While SLS and CTAB show negligible or no antibacterial activities at the concentration of 1 mM, their micelles SLS@TC and CTAB@TC do. Moreover, as both SLS and CTAB are already used for drug delivery [30][31][32][33][34], the CTAB cation is an effective antiseptic agent against bacteria and fungi [35]; the SLS is an anionic body detergents-cleansers (shower gels and toothpastes) ingredient, without any harmful irritations [36], which has been recognized by USFDA as a harmless food ingredient (21 CFR 172,822); the antimicrobial activities of SLS and CTAB are expected to be limited.

Effect on Biofilm Formation by SLS@TC and CTAB@TC
Nowadays, 80% of the clinical infections are biofilm related. Especially, the ophthalmic biofilm infections are difficult to treat. Biofilm elimination can be achieved by applying metalo-antibiotic or their micelles [24][25][26][27].
The effect of SLS@TC, CTAB@TC, and TC against biofilms formed by P. aeruginosa and S. aureus was evaluated by the biofilm elimination concentration (BEC) using crystal violet assay [24][25][26][27]. The BEC is defined as the required concentration to achieve at least a 99.9% reduction in the viability of biofilm bacteria. The BEC values of SLS@TC, CTAB@TC, and TC are 557, 1875, and 427 µM, respectively, against P. aeruginosa, while they are 1913, 983, and >2300 µM [53], respectively, against S. aureus (Figures S15-S17). Thus, SLS@TC and CTAB@TC are more efficient against the biofilm of S. aureus, while they are less towards P. aeruginosa than free TC.  Table 2). The synthesized micelles are more toxic than TC.

Evaluation of In Vitro Genotoxicity
The micronucleus assay was employed for the evaluation of in vitro genotoxicity. It is widely used to monitor genetic damage in the cells and in order to evaluate the mutagenic, genotoxic, or teratogenic effect of metallodrugs [54]. The in vitro genotoxicity of SLS@TC, CTAB@TC, and TC was tested towards HCECs at their IC 50 values (Figure 14). The corresponding frequency of micronucleus after treatment of HCECs upon their treatment with SLS@TC, CTAB@TC, and TC is (2.0 ± 0.2), (1.9 ± 0.7), and (2.5 ± 0.7)%, respectively, in contrast to (2.4 ± 0.2)% for non-treated cells. The micronucleus frequencies of the micelles and TC indicate their non-mutagenic or genototic activity towards HCECs.

In Vivo Toxicity Evaluation, by Brine Shrimp Artemia Salina
The in vivo toxicity of the SLS@TC and CTAB@TC was screened by brine shrimp Artemia salina assay. The brine shrimp Artemia salina lethality test is a preliminary toxicity test, useful in predicting biological activities such as cytotoxic, phototoxic, and pesticidal activities [28,55].
The % survival of Artemia salina larvae in increasing concentrations of solutions with or without SLS@TC and CTAB@TC after 24 h is summarized in Table 3. The lethality was noted in terms of deaths of larvae. The mortality rate of brine shrimp larvae was found upon their incubation with SLS@TC and CTAB@TC in concentrations equal to MIC min , MIC max , 2 × MIC max , and 3 × MIC max against P. aeruginosa, E. coli, S. epidermidis, or S. aureus respectively, suggesting low toxicity at MIC max (Table 3).

Conclusions
The disadvantage of the negligible water solubility of tetracycline (TC), a well-known antibiotic of clinical use, is overcome by the development of its micelles with formulae SLS@TC and CTAB@TC. Both micelles were characterized by spectroscopy techniques and ultrasonic imaging. The hydrodynamic volumes are obtained from transient acoustically-induced birefringence measurements. The results indicate the incorporation of TC in the parental micelles for both surfactants. While the diameters of SLS or CTAB micelles are twice the corresponding ones of SLS or CTAB, respectively, the diameters of the SLS@TC and CTAB@TC are the sum of the diameters of

In Vivo Toxicity Evaluation, by Brine Shrimp Artemia Salina
The in vivo toxicity of the SLS@TC and CTAB@TC was screened by brine shrimp Artemia salina assay. The brine shrimp Artemia salina lethality test is a preliminary toxicity test, useful in predicting biological activities such as cytotoxic, phototoxic, and pesticidal activities [28,55].
The % survival of Artemia salina larvae in increasing concentrations of solutions with or without SLS@TC and CTAB@TC after 24 h is summarized in Table 3. The lethality was noted in terms of deaths of larvae. The mortality rate of brine shrimp larvae was found upon their incubation with SLS@TC and CTAB@TC in concentrations equal to MIC min , MIC max , 2 × MIC max , and 3 × MIC max against P. aeruginosa, E. coli, S. epidermidis, or S. aureus respectively, suggesting low toxicity at MIC max (Table 3).

