pH-Responsive Release of Ruthenium Metallotherapeutics from Mesoporous Silica-Based Nanocarriers

Ruthenium complexes are attracting interest in cancer treatment due to their potent cytotoxic activity. However, as their high toxicity may also affect healthy tissues, efficient and selective drug delivery systems to tumour tissues are needed. Our study focuses on the construction of such drug delivery systems for the delivery of cytotoxic Ru(II) complexes upon exposure to a weakly acidic environment of tumours. As nanocarriers, mesoporous silica nanoparticles (MSN) are utilized, whose surface is functionalized with two types of ligands, (2-thienylmethyl)hydrazine hydrochloride (H1) and (5,6-dimethylthieno[2,3-d]pyrimidin-4-yl)hydrazine (H2), which were attached to MSN through a pH-responsive hydrazone linkage. Further coordination to ruthenium(II) center yielded two types of nanomaterials MSN-H1[Ru] and MSN-H2[Ru]. Spectrophotometric measurements of the drug release kinetics at different pH (5.0, 6.0 and 7.4) confirm the enhanced release of Ru(II) complexes at lower pH values, which is further supported by inductively coupled plasma optical emission spectrometry (ICP-OES) measurements. Furthermore, the cytotoxicity effect of the released metallotherapeutics is evaluated in vitro on metastatic B16F1 melanoma cells and enhanced cancer cell-killing efficacy is demonstrated upon exposure of the nanomaterials to weakly acidic conditions. The obtained results showcase the promising capabilities of the designed MSN nanocarriers for the pH-responsive delivery of metallotherapeutics and targeted treatment of cancer.


Introduction
Cancer treatments typically cause a range of side effects [1,2], induced by poor selectivity for cancerous over healthy cells, and researchers are making a significant effort to develop methodologies for controlled and site-specific drug delivery to cancer [3,4]. Towards this end, the application of nanomaterials is seen as very encouraging [5,6], particularly due to the enhanced permeability and retention (EPR) effect, which enables selective accumulation of nanoparticles at tumour tissues [7]. Moreover, some nanomaterials, such as mesoporous silica nanoparticles (MSN) allow the employment of additional cancertargeting modalities, through devising nanocarriers with stimuli-responsive drug delivery capabilities [8]. Further beneficial attributes of MSN include their large surface area, uniform mesoporosity, tunable morphology, facile surface functionalization and proven biocompatibility [9][10][11].

Synthesis of MSN and Functionalization with Ligands
MSN were obtained by the sol-gel templating method. In brief, CTAB (1 g) was dissolved in deionized water (480 mL) with the addition of NaOH (2 M, 3.5 mL) at 80 • C. Then, TEOS (5 mL) was added dropwise under vigorous stirring. The reaction was continued at 80 • C for 2 h at 500 rpm. The material was collected by filtration and washed twice with water and once with ethanol. After drying at 80 • C, it was calcinated at 500 • C (ramp 2 deg/min) to remove the surfactant.
The amino-functionalized MSN was prepared through the grafting procedure. MSN (650 mg) was refluxed at 110 • C in anhydrous toluene containing APTES (0.65 mmol) overnight. The material (MSN-AP) was collected by filtration, washed twice with 2propanol, and dried at 80 • C.
The functionalization with PA was performed in an aqueous solution at pH 6.0. First, the carboxyl groups of PA (960 mg) were activated through stirring for 2 h at room temperature in a solution containing NHS (240 mg) and EDC (480 mg). Then, an aqueous dispersion (pH 6.0) containing MSN-AP (640 mg) was added to the above solution under stirring. The reaction was continued for 24 h at RT. The material (MSN-PA) was collected by filtration, washed with water and ethanol, and dried at 80 • C.
In order to obtain a pH-responsive hydrazone bond, 250 mg of MSN-PA was dispersed in 100 mL of ethanol containing 0.02 M of appropriate hydrazine (H1/H2), followed by the addition of 100 µL of TFA and 2 mL of TEA under stirring. The reaction solution was degassed with nitrogen for 30 min. Afterward, the reaction mixture was protected from light and refluxed at 80 • C for 24 h. The final products (MSN-H1 and MSN-H2) were obtained by filtration, washed with boiled ethanol, and dried at 80 • C. ) was stirred at room temperature for 48 h. In order to investigate the drug release kinetics, suspensions were centrifuged at 12,066× g rcf at certain time intervals (15 min, 30 min, 1 h, 2 h, 4 h, 24 h, and 48 h) to separate a supernatant (4 mL), which was analysed using UV/VIS spectrophotometer DSH-L6/L6S (Dshing Instrument Co., Zhuhai, China) by measuring absorbance in the range from 250 to 600 nm. After the measurement, materials were re-dispersed in the supernatant and returned to bulk suspension under stirring.

