Applied Surface Science 561 (2021) 150011

Contents lists available at ScienceDirect

Applied Surface Science

journal homepage: www.elsevier.com/locate/apsusc

Full Length Article

pH-responsive mesoporous silica drug delivery system, its biocompatibility

and co-adsorption/co-release of 5-Fluorouracil and Naproxen
Eva Beňová a, b, Virginie Hornebecq b, *, Vladimír Zeleňák a, *, Veronika Huntošová c,
Miroslav Almáši a, Mariana Máčajová d, David Bergé-Lefranc e
a
Department of Inorganic Chemistry, Faculty of Science, P.J. Šafárik University, Moyzesova 11, SK-041 54 Košice, Slovakia
b
Aix Marseille Univ, CNRS, Madirel, Marseille, France
c
Center for Interdisciplinary Biosciences, Technology and Innovation Park, P. J. Šafárik University in Košice, Jesenná 5, 041 54 Košice, Slovakia
d
Institute of Animal Biochemistry and Genetics, Centre of Biosciences, Slovak Academy of Sciences, Dúbravská cesta 9, 840 05 Bratislava, Slovakia
e
Aix Marseille Univ, CNRS, IMBE, Faculté de Pharmacie, Marseille, France

A R T I C L E I N F O A B S T R A C T

Keywords: A pH-responsive drug delivery system composed of N-(propyl)aniline (N-ANI) modified mesoporous silica
Drug delivery system nanoparticles as a carrier, and β-cyclodextrin (β-CD) as a cap was prepared and investigated. The system was
pH-responsive silica studied for co-adsorption and co-delivery of two different drugs; an anticancer drug 5-fluorouracil (5-FU), and
Co-adsorption
anti-inflammatory agent naproxen (NAP). Detailed adsorption study of the drugs themselves as well as in their
Co-delivery
Cytotoxicity
mutual mixture was performed. The drug release experiments were realized in simulated body fluid (pH = 7.4)
and slightly acidic medium (pH = 5) imitating the pH of tumor affected area. The experiments showed, that at
physiological pH 7.4, the prepared drug delivery system released no drug, at slightly acidic medium (pH = 5) the
β-CD caps were loosened, and drugs were released (88.5% and 98.7% for 5-FU and NAP, respectively). To
investigate the efficiency of the designed system for the delivery in biological environment, the biocompatibility
and cytotoxicity experiments were performed using U87 MG and SKBR3 cancer cells by fluorescence microscopy,
flow-cytometry, MTT. apoptotic and CAM assays.

1. Introduction internalized efficiently by alveolar cancerous cells and produce an

enhanced apoptotic effect in comparison with the single drug model [6].
The extensive uses of drugs, especially those targeted for the anti­ Co-delivery approach was studied either by co-delivery of two anti­
microbial and antineoplastic treatments, lead to various health prob­ cancer drugs or in combination of anticancer drug with another thera­
lems. One of the most important is resistance of surviving cells against peutic molecule showing anti-inflammatory, immunologic, antiallergic
the employed drugs, which drastically reduces the efficiency of che­ or anti-angiogenic properties [7]. It was demonstrated, that when
motherapeutics or antibiotics in further treatments. This phenomenon is doxorubicin (DOX) was loaded with an anti-angiogenic drug Com­
called multidrug resistance (MDR) [1,2]. MDR problem can be reduced bretastatin A4 (denoted as CA4) the DOX/CA4 loaded silica nano­
by co-delivery approach (or so-called combination therapy) using suit­ particles showed the clear inhibition of tumor growth in vivo although
able porous carriers like mesoporous silica nanoparticles (MSN) [3,4]. the in vitro apoptotic effect showed similar values as for single drug
Such systems can exert simultaneous and synergistic action on more loading [7].
than one critical metabolic pathway and can increase the apoptotic ef­ In the case of antibiotics, combination/co-delivery therapy is used
fect throughout the co-delivery of several drugs. It was shown that co- against polymicrobial infections or recalcitrant infections that are not
delivery of doxorubicin and camptothecin showed great potential for eradicated by standard treatment regimes. Indeed, this combination can
tumor-trigged drug release and for use in the synergistic chemotherapy broaden the spectrum of antibacterial activity and provide synergistic
of tumors like glioblastoma [5]. Liu et al. reported that the silica effects, e.g. if one drug stimulates the uptake of another. This behavior
nanoparticles prepared by co-loading of doxorubicin and paclitaxel, a was observed when vancomycin and polymyxin B where encapsulated
hydrophilic and a hydrophobic chemotherapeutics, could be and delivered using MSN [8]. Loading of both antibiotics had synergistic

