International Journal of Molecular Sciences Article Toxicological Profiling and Long-Term Effects of Bare, PEGylated- and Galacto-Oligosaccharide-Functionalized Mesoporous Silica Nanoparticles Irene Barguilla 1,† , Vicente Candela-Noguera 2,3,† , Patrick Oliver 1 , Balasubramanyam Annangi 1 , Paula Díez 2,3,4,5 , Elena Aznar 2,3,4,5, * , Ramón Martínez-Máñez 2,3,4,5 , Ricard Marcos 1 , Alba Hernández 1, * and María Dolores Marcos 2,3,4,5 1 Grup de Mutagènesi, Departament de Genètica i de Microbiologia, Facultat de Biociències, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Spain; ricard.marcos@uab.cat (R.M.) 2 Instituto Interuniversitario de Investigación de Reconocimiento Molecular y Desarrollo Tecnológico, Universitat Politècnica de València, Universitat de València, 46022 Valencia, Spain 3 Unidad Mixta UPV-CIPF de Investigación en Mecanismos de Enfermedades y Nanomedicina, Universitat Politècnica de València, Centro de Investigación Príncipe Felipe, 46012 Valencia, Spain 4 Unidad Mixta de Investigación en Nanomedicina y Sensores, Instituto de Investigación Sanitaria La Fe (IISLAFE), Universitat Politècnica de València, 46026 Valencia, Spain 5 CIBER de Bioingeniería, Biomateriales y Nanomedicina (CIBER-BBN), Instituto de Salud Carlos III, 28029 Madrid, Spain * Correspondence: elazgi@upvnet.upv.es (E.A.); alba.hernandez@uab.cat (A.H.) † These authors contributed equally to this work. Abstract: Mesoporous silica nanoparticles (MSNs) are amongst the most used nanoparticles in biomedicine. However, the potentially toxic effects of MSNs have not yet been fully evaluated, being a controversial matter in research. In this study, bare MSNs, PEGylated MSNs (MSNs-PEG), Citation: Barguilla, I.; and galacto-oligosaccharide-functionalized MSNs (MSNs-GAL) are synthesized and characterized Candela-Noguera, V.; Oliver, P.; to assess their genotoxicity and transforming ability on human lung epithelial BEAS-2B cells in Annangi, B.; Díez, P.; Aznar, E.; short- (48 h) and long-term (8 weeks) exposure scenarios. Initial short-term treatments show a Martínez-Máñez, R.; Marcos, R.; dose-dependent increase in genotoxicity for MSNs-PEG-treated cells but not oxidative DNA damage Hernández, A.; Marcos, M.D. for MSNs, MSNs-PEG, or for MSNs-GAL. In addition, after 8 weeks of continuous exposure, neither Toxicological Profiling and induced genotoxic nor oxidative DNA is observed. Nevertheless, long-term treatment with MSNs- Long-Term Effects of Bare, PEG and MSNs-GAL, but not bare MSNs, induces cell transformation features, as evidenced by PEGylated- and Galacto- the cell’s enhanced ability to grow independently of anchorage, to migrate, and to invade. Further, Oligosaccharide-Functionalized the secretome from cells treated with MSNs and MSNs-GAL, but not MSNs-PEG, shows certain Mesoporous Silica Nanoparticles. Int. tumor-promoting abilities, increasing the number and size of HeLa cell colonies formed in the J. Mol. Sci. 2023, 24, 16158. https:// doi.org/10.3390/ijms242216158 indirect soft-agar assay. These results show that MSNs, specifically the functionalized ones, provoke some measurable adverse effects linked to tumorigenesis. These effects are in the order of other Academic Editor: Soo-Jin Choi nanomaterials, such as carbon nanotubes or cerium dioxide nanoparticles, but they are lower than Received: 16 October 2023 those provoked by some approved drugs, such as doxorubicin or dexamethasone. Revised: 3 November 2023 Accepted: 3 November 2023 Keywords: mesoporous silica nanoparticles; long-term effects; toxicological profiling Published: 10 November 2023 1. Introduction Copyright: © 2023 by the authors. Licensee MDPI, Basel, Switzerland. The applications of nanomaterials are widely and rapidly growing. The interest in This article is an open access article their use in biomedicine has increased greatly in the past years as such materials embody distributed under the terms and an attractive framework where new tools for imaging, diagnostic, and therapeutic purposes conditions of the Creative Commons are flourishing [1]. Within this field, mesoporous silica nanoparticles (MSNs) are rising Attribution (CC BY) license (https:// as some of the most promising and customizable drug delivery systems [2,3]. MSNs are creativecommons.org/licenses/by/ formed from silicon dioxide arranged to obtain spherical particles with tunable pores from 4.0/). 