Determination of Cell Viability
B16F1 cells were maintained routinely using a complete medium (RPMI 1640, 10% FCS, 1% glutamine and 1% penicillin/streptomycin) with an atmosphere of 5% CO 2 at 37 • C. For cell viability assay, the cell suspension was prepared at a density of 5000 cells/100 µL.
The stock solutions of the tested materials were prepared at a concentration of 20 mg/mL in PBS (pH 7.2) or acetate buffer (pH 5.0). The stock solutions of prepared materials were incubated on a shaker for 4 h. The materials were tested at different concentrations, prepared by dilution of stock solutions in a completed medium, against mouse melanoma B16F1 cell line and incubated for 48 h at 37 • C and 5% CO 2 . Subsequently, MTT and CV assays were performed according to the literature [48]. The absorbance was measured using plate reader Spectramax (Molecular Devices, San Jose, CA, USA) at 570 and 670 nm. The cell viability is represented as a percentage compared to untreated cells and the mean calculated using a four-parametric logistic function.

Results and Discussion
The SEM of MSN ( Figure 2a) demonstrated that the synthesized material consists of uniform spherical nanoparticles with diameters in the range of 160 to 220 nm. Nitrogen sorption analysis (Figure 2b) revealed that the initial MSN material possesses a high BET surface area (1046 m 2 /g). The BET isotherms of non-functionalized MSN, as well as for MSN-PA, MSN-H1 and MSN-H2 displayed a typical type IV isotherm without an apparent hysteresis loop, confirming a narrow, uniform, and well-defined mesoporous structure. Though, with the introduction of functional groups and, particularly upon coordination of Ru(II) metal complex, the surface areas, total volumes of mesopores, average and BJH pore diameters decreased (Table 1), evidencing successful functionalization. After functionalization of MSN with PA, the surface area decreased without the influence on the total volume of mesopores, due to the small size of the PA molecule. Upon subsequent functionalization with larger H1 and H2 molecules, the surface area but also the mesopore pore volumes decreased. Further modification of the functional groups by coordination of the Ru(II) complex, leads to the change of the BET isotherm to type II, also followed by a decrease in the BJH pore diameter below 2 nm.
These results indicate blocking of the MSN mesopores after coordination of the Ru(II) complex, which is in agreement with our previous observation that Ru(II) complexes are indeed capable of capping the MSN pores, which was also utilized for entrapping the cargo molecules and their subsequent release by exposure to visible light [40,49]. Small angle X-ray scattering measurements evidence the presence of hexagonally ordered porosity, typical for MCM-41-based MSN, with the following peak positions: (100) at 2.35 degrees (2θ), (110) at 4.17 degrees (2θ) and (200) at 4.70 degrees (2θ) (Figure 2d). Covalent surface functionalization lead to shifting of the peaks to higher Bragg angles due to decreasing pore size and decreased intensity of higher reflections ( (110) and (200)) due to disruption of symmetry upon covalent surface modifications. These results indicate blocking of the MSN mesopores after coordination of the Ru complex, which is in agreement with our previous observation that Ru(II) complexes indeed capable of capping the MSN pores, which was also utilized for entrapping cargo molecules and their subsequent release by exposure to visible light [40,49]. Sm angle X-ray scattering measurements evidence the presence of hexagonally ordered rosity, typical for MCM-41-based MSN, with the following peak positions: (100) at 2 degrees (2θ), (110) at 4.17 degrees (2θ) and (200) at 4.70 degrees (2θ) (Figure 2d). Coval surface functionalization lead to shifting of the peaks to higher Bragg angles due to creasing pore size and decreased intensity of higher reflections ( (110) and (200)) due disruption of symmetry upon covalent surface modifications. FTIR spectroscopy (Figure 3a and Figure S1), thermogravimetric analysis (Fig  3b,c) and zeta potential measurements were further employed to evidence the success surface functionalization. All   FTIR spectroscopy (Figure 3a and Figure S1), thermogravimetric analysis (Figure 3b,c) and zeta potential measurements were further employed to evidence the successful surface functionalization. All FTIR spectra are dominated by bands at 441 cm −1 (Si-O rocking vibration), 809 cm −1 (internal Si-O-Si symmetric stretching vibration) and 1062 cm −1 (internal Si-O-Si asymmetric stretching vibration) [50,51]. In the region 1300 to 1800 cm −1 non-functionalized MSN and MSN[Ru] material, which contains the surface adsorbed Ru(II) complex, exhibit similar spectra, with the dominant band at 1640 cm −1 , characteristic for stretching vibration of surface adsorbed water. This result hints that a small amount of Ru(II) precursor was adsorbed on the MSN surface in the absence of any functionalization and, therefore, no significant changes in the spectra are observed. After grafting with APTES a new band appeared at 1595 cm −1 (Figure 3a), assigned to N-H asymmetric bending vibration. The appearance of a new intense band in the region 1600-1700 cm −1 in the spectrum of MSN-PA can be ascribed to C=O stretching vibration