* Corresponding authors.
E-mail addresses: virginie.hornebecq@univ-amu.fr (V. Hornebecq), vladimir.zelenak@upjs.sk (V. Zeleňák).

https://doi.org/10.1016/j.apsusc.2021.150011
Received 9 January 2021; Received in revised form 7 April 2021; Accepted 30 April 2021
Available online 8 May 2021
0169-4332/© 2021 Elsevier B.V. All rights reserved.
E. Beňová et al. Applied Surface Science 561 (2021) 150011

effect against Gram-negative bacteria, and the antimicrobial efficacy was further calcined to obtain the final mesoporous SBA-15 material.
was higher for MSN loaded with both antibiotics, compared to free an­ The surface functionalization was done by the post-synthetic grafting
tibiotics at the same concentration [8]. of SBA-15 using N-[3-(trimethoxysilyl)propyl] aniline (N-ANI) to form
In addition to scientific works mentioned above, the co-delivery amine-functionalized silica matrix. In a typical procedure, 1 g of pre-
approach using MSN was also applied in the co-delivery of antitumor heated SBA-15 (pre-heated to 150 ◦ C) was dispersed in anhydrous
drug paclitaxel (PTX) and anti-inflammatory, immunologic and antial­ toluene and mixed with 3 mL of N-ANI (12.5 mmol). The mixture was
lergenic drug, tetrandrine (TET) [9]. It was shown that silica particles refluxed under N2 for 20 h. The obtained product was centrifuged,
containing both PTX and TET suppressed tumor cells growth more several times washed with toluene and ethanol, and dried at 40 ◦ C for 24
efficiently than the delivery of PTX from silica particles alone, or the free h. The resulting product was denoted as SBA-15_N-ANI.
PTX alone. Moreover, the nanoparticles loaded with a PTX/TET
completely reversed the resistance of human mammary cells used [9].
It is obvious that the use of mesoporous silica particles for targeted 2.2. Drugs loading
drug delivery was intensively studied in recent years [10–13]. The
loading and release of drugs from the silica mesopores is a complex 2.2.1. Adsorption properties of silica supports
process since it depends on many parameters such as the different sol­ Antineoplastic agent, 5-Fluorouracil (5-FU) and anti-inflammatory
ubility of guest molecules in the solvent, the different diffusion rates agent Naproxen (NAP) were taken as model drugs to estimate the co-
throughout the pores and/or the strength of the interactions between adsorption and co-release performance of pH-responsive silica sup­
loaded molecules and silica nanoparticles. The first studies of the silica port. Initially, adsorption properties of silica materials SBA-15 and SBA-
as a drug delivery system focused on dissolution/diffusion mechanism of 15_N-ANI were studied with 5-FU and NAP separately. Adsorption was
delivery [10,11]. However, at present more sophisticated systems can be performed on both the SBA-15 and SBA-15_N-ANI for comparative
prepared, that hamper an undesired (premature) release of loaded purposes.
compounds and the release can be triggered on demand, using various Drugs were loaded into the supports by the impregnation method
physical or chemical stimuli [14–19]. Since the mesoporous silica par­ using drug solutions of known concentrations. The limiting factor was
ticles can deliver drugs inside the cells [20], such vectored delivery the solubility of 5-FU, which is sparingly soluble in water. Therefore, the
directly into cells, in combination with co-delivery approach and the highest possible concentration for 5-FU was set at 6.5 mg/mL (5.10− 2
release on demand, can partially solve the problem of above-mentioned M), as already discussed in our previous study [31]. For the adsorption
resistance (MDR), by the targeted killing of tumor cells. The first mes­ studies, the concentration of naproxen was therefore adjusted also to
oporous silica-based DDS was MCM-41 material [21]. After this dis­ this value and the highest concentration was the same for both drugs
covery, different silicas were tested for drug loading and release and, used.
among them, the SBA-15 silica has been extensively reported since its For the construction of adsorption isotherms, the materials SBA-15
two-dimensional hexagonal porous structure similar to MCM-41, but and SBA-15_N-ANI (0.02 g) were dispersed in a series of plastic vials