2 to 50 nm, which can be loaded with selected cargos such as fluorophores, dyes, drugs, Int. J. Mol. Sci. 2023, 24, 16158. https://doi.org/10.3390/ijms242216158 https://www.mdpi.com/journal/ijms Int. J. Mol. Sci. 2023, 24, 16158 2 of 14 and enzymes or small peptides [4]. Moreover, MSNs can turn into stimuli-responsive delivery systems by capping the pores with different molecules that act as gatekeepers (also known as molecular gates) allowing cargo delivery only when specific stimuli are present [5,6]. Such systems have been used not only for drug delivery purposes [7–10] but also for chemical communication strategies [11] and sensing [12,13]. As MSNs gain popularity for potential biomedical applications, their biocompatibility has been often evaluated mostly analyzing different short- to mid-term endpoints. Gener- ally, MSNs are described as biocompatible [3]. In fact, silica and its degradation products have been ‘generally recognized as safe’ (GRAS) by regulatory agencies such as the FDA and silica is used in cosmetics and in the food industry as an additive [14]. Nevertheless, the potential for MSNs to cause adverse effects in biomedical applications still remains con- troversial. In many in vitro and in vivo studies, MSNs revealed no cytotoxicity, including histopathological evaluation, even when relatively high concentrations were used [15,16]. Nonetheless, some studies found a correlation of MSN uptake and in vitro cytotoxicity, inducing ROS production which can damage cells and tissues, either inducing necrosis or apoptosis, inflammatory processes, and genotoxicity [17–20]. This cytotoxicity and ROS production have been suggested to be associated with silanol abundance on the surface of MSNs, which can disrupt the membrane components, the cytoskeleton, and nuclei components [21]. For instance, a recent in vitro finding demonstrated that bare MSNs could induce apoptosis, ROS, and DNA damage at high doses and repeated exposures in human corneal epithelial cells [22]. Another study described that MSNs were involved in intestinal damage in mice after sub-acute exposures, as well as ROS production, inflammasome activation, apoptosis, and loss of autophagy flux in intestinal epithelial cells [23]. Cardiac and pulmonary toxicity due to increased ROS levels and inflammation after MSN exposure have also been described [24]. Moreover, it has been widely reported that surface functionalization of MSNs can significantly increase, diminish, or even prevent adverse effects [25]. In fact, function- alization of MSNs greatly affects the chemical properties of the surface, such as the net surface charge and polarity, which influences the interactions with cells, the adsorption of proteins, and the nanoparticles’ stability [26]. For example, no toxicity was found for MSN functionalized with folate in cell culture models [27,28]. Another study analyzed the acute toxicity of MSNs in immune-competent mice, establishing the highest maximum tolerated dose of amino-functionalized MSNs at 100–150 mg/kg and non-functionalized MSNs at 30–65 mg/kg [29], suggesting that they have a low toxicity. In addition, nanoparticles’ morphology can also affect some cell functions such as apoptosis, adhesion, migration, and proliferation, which are related to cytoskeleton organization [30]. The contradictory results found in the literature in relation to the toxicity of MSNs bring out the necessity to carry out additional thorough biocompatibility studies. These inconsistent conclusions are evidenced even more in the long-term effects of MSNs. While some evaluations did not find acute or chronic toxicity or inflammation in vivo after treatment with MSNs for a month [31] or one year [32], other authors claimed