and evidenced successful functionalization with pyruvic acid. Further functionalization of nanomaterials with H1 and H2 leads to the decrease in the carbonyl group vibration The introduction of functional groups onto the surface of MSN was also revealed by thermogravimetric analysis (Figure 3b,c). All the materials show weight loss below 150 °C due to surface-adsorbed water. The weight loss patterns between 150 °C and 750 °C show different bands for different materials, indicating the presence of different surface moieties and successful surface functionalization. Moreover, the amounts of functional group grafted with respect to the starting MSN were 9.87 wt %, 11.22 wt %, 12.45 wt %, and 13.14 wt % in the case of MSN-AP, MSN-PA, MSN-H1, and MSN-H2, respectively. A noticeable agreement in the rise of weight loss with every synthesis step is evidenced.  Chlorine was also quantified in the atomic ratio Ru:Cl ca. 1:2 (Table S1), which is in agreement with the suggested structure of the coordinated Ru(II) complex. EDS chromatograms are provided in Supplementary Materials (Figure S3).
The  As can be noted, the release kinetics were clearly pH-dependent with enhanced cargo release upon acidification of the environment. . This result evidences the crucial role of surface-bound ligands for constructing efficient delivery systems and for achieving the pH-responsive release of these types of Ru(II) complexes. Furthermore, by comparing the release kinetics, it is evident that the drug release from MSN[Ru] reaches its maximum within one hour of measurements, while for the other two materials the release kinetics is evidently slower, reaching the plateau only after 4 h. As the release of Ru(II) complex in the case of MSN[Ru] occurs rapidly due to simple desorption of the adsorbed species, the slower release kinetics in the case of MSN-H1[Ru] and MSN-H2[Ru] supports the assumption that these release processes are governed by a more complex mechanism than desorption, such as the process of hydrolysis of the hydrazone linkages. The measurements at pH 6.0 reveal a stronger initial burst of the cargo release after 2 h of measurements, followed by the decrease in the measured absorbances. This result is not observed at pH 5.0 and may be related to the reversibility of the hydrazone formation [52], which is less favored at lower pH.
UV/VIS spectra of Ru(II) complexes in supernatants after the release kinetics measurements ( Figure S4) exhibited different bands, which were highly dependent on the pH values. This result hints at possible hydrolysis and substitution of ligands, giving rise to different possible mononuclear and binuclear Ru(II) complexes coexisting in solution, containing different combinations of Cl, OH and H 2 O ligands [53]. Hence, as the UV/VIS spectra change with pH, comparison of absorbances at the same wavelength may not give a reliable estimation of the concentration of the released Ru(II) complexes due to the probable differences in extinction coefficients at 410 nm. However, the final released amounts of the ruthenium were quantified from the solution by ICP-OES, after 48 h of stirring in solutions of different pH values (Table 2), and the results of this analysis also evidence the beneficial effects of acidification on the release of Ru(II) complexes. To evaluate the potential of prepared nanomaterials for cancer treatment, in vitro cell viability experiments were performed against B16F1 melanoma cell lines. The materials were first incubated for 4 h in buffers at pH 5.0 and 7.2 and then different dilutions of the materials in the medium were prepared for treating the cells for 48 h. The half maximal inhibitory mass concentration (MC 50 ) values of Ru(II)-functionalized MSN, calculated as a mass concentration (µg/mL) of the Ru(II)-containing MSN needed to inhibit the cell viability by 50%, are listed in Table 3, while dose-dependent results of B16F1 cells treated with [Ru] immobilized on MSN are shown in Figure 5.  As can be seen, the investigated materials showed high activity upon pre-incubati of materials in an acidic environment (pH 5.0), while after pre-incubation at pH 7.2 ma rials were found inactive against B16F1 cells. This substantial difference can be eviden associated with the cleavage of hydrazone bonds and drug release differences from t tested materials. Results from the MTT and CV assays indicate that both nanomateri are slowing the metabolic profile of the cells. Comparing to the MC50 values of mesop rous silica loaded with cisplatin (CV assay: MC50 = 1.23 ± 0.13 μg/mL), materials report herein, preincubated for 4 h at pH 5.0, showed two times higher potential against B16 cells [54].  Figure S5). Furthermore, previous research showed that sta ing Ru(II) precursor for the preparation of MSN-metallotherapeutics did not exhibit cy toxicity against different cell lines, such as human colon adenocarcinoma (Colo205 and multidrug-resistant counterpart Colo320), as well as human embryonal lung fibrobla cell line (MRC-5) [55]. Such results strengthened our conviction that functionalization MSN with coordination-capable ligands, such as H1 and H2, improves the loading capa ity, but also enhances the cytotoxic activity of Ru(II) metallotherapeutic though pHsponsive release of H1-and H2-containing Ru(II) complexes.