larger pore size (5–10 nm in diameter). The larger pore size makes SBA- containing 2 mL of aqueous solution of drug in growing initial concen­
15 material suitable for loading larger amounts of drugs as well as more trations of 0.13, 0.325, 0.65, 0.975, 1.3, 3.25, 6.5 mg/mL. Suspensions
bulky drug molecules or even macromolecules [22–26]. were stirred at 300 rpm for 24 h at 37 ◦ C while the evaporation of water
Considering the above-mentioned studies, the present work aims to was prevented. The suspensions were then centrifugated at 10 000 rpm
explore the possibility of co-adsorption and co-delivery of two different for 10 min; after centrifugation the supernatants were separated from
drugs, an anticancer drug (5-fluorouracil; 5-FU) and an anti- the solids and the impregnated solids were dried overnight at 40 ◦ C in an
inflammatory agent (Naproxen; NAP). In our previous works, we have oven. The samples with the 5-FU and NAP drugs separately loaded in
prepared and described the different mesoporous silica particles, that SBA-15 and SBA-15_N-ANI were denoted SBA-15_5-FU, SBA-15_N-
release drugs either on diffusion principle [27,28] or using external ANI_5-FU or SBA-15_NAP, SBA-15_N-ANI_NAP, respectively.
stimulus, e.g. light driven drug release [29,30] or pH driven drug release The co-adsorption process followed the same procedure as
[31]. In the latter, the system releasing the drug by decreasing pH is mentioned above. 0.02 g of the silica support was dispersed in a solution
especially suitable for the delivery of anticancer drugs. In the present of drugs mixture in different ratio, 5-FU:NAP 3.5:0.5, 3:1, 2.5:1.5, 2:2,
paper, we study the co-adsorption and co-release of 5-FU and NAP using 1.5:2.5, 1:3, 0.5:3.5 (final volume of the solution was 2 mL). The initial
MSN system composed of SBA-15 nanoparticles as a carrier, which were highest concentration of drug solution was fixed to 5.10− 2 M. The sub­
modified by β-cyclodextrin (β-CD) as a cap closing the carrier nanopores sequent procedure was the same as for simple adsorption. The observed
and preventing the premature release of the drugs. The properties of samples were denoted as SBA-15_5-FUN and SBA-15_N-ANI_5-FUN.
such carrier, its co-delivery characteristics and biocompatibility were Adsorbed amounts of the drugs were determined using UV spec­
evaluated by different biological assays like MTT, apoptosis and CAM troscopy at λ = 266 nm for 5-FU and λ = 330 nm for NAP. Before the
assay or cytotoxicity study. determination of the concentration of 5-FU and NAP, calibration curves
have been established based on solutions with different concentrations
2. Experimental section of 5-FU and NAP (see Figures S1, S2 in the Supplementary Information).
The linearity of the calibration curves was confirmed by the correlation
All chemicals used in the syntheses were obtained by Sigma-Aldrich coefficient r2 = 0.99996 for 5-FU and r2 = 0.99989 for NAP. Then, the
and Across Organics companies in the highest purities and used without adsorbed amount of drug, QADS, was calculated considering the differ­
further purification. Anhydrous solvents were obtained after standard ence between the initial concentration and the concentration of drugs in
procedures described in the reference [32] and stored over molecular the supernatant, as follows:
sieves. V(Ci − Ceq )
QADS = (1)
m
2.1. Synthesis and functionalization of silica SBA-15
where Ci (M) is the initial concentration, Ceq (M) is the equilibrium
SBA-15 silica was prepared by the template synthetic method using a concentration after 24 h at 37 ◦ C, V is the volume of the drug solution,
procedure reported by Zhao et al. [33]. Triblock copolymer Pluronic and m is the mass (g) of adsorbent used.
P123 was used as a structure reacting agent and tetraethoxysilane Microcalorimetric experiments were performed to determine the
(TEOS) as a silica source. The molar ratio of used reagents was 1 TEOS : adsorption enthalpies of drugs on the prepared supports surface. During
5.9 HCl : 193 H2O : 0.017 P123. The prepared as-synthesized material these experiments, solids (SBA-15 and SBA-15_N-ANI) were maintained