that the low degradability of MSNs in vivo could lead to long-term toxicity in various tissues or cause diseases [33,34]. Another study with in vivo models reported the toxicity of submicron MSNs with acute exposure (10 days), but this was considerably alleviated with chronic exposure (180 days) [35]. Nonetheless, there is a remarkable lack of studies in which the long-term effects of MSN exposure are analyzed. Based on the above, and to advance our understanding of MSN toxicity, we report herein in vitro short- and long-term toxicity and genotoxicity studies on human bronchial epithelial cells (BEAS-2B) using three different MSNs: (i) bare MSNs (MSNs), (ii) PEGylated MSNs (MSNs-PEG), and (iii) galacto-oligosaccharide-functionalized MSNs (MSNs-GAL). MSNs-PEG [36] and MSNs-GAL [37] have already been used as drug delivery systems in in vitro and in vivo studies. Int. J. Mol. Sci. 2023, 24, 16158 3 of 14 2. Results 2.1. Synthesis and Characterization of Materials Bare MSNs, MSNs-PEG and MSNs-GAL were synthetized following the protocols described in the experimental section and were characterized using standard techniques. Powder X-ray diffraction (PXRD) patterns of the bare MSNs (as-made and calcined) are shown in Supplementary Figure S1A. The PXRD patterns display a typical MCM-41 hexago- nal arrangement of pores, whose peaks can be indexed as (100), (110), (200), and (210) Bragg reflections. In addition, it is observable that there is a right drift of peaks in calcined MSNs compared to as-made MSNs. This fact can be attributed to an approximate cell contraction produced by the reorganization of the mesoporous structure and further condensation of the silanol groups due to the surfactant removal process in the calcination process. N2 adsorption−desorption isotherms of bare MSNs were recorded (Supplementary Figure S1B), in which a typical curve for MCM-41-like mesoporous solids was obtained, corresponding to a type IV isotherm. This indicates nitrogen condensation inside the mesopores by capillarity with an acute adsorption step at intermediate P/P0 values (0.3–0.4). Moreover, the minimum hysteresis loop confirms the proper formation of uniform cylindrical mesopores, in which nitrogen freely adsorbs and desorbs. Using the adsorption curve and the Barrett-Joyner-Halenda (BJH) model, a narrow pore distribution centered at 2.88 nm and a pore volume of 0.944 cm3 g−1 were calculated. Furthermore, applying the Brunauer-Emmett-Teller (BET) model resulted in a value of 1118.6 m2 g−1 for the relative surface area of the calcined MSNs. In addition to mentioned data, a second feature can also be observed at high relative pressures (P/P0 > 0.9), corresponding to the filling of the interparticle space. Regarding the gated MSNs, the isotherms showed a reduction in capillary condensation because the pores were filled by dyes. The pore volume and surface area were reduced to ca. 0.1 cm3 g−1 and ca. 150 m2 g−1 , respectively, which was in agreement with reported data for MSNs-PEG and MSNs-GAL [36,37]. The mesoporous structure of bare MSNs, MSNs-PEG and MSNs-GAL was also con- firmed by transmission electron microscopy (TEM) analysis. Representative images in Supplementary Figure S2 confirmed that the solids were obtained as spherical particles with a diameter of around 82 nm ± 15 nm. Moreover, the MCM-41-like hexagonal structure of the mesopores can also be observed in the samples as pores and channels. Furthermore, the hydrodynamic diameters of the MSNs, MSNs-PEG, and MSNs-GAL were determined by dynamic light scattering (DLS) (Supplementary Figure S3). The size distribution of the hydrodynamic diameter obtained for MSNs was centered at 148 nm, whereas that found for MSNs-PEG and MSNs-GAL were centered at 245 nm and 300 nm, respectively. These results agree with the functionalization of the MSNs with PEG and GAL, which forms an organic layer around the nanoparticle increasing the hydrodynamic diameter of the material. The surface charge of the nanoparticles was monitored measuring their ζ potential, whose results were −37.1 mV for bare MSNs, −26.4 mV for MSNs-PEG, and −12.3 mV for MSNs-GAL. The negative surface charge of MSNs can be explained due to the presence of silanol groups on their surface, which can be deprotonated and form silanolate anions. When nanoparticles are functionalized, their negative charge decreases due to two processes: the reaction of silanol groups with alkoxysilanes and the shielding of the charge of free silanol groups by the anchored oligomers, i.e., PEG and GAL, which have no charge. The characterization results are summarized in Table 1. The organic contents in the solids were also determined by thermogravimetric studies (Supplementary Figure S4). As summarized in Table 1, thermogravimetric data show that organic content increases with cargo loading and functionalization. Considering that the loss of mass in MSNs due to silanol condensation is ca. 3.5%, the organic contents are 23.1% and 18.2% for MSNs-PEG and MSNs-GAL, respectively. Table 1. Characterization data of MSNs and gated MSNs under TEM, Zetasizer, and thermogravi- metric analysis. Int. J. Mol. Sci. 2023, 24, 16158 4 of 14 MSNs MSNs-PEG MSNs-GAL Size (nm) (TEM) 81.7 ± 11.2 82.4 ± 10.5 80.9 ± 14.0 Size (nm) (DLS) 148 245 300 Characterization Table 1.PDI data of MSNs(DLS) (width of the distribution) and gated MSNs 0.242under TEM, Zetasizer, and0.215 0.307 thermogravi- metric analysis. ζ Potential (mV) −31.7 ± 5.6 −23.1 ± 6.2 −12.3 ± 4.37 Organic content (%) – 23.1 18.2 MSNs MSNs-PEG MSNs-GAL Size (nm) (TEM) 81.7in±the The organic contents 11.2solids were also 82.4 ± 10.5 by thermogravimetric determined 80.9 ± 14.0stud- Size (nm) (DLS) ies (Supplementary Figure148 S4). As summarized in Table 1, thermogravimetric300 245 data show PDI (width of the distribution) (DLS) 0.242 that organic content increases with cargo loading and 0.307 0.215 that functionalization. Considering ζ Potential (mV) −31.7 ± 5.6 −23.1 ± 6.2 −12.3 ± 4.37 the loss of mass in MSNs due to silanol condensation is ca. 3.5%, the organic contents are Organic content (%) – 23.1 18.2 23.1% and 18.2% for MSNs-PEG and MSNs-GAL, respectively. Nanoparticles MSNs-PEG and MSNs-GAL were loaded with cargo (SafO and RhoB, Nanoparticlesto correspondingly) MSNs-PEG demonstrate andtheMSNs-GAL wereability correct capping loadedofwith cargoand the PEG (SafO GAL and RhoB, func- correspondingly) to demonstrate the correct capping ability of the PEG and tionalization. The controlled release performance of the gated materials was evaluated GAL func- tionalization. The controlled release performance of the gated materials was following the methodology described in the Supplementary Information (release assays). evaluated following theassays The release methodology described were monitored forin24the Supplementary h. Both MSNs-PEG and Information MSNs-GAL (release assays). showed a The specific release of their cargo just in the presence of the corresponding stimulus (GSH and a release assays were monitored for 24 h. Both MSNs-PEG and MSNs-GAL showed specific βGAL,release of theirand respectively) cargo just in therelease a negligible presence of the corresponding in absence stimulus (GSH of them (Supplementary Figureand S5). respectively) and a negligible release in absence of them (Supplementary Figure S5). βGAL, 2.2. Acute 2.2. AcuteMSNs MSNsExposure Exposure Does Does Not Lead to Not Lead toToxic ToxicEffects EffectsininBEAS-2B BEAS-2BCells Cells TheTheanalysis analysisof ofcell cell viability after aa48 viability after 48hhshort-term short-termexposure exposure to to bare bare MSNs, MSNs, MSNs- MSNs- PEG, and MSNs-GAL shows a survival rate over 80% for the selected PEG, and MSNs-GAL shows a survival rate over 80% for the selected doses, ranging fromdoses, ranging from 5 to 5100 to 100 µg/mL. µg/mL. No differential No differential effect ofeffect of the nanoparticles’ the nanoparticles’ coating coating was was(Figure observed observed 1A). Therefore, (Figure a non-cytotoxic 1A). Therefore, dose of dose a non-cytotoxic 10 µg/mL of 10was selected µg/mL wasfor the three selected forsolids to con- the three solids totinue forward continue with awith forward chronic BEAS-2B a chronic exposure. BEAS-2B exposure. Figure 1. Cytotoxicity and genotoxicity evaluation of bare and gated MSNs on BEAS-2B cells. Figure 1. Cytotoxicity and genotoxicity evaluation of bare and gated MSNs on BEAS-2B cells. (A) (A)Cell Cell viability viability 48after 48 h h after the treatment the treatment with with different different doses doses of MSNs, of MSNs, MSNs-PEG, MSNs-PEG, and MSNs- and MSNs-GAL. GAL. (B) Genotoxic (B) Genotoxic and oxidative and oxidative DNADNA damagedamage induced induced by MSNs, by MSNs, MSNs-PEG, MSNs-PEG, and MSNs-GAL and MSNs-GAL at dif- at ferent concentrations different after4848h hofofshort-term concentrationsafter short-termtreatment. treatment.(C)(C) Long-term Long-term genotoxic andand genotoxic oxidative oxidative DNAdamage DNA damageinduced induced byby MSNs, MSNs, MSNs-PEG, MSNs-PEG,and andMSNs-GAL MSNs-GAL at 10 µg/mL at 10 concentration µg/mL after 8after concentration weeks of continuous exposure. Data represented as mean ± SEM (n = 3), one-way ANOVA with 8 weeks of continuous exposure. Data represented as mean ± SEM (n = 3), one-way ANOVA with Dunnett’s post-test for exposed vs. unexposed or passage-matched controls. ** p < 0.01, * p < 0.05. Dunnett’s post-test for exposed vs. unexposed or passage-matched controls. ** p < 0.01, * p < 0.05. 2.3. MSNs Can Induce Slight Genotoxic Damage with Short-Term Exposure While This Effect Is Not Observed in Long-Term Exposed Cells The genotoxic damage when cells are treated with nanoparticles is shown in Figure 1B for acute exposure (48 h at high doses) and in Figure 1C for chronic exposure (8 weeks). Regarding the acute exposure, neither MSNs nor the coated particles induce any oxidative DNA damage (ODD), with the exception of MSNs-GAL (100 µg/mL) where a slight increase was observed. In addition, bare particles do not lead to genotoxic damage, and there is barely any in the case of MSNs-GAL-exposed cells. Nevertheless, we have found that there is a slight dose-dependent increase in MSNs-PEG-induced genotoxicity. On the other hand, after 8 weeks of chronic exposure, no genotoxic or ODD was observed neither The genotoxic damage when cells are treated with nanoparticles is shown in Figure 1B for acute exposure (48 h at high doses) and in Figure 1C for chronic exposure (8 weeks). Regarding the acute exposure, neither MSNs nor the coated particles induce any oxidative DNA damage (ODD), with the exception of MSNs-GAL (100 µg/mL) where a slight in- Int. J. Mol. Sci. 2023, 24, 16158 crease was observed. In addition, bare particles do not lead to genotoxic damage, and 5 of 14 there is barely any in the case of MSNs-GAL-exposed cells. Nevertheless, we have found that there is a slight dose-dependent increase in MSNs-PEG-induced genotoxicity. On the other hand, after 8 weeks of chronic exposure, no genotoxic or ODD was observed neither for the non-capped nor the gated MSNs, suggesting that cells adapt to the presence of low for the non-capped nor the gated MSNs, suggesting that cells adapt to the presence of low doses of MSNs. doses of MSNs. 2.4. Gated MSNs Show a Distinctive Transforming Potential after Chronic Exposure 2.4. Gated MSNs Show a Distinctive Transforming Potential after Chronic Exposure To evaluate the effects of a longer exposure to the MSNs and assess the potential To evaluate the effects of a longer exposure to the MSNs and assess the potential transforming ability of these particles, different transformation biomarkers were evaluated transforming ability of these particles, different transformation biomarkers were evalu- after 8 weeks of chronic exposure. ated after 8 weeks of chronic exposure. NoNo remarkable remarkablechanges changeswerewere observed observed in inthe theproliferation proliferation rate rate of of thethe exposed exposed BEAS- BEAS- 2B2Bcells. Nevertheless, the anchorage-independent growth analysis in soft cells. Nevertheless, the anchorage-independent growth analysis in soft agar showed aagar showed a gentle increasedability gentle increased abilityto to form form colonies colonies for MSN-PEG- for the the MSN-PEG- and MSN-GAL-exposed and MSN-GAL-exposed cells cells when compared with the non-exposed controls and when compared with the non-exposed controls and those exposed to those exposed to bare bare MSNs MSNs (Figure (Figure 2A,B). No further effects in the colony size were observed 2A,B). No further effects in the colony size were observed (Figure 2C). (Figure 2C). Figure Figure2. 