Conclusions
In summary, we constructed two types of mesoporous silica nanoparticle-bas nanocarriers, containing surface-attached ligands and coordinated Ru(II)-based metal therapeutic. The ligands were attached to MSN through a pH-responsive hydrazone lin age and the enhanced release of Ru(II) complexes was successfully achieved at weak acidic conditions in comparison to the release at physiological pH. In Vitro evaluation the prepared materials against B16F1 cells evidenced their potent anticancer activity up exposure to weakly acidic conditions, which is encouraging toward further investigati in utilization of functionalized MSN as novel cancer-targeting nanotherapeutics for p responsive delivery of cytotoxic Ru(II) complexes. As can be seen, the investigated materials showed high activity upon pre-incubation of materials in an acidic environment (pH 5.0), while after pre-incubation at pH 7.2 materials were found inactive against B16F1 cells. This substantial difference can be evidently associated with the cleavage of hydrazone bonds and drug release differences from the tested materials. Results from the MTT and CV assays indicate that both nanomaterials are slowing the metabolic profile of the cells. Comparing to the MC 50 values of mesoporous silica loaded with cisplatin (CV assay: MC 50 = 1.23 ± 0.13 µg/mL), materials reported herein, preincubated for 4 h at pH 5.0, showed two times higher potential against B16F1 cells [54].  Figure S5). Furthermore, previous research showed that starting Ru(II) precursor for the preparation of MSN-metallotherapeutics did not exhibit cytotoxicity against different cell lines, such as human colon adenocarcinoma (Colo205 and its multidrug-resistant counterpart Colo320), as well as human embryonal lung fibroblast cell line (MRC-5) [55]. Such results strengthened our conviction that functionalization of MSN with coordinationcapable ligands, such as H1 and H2, improves the loading capacity, but also enhances the cytotoxic activity of Ru(II) metallotherapeutic though pH-responsive release of H1-and H2-containing Ru(II) complexes.

Conclusions
In summary, we constructed two types of mesoporous silica nanoparticle-based nanocarriers, containing surface-attached ligands and coordinated Ru(II)-based metallotherapeutic. The ligands were attached to MSN through a pH-responsive hydrazone linkage and the enhanced release of Ru(II) complexes was successfully achieved at weakly acidic conditions in comparison to the release at physiological pH. In Vitro evaluation of the prepared materials against B16F1 cells evidenced their potent anticancer activity upon exposure to weakly acidic conditions, which is encouraging toward further investigation in utilization of functionalized MSN as novel cancer-targeting nanotherapeutics for pH-responsive delivery of cytotoxic Ru(II) complexes.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/pharmaceutics13040460/s1, Figure S1: Full range FTIR spectra of the synthesized materials, Figure S2: Particle size distribution of Ru-modified nanoparticles in water (top) and culture medium (bottom), Table S1: Variation of ruthenium and chloride concentrations from EDS measurement, Figure