2
E. Beňová et al. Applied Surface Science 561 (2021) 150011

in water suspension using a stirring system and a stock solution of NAP the cells seeded in glass coverslip bottom Petri dishes (MatTek, USA)
and (NAP+5-FU) was added by 10 μL aliquots step by step (10 in­ were observed in the bright - field mode with the inverted LSM700
jections). Next, the heat flow peaks were time-integrated and after confocal microscope (Zeiss, Germany), equipped with a 20X Fluar (NA
correction of dilution effects, the integral enthalpies were obtained. = 0.75,∞, Zeiss, Germany) and a CCD camera (AxioCam HRm, Zeiss,
Germany).
2.2.2. Capping of loaded SBA-15_N-ANI_5-FUN nanoparticles by β-CD
A solutions of 5-FU and NAP were prepared in distilled water at 2.6. Cell viability - MTT assay
ambient temperature with the concentration of each 6.5 mg/mL (5.10− 2
M) and then mixed. In the next step 300 mg of the solid SBA-15_N-ANI The MTT (3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium
was added to the drugs mixture solution and stirred at 37 ◦ C for 24 h bromide, Sigma-Aldrich, Germany) assay was performed 24 and 48 h
until the equilibrium was achieved. Then, the pH of the suspension was after the administration of 5-FU (from 5.7 up to 57 µg/mL), NAP (from
adjusted to 7.4 and 800 mg of β-CD was added to the solution. The 45.6 up to 114 µg/mL), and SBA-15_N-ANI_5-FUN_ β-CD (from 0.2 up to
mixture was stirred overnight. Then, nanoparticles were centrifuged, 2 mg/mL). Two molar ratios (1:1 and 1:2) of 5-FU:NAP in SBA-15_N-
washed with water, and used in further experiments. At each further ANI_5-FUN_ β-CD were tested. The amount of drugs adsorbed were
step, the pH of the suspensions was carefully checked, and supernatants determined by TGA (see Figure S3 in the Supplementary information)
were analyzed by UV/Vis spectroscopy. The obtained sample was and by solid/liquid adsorption measurements (see Table T1 in the
denoted as SBA-15_N-ANI_5-FUN_ β-CD. Supplementary Information). The same protocol was kept for MTT-assay
[31]. The results were detected in the triplicates. The significance of
2.3. Release of the drugs difference was estimated with the student t-test with *p < 0.05, **p <
0.01 and ***p < 0.001.
The release of the drugs was checked at selected time intervals: 0.08
h, 0.33 h, 0.66 h, 1 h, 2 h, 3 h, 4 h, 5 h, 6 h and 24 h. During these time 2.7. Apoptotic assay
intervals, the mixtures were stirred gently at 37 ◦ C in a tube rotator. In
each defined time interval, the suspension was centrifuged and the su­ The cells (treated 48 h with 5-FU, NAP, and SBA-15_N-ANI_5-FUN_
pernatant was removed and analyzed. The released amounts of 5-FU or β-CD) were 15 min labelled with the mitochondria potential probe a
NAP from the samples, where they were loaded alone, or co-adsorbed MitoTrackerTMOrange CMTM/Ros (MTO, 0.2 μM, ThermoFisher­
from their mutual mixture 5-FU+NAP were calculated from calibra­ Scientific, USA) and an AnnexinV/FITC (FITC Annexin V Apoptosis
tion curves (see Figures S1 and S2 in the Supplementary Information). Detection Kit, BD PharmingenTM, Belgium) to detect phosphatidylser­
The 5-FU was recorded at the wavelength of 266 nm and NAP at 330 nm. ine in the plasma membrane. The cells were detached with the trypsin/
The absorbance of 5-FU and NAP at 266 nm and 330 nm does not ethylene-diaminetetraacetic acid (Gibco-Invitrogen, Life Technologies