Conclusions
The disadvantage of the negligible water solubility of tetracycline (TC), a well-known antibiotic of clinical use, is overcome by the development of its micelles with formulae SLS@TC and CTAB@TC. Both micelles were characterized by spectroscopy techniques and ultrasonic imaging. The hydrodynamic Antibiotics 2020, 9,845 18 of 23 volumes are obtained from transient acoustically-induced birefringence measurements. The results indicate the incorporation of TC in the parental micelles for both surfactants. While the diameters of SLS or CTAB micelles are twice the corresponding ones of SLS or CTAB, respectively, the diameters of the SLS@TC and CTAB@TC are the sum of the diameters of the parental micelles and that of tetracycline itself.
SLS@TC and CTAB@TC show a higher antibacterial effect than TC towards the Gram negative and Gram positive bacteria, because of their water solubility and their bioavailability in contrast to insoluble TC. Especially, SLS@TC and CTAB@TC exhibit up to 6-fold and up to 540-fold stronger activity than free TC towards S. epidermidis, respectively. Both micelles are bacteriostatic agents towards the tested strains. Moreover, SLS@TC and CTAB@TC exhibit greater inhibitory efficiency against the biofilm of S. aureus than TC itself, while they are less towards the corresponding one from P. aeruginosa. The in vitro genotoxic studies show that the micronucleus frequencies indicate the non-mutagenic or genotoxic activity of micelles towards HCECs. The in vivo toxicity studies towards brine shrimp larvae suggest low toxicity at MIC max for both micelles. The irritating side effect that both agents might be caused by the presences of SLS or CTAB is overcome because of their low effective concentration (MIC max < 120 µM or 0.01% w/w), which is lower than the irritant one (<0.1% w/w) [36]. Therefore, SLS@TC and CTAB@TC can be candidates for the development of new antibiotics. Bacterial Strains: For the antibacterial experiments, the strains of Staphylococcus epidermidis (ATCC ® 14990™), S. aureus subsp. aureus (ATCC ® 25923™), and P. aeruginosa PAO1 and Escherichia coli were used.

Materials and instruments
Effects of SLS@TC, CTAB@TC, TC, SLS, and CTAB on the growth of microbial strains: This study was performed according to standard procedure, which is also described elsewhere [21][22][23][24][25][26][27][28][29]. Briefly, the bacterial strains were streaked onto trypticase soy agar. The plates were incubated for 18-24 h at 37 • C. Three to five isolated colonies were selected of the same morphological appearance from the fresh agar plate using a sterile loop and transferred into a tube containing 2 mL of sterile saline solution. The optical density at 620 nm is 0.1, which corresponds to 10 8 cfu/mL. The final inoculum size for broth dilution is 5 × 10 5 cfu/mL. The total volume of the culture solution treated by micelles, TC, and surfactants, as well as the total volume of the positive and negative control, was 2 mL. The range of concentrations of micelles, TC, and surfactants is 0.02-300 µM. The growth is assessed after incubation for 20 h. The minimal inhibitory concentration (MIC) was determined as the concentration of the compound that inhibits the visible growth of the bacterium being investigated. The optical density of the solution at 620 nm versus concentrations graph leads to the MIC determination [21][22][23][24][25][26][27][28][29].
Effects on biofilm formation by micelles: Bacteria with a density of 6.7 × 10 6 cfu mL −1 were inoculated into LB medium for P. aeruginosa or tryptic soy broth for S. aureus (total volume = 1500 µL) and cultured for 24 h at 37 • C. Afterwards, the content of each tube was carefully removed and the tubes were washed with 1 mL 0.9% saline dilution. The negative control contained broth only. After the bacteria were incubated with 50-600 µM of the micelles for 20 h at 37 • C, the content of each tube was aspirated; each tube was washed three times with 1 mL methanol and 2 mL 0.9% saline and left to dry. Then, the tubes were stained for 15 min with crystal violet solution (0.3% w/v). Excess stain was rinsed off with 1 mL methanol and 2 mL 0.9% saline solution and then 3 mL 0.9% saline solution. The tubes were left to dry for 24 h and the bounded crystal violet was released by adding 30% glacial acetic acid. The optical density of the solution yielded was then measured at 550 nm to give the biofilm biomass.
Evaluation of genotoxicity by micronucleus assay. The micronucleus assay was carried out as previously described [21][22][23][24][25][26][27][28][29]. The procedure is described here briefly. HCECs were seeded (at a density of 15 × 10 4 cells/well cells per well) in glass cover slips, which were afterwards placed in six-well plates, with 3 mL of cell culture medium, and incubated for 24 h. HCECs were exposed with micelles at their IC 50 values for a period of 48 h. Afterwards, the coverslips were washed three times with Phosphate Buffer Solution (PBS), and with a hypotonic solution (75 mM KCl) for 10 min at room temperature. The hypotonized cells were fixed by at least three changes of 1/3 acetic acid/ethanol. The coverslips were also washed with cold methanol containing 1% acetic acid. The coverslips were stained with acridine orange (5 µg mL −1 ) for 15 min at 37 • C. After, the coverslips were rinsed three times with PBS to remove any excess acridine orange stain. The number of micronucleated cells per 1000 cells was determined using a fluoresce microscope. The MN were identified according to the previously established criteria: (1) the area of MN should correspond to 1/256 and 1/9 of the area of the main nuclei; (2) the shape of MN should be round or oval; (3) the MN should be readily distinguished from artifacts; (4) MN are not linked or connected to the main nucleus; (5) MN may touch, but not overlap the main nuclei, and the micronuclear boundary should be distinguishable from the nuclear boundary; and (6) the MN have the same staining intensity as the main nuclei [24].
Evaluation of toxicity with brine shrimp assay: Brine shrimp assay was performed by a method previously described [28]. Here, 1 g cysts were initially hydrated in freshwater for one hour in a separating funnel or cone shaped container. Seawater was prepared by dissolving 17 g of sea salt in 500 mL of distilled water [28]. The cone was facilitated with good aeration for 48 h at room temperature and under continuous illumination. After hatching, nauplii released from the egg shells were collected at the bright side of the cone (near the light source) using a micropipette. The larvae were isolated from the eggs by aliquoting them in a small beaker containing NaCl 0.9% [28]. An aliquot (0.1 mL) containing about 5 to 15 nauplii was introduced to each well of a 24-well plate and micelles were added in each well at MIC min , MIC max , 2 × MIC max , and 3 × MIC max . The final volume of each well is 1 mL with NaCl 0.9%. The brine shrimps were observed after of 24 h using a stereoscope. Larvae are considered dead if they do not exhibit any internal or external movement in 10 s of observation. Each experiment was repeated three times.