2.Long-term Long-termexposure to gated exposure to gatedMSNs MSNsincreases increasesthethe cells’ cells’ colony-forming colony-forming ability. ability. The rep- The repre- sentative images resentative images (A)(A) and andbar bargraphs graphsshowing showingthe thecolony colonynumber number (B)(B) andandthethe colony colonysizesize (C) (C) of of BEAS-2B cells treated with MSNs, MSNs-PEG, and MSNs-GAL compared to untreated BEAS-2B cells treated with MSNs, MSNs-PEG, and MSNs-GAL compared to untreated controls after controls af- ter 8 weeks of long-term exposure. (C) Data represented as mean ± SEM (n = 3), one-way ANOVA 8 weeks of long-term exposure. (C) Data represented as mean ± SEM (n = 3), one-way ANOVA with with Dunnett’s post-test for exposed vs. passage-matched controls. * p < 0.05. Dunnett’s post-test for exposed vs. passage-matched controls. * p < 0.05. The evaluation of the migration (Figure 3A,B) and invasion (Figure 3C,D) potential The evaluation of the migration (Figure 3A,B) and invasion (Figure 3C,D) potential of the of the exposed cells consistently showed that the cells chronically exposed to PEG- or exposed cells consistently GAL-coated-MSNs wereshowed more able thattothe cellsthe cross chronically transwell exposed to PEG- or when compared GAL-coated- both to bare MSNs MSNs-exposed cells and non-exposed controls. The increase in the migrating cells is ca.cells were more able to cross the transwell when compared both to bare MSNs-exposed and non-exposed 30% and 65% when controls. The they are increase treated withinMSNs-PEG the migrating andcells is ca. 30%respectively. MSNs-GAL, and 65% when they In the arecase treated with of the MSNs-PEG invading cells,and the MSNs-GAL, increment is respectively. ca. 50% and In 90%thefor case of the invading MSNs-PEG cells, the and MSNs- increment is ca. 50% and 90% for MSNs-PEG and MSNs-GAL, respectively. GAL, respectively. 2.5. Chronically Exposed Cells Undergo Secretome Changes Leading to Cell Growth Promotion Cells chronically exposed to contaminants or other agents tend to undergo changes in their secretome. The indirect soft-agar assay evidenced that the secretome from MSNs and MSNs-GAL-exposed BEAS-2B cells was able to increase the colony growth of the already well-established tumoral HeLa cells (Figure 4A,B), and the colonies’ diameter slightly increases in these cases (Figure 4C). Int. J. Mol. Sci. 2023, 24, 16158 6 of 14 Int. J. Mol. Sci. 2023, 24, x FOR PEER REVIEW 6 of 15 Figure Figure The migration 3. migration 3. The and invasionand invasion potential potential of BEAS-2B of BEAS-2B cells increases cells increases when chronically exposed when chronically ex- to gated MSNs. The representative images (A) and histogram (B) depicting the proportion of BEAS- posed to gated MSNs. The representative images (A) and histogram (B) depicting the proportion 2B cells able to migrate after continuous exposure to MSNs, MSNs-PEG, and MSN-GAL compared ofpassage-matched to BEAS-2B cells ableforto8 weeks. control migrate after continuous The representative exposure images (C) to MSNs, and histogram MSNs-PEG, and MSN-GAL (D) revealing the invasion capability of BEAS-2B cells induced by MSNs, MSNs-PEG, and MSN-GAL compared to passage-matched control for 8 weeks. The representative images (C) and histogram in compar- ison to passage-matched control at 8 weeks. Data represented as mean ± SEM (n = 3), one-way (D) revealing ANOVA the invasion with Dunnett’s post-test for capability of BEAS-2B controls. exposed vs. passage-matched cells induced ** p < 0.01,by * p