overlay on each other. Ltd., France), centrifuged at 1200 rpm/10 min and resuspended in 0.5
mL of phosphate saline buffer (PBS) at pH = 7.4. The harvested cells
2.3.1. Release experiments from the supports containing single drug were measured with the flow cytometer (MACSQuant®Analyzer, Mil­
The release studies of 5-FU and NAP from samples SBA-15_N-ANI_5- tenyi, Germany) in B1 (525/50 nm) and B3 (655–730 nm) channel at the
FU and SBA-15_N-ANI_5-NAP, i.e. when each drug was encapsulated excitation 488 nm.
alone, were performed as pH dependence. For this purpose, the pH of a
saline solution, used as a release medium, was set to pH equal to 7.4 and 2.8. CAM assay
5.
Chorioallantoic membrane (CAM) of the fertilized Japanese quail’s
2.3.2. Release experiments from the supports containing co-adsorbed drugs (Coturnix japonica) eggs (a breeding colony of IABG SASci) was used for
To analyze the release of 5-FU and NAP from SBA_15_N-ANI_5-FUN_ in vivo biocompatibility study. The eggs were kept in an incubator (MIDI
β-CD, the particles were placed into the cuvette of spectrometer and 2 BIOS, Czech Republic) at 37 ◦ C and 50–60% relative humidity. The eggs
mL of saline solution (pH 7.4 or 5) was added carefully. were disinfected in a sterile condition with 70% ethanolic solution at
embryonic day 3 (ED3). The eggs were opened and the embryos were
2.4. Cell culture preparation placed into the 6-well tissue culture plates and cultivated in 37 ◦ C hu­
midified incubator (Memmert, Germany) until ED10. The silicon rings
The cell cultures were cultivated according to the standard propa­ were placed on the CAM. The volume of 30 µL of the studied solutions
gation protocols [34]. The U87 MG glioma cells (Cells Lines Services, was administered inside the ring. The image was detected with a digital
Germany) and SKBR3 human breast carcinoma cells (a gift from prof. camera (Canon EOS 6D II with Canon MP-E 65 mm f/2.8, Japan).
Pluckthun laboratory, University of Zurich, Switzerland) were grown in
the cell culture medium (high glucose, GlutaMAXTM supplement, py­ 2.9. Histology
ruvate Dulbecco’s modified Eagle medium, D-MEM, Gibco-Invitrogen,
Life Technologies Ltd., France) and RPMI 1640 (LM-R1638/500, bio­ The treated tissue of the CAM was fixed with 4% paraformaldehyde
sera, France), respectively in the dark at 37 ◦ C, 5% of CO2 and humid­ (Sigma-Aldrich, Germany) and separated. The paraffin sections with the
ified atmosphere. The complete cell culture medium was prepared with 5 µm thickness were prepared for histopathological analysis to deter­
10% fetal bovine serum (FBS, biosera, France). One day before experi­ mine the damage of the tissue (hematoxylin (BAMED, Slovakia) and
ments the cells were seeded into the 35 mm in diameter Petri dishes or eosin (BAMED, Slovakia) staining). The sections were evaluated with a
96-well plates and grown in the complete medium at pH = 7.3 or 6.3. light microscope Kapa 2000 (Kvant, Slovakia) at 10X magnification, and
The next day, the aliquots of the 5-FU, NAP, and SBA-15_N-ANI_5-FUN_ digital camera Nikon E995 (Japan).
β-CD were administered. The sterile powders of the substances were
dissolved in distilled water and sonicated 15 min before administration. 2.10. Characterization of porous solids with analytical techniques

2.5. Microscopy Nitrogen sorption measurements were carried out using an ASAP
2020 Micromeritics apparatus at − 196 ◦ C. Before adsorption, samples
The 24 h after the administration of 5-FU (5.7 and 57 µg/mL), NAP (~80–100 mg) were degassed at 120 ◦ C (40 ◦ C for grafted samples and
(114 µg/mL) and SBA-15_N-ANI_5-FUN_ β-CD (from 0.2 up to 2 mg/mL) samples containing drug molecules) during 12 h under the vacuum of

3
E. Beňová et al. Applied Surface Science 561 (2021) 150011

2.10− 3 mbar. The specific surface area was determined with Brunauer- with trapping drug molecules in pores while causing dissociation with
Emmet-Teller (BET) method, the pore size distribution was calculated the β-CD caps at pH < 6 [38].
from the adsorption branch using the Barrett-Joyner-Halenda (BJH) The delivery of doxorubicin was tested in the work of Bai et al. who
method [35]. constructed a pH-responsive controlled release system using p-anisidine
TEM micrographs were obtained using a JEOL 2000FX microscope. amine stalks and different CD (αand β) as the pore caps to regulate the
Samples were ground and suspended in methanol. The suspension was release of drug from silica channels. The authors studied how stalk
dropped on a carbon grid and dried in air. density and the type of CD used to affect release behavior. They found
UV–Visible spectroscopy measurements in the liquid phase were that the ability of CD to produce a better pH-responsive release was in
performed on a Varian Cary 300 spectrometer in the 200–400 nm range. the order β-CD > α-CD [39]. Therefore, based on these results in our
Thermogravimetric (TGA) measurements were carried out with a work we focused on β-CD system.
TGA Q500 apparatus (TA Instruments) using air as a carrier gas. The
samples with weight ~ 5–20 mg were treated from ambient temperature 3.1. Small angle X-ray diffraction
up to 800 ◦ C.
Small-angle X-Ray diffraction measurements were recorded on a X-ray diffraction patterns of unmodified/modified SBA-15 samples
Siemens D500R XRD diffractometer using Cu Kα radiation (λ = 1.54060 (see Fig. 2) match well with the typical XRD pattern reported for SBA-15
Å) in the 0.5 − 3◦ 2θ range with a 0.04◦ step associated with a step time silica. The pattern is associated with a highly ordered, two-dimensional
of 4 s. hexagonal mesostructure (p6mm symmetry) [33,40].
Microcalorimetry experiments were performed on TAM 2277 The X-ray diffraction pattern of SBA-15 silica is well resolved with a
microcalorimeter. strong diffraction peak at 2θ = 0.89 Å, and two weaker peaks at 2θ =
1.54 Å and 1.78 Å. They can be indexed as (1 0 0), (1 1 0) and (2 0 0) in
3. Results and discussion hexagonal lattice with a d(1 0 0) spacing 99.2 Å. The d-spacing corre­
sponds to the large unit cell parameter a0 = 114.5 Å as calculated ac­
In the present paper, we study the co-adsorption and co-release of 5- cording to the Bragg equation for hexagonal symmetry, a0 = 2 × d100/
FU and NAP using MSN system composed of amine grafted SBA-15 √3. The XRD patterns corresponding to SBA-15_N-ANI and SBA-15_N-
material and β-cyclodextrin (β-CD). For the preparation of the func­ ANI_5-FUN samples show only the (1 0 0) and (2 0 0) diffractions peaks
tional material, the mesoporous silica SBA-15 was surface modified by well resolved. The diffraction peak (1 1 0) has lower intensity due to
grafting procedure, during which silica surface hydroxyl groups reacted surface modification and drugs loading.
with N-[3-(trimethoxysilyl)propyl]aniline (see Scheme 1). During this
grafting procedure amine was covalently attached on the silica surface. 3.2. Nitrogen adsorption/desorption
This covalently attached amine represents an immobilized stalk
molecule covalently attached to silica surface and β-cyclodextrin rep­ To obtain more precise information about the structure and textural
resents a mobile cyclic molecule encircling the stalk via non-covalent properties of prepared samples, N2 adsorption/desorption measure­
interactions. The non-covalent interactions between stalk and cyclic ments were performed. The measurements were used to determine the
molecule β-CD are therefore crucial for pore blocking effect (see Fig. 1). specific surface area, pore volume, and pore size distribution of the
Such supramolecular nanovalves responsible for opening and closing the samples. The adsorption/desorption isotherms of the studied samples
pores of the carrier and preventing the premature release of the drugs. are shown in Fig. 3a. The isotherm of SBA-15 can be considered as a
When the pH of the environment is decreased from its initial values typical IV(a) type according to IUPAC classification, with well-defined
(~7.4), the molecules on the silica surface become protonated resulting H1 hysteresis loop associated with presence of mesopores [41]. The
that binding affinity with cyclodextrin is decreased. As it was already H1 hysteresis loop confirms open-ended hexagonal cylindrical pore ge­
demonstrated the β-CD caps are thus dispersed around from the stalks ometry typical for SBA-15 silica material. The steepness of the hysteresis
and pores are opened [36,37]. loop in the relative pressures range p/p0 = 0.6 – 0.8 shows on the well-
Different amines and cyclodextrin molecules have been demon­ ordered structure with narrow pore size distribution. The presence of
strated to be an effective supramolecular pore gatekeepers for meso­ micropores interconnecting mesoporous channels in the SBA-15 mate­
porous silica materials. For example Meng et al. reported a MSN delivery rial is reflected by an initial sharp nitrogen uptake at relative pressures
system based on the β-CD nanovalves and N-methyl benzimidazole below 0.05. When the surface of SBA-15 is modified with organic ligands
molecules as a stalks, which bind to the β-CD rings strongly at pH 7.4 and/or the pores are filled with drugs and capped with β-CDs, the whole
isotherm is shifted to lower relative pressures also the nitrogen adsorbed
volume is lower. Furthermore, the microporosity as well as the shape of
the hysteresis is changed upon modification and loading. The pore size
distributions, calculated from the adsorption branches of isotherms are
shown in Fig. 3b. Calculated specific surface areas, mesopore volumes
and sizes are summarized in Table 1.
The specific surface areas were calculated using the BET method. As
expected, the grafting of the N-ANI ligand leads to a decrease in specific
surface areas and total pore volumes. Functionalization of SBA-15 by N-
ANI ligand also downshifts the capillary condensation from p/p0 =
0.63–0.76 (pure SBA-15) to lower relative pressures p/p0 = 0.58–0.65
(SBA-15_N-ANI) reflecting the filling of the pores and decrease in pore
size from 6.5 nm (pure SBA-15) to 5.3 nm (SBA-15_N-ANI). The
adsorption of drugs 5-FU and NAP and β-CD capping caused the further
decrease in the specific surface area and total pore volumes for SBA-
15_N-ANI_5-FUN and SBA-15_N-ANI_5-FUN_ β-CD samples, as it is
obvious from Table 1.

Scheme 1. Grafting of N-[3-(trimethoxysilyl)propyl]aniline on the SBA-15

silica surface.

4
E. Beňová et al. Applied Surface Science 561 (2021) 150011

Fig. 1. Representation of non-covalent interaction between N-ANI stalks and cyclic β-CD molecule.

to determine the amount of N-ANI groups grafted on the surface and the
amount of the encapsulated drugs [27]. The results are shown in Fig. S4
in the Supplementary Information. Since the adsorbed moisture and
solvents loss occurs in the temperature range 25–150 ◦ C, all TGA curves
were normalized with a starting temperature of 150 ◦ C to evaluate mass
loss only due to decomposition of organic ligands and/or drugs. The
TGA curve of calcined SBA-15 shows that the sample is thermally stable,
and no significant mass loss is observed in the whole measured range.
For SBA-15 sample only a small weight loss (0.63 wt%) is observed
above 500 ◦ C, corresponding to the silanol groups condensation and the
dehydroxylation of the sample. Functionalized mesoporous silica SBA-
15_N-ANI is thermally stable up to 200 ◦ C. The mass loss, observed
above this temperature, corresponds to the thermal decomposition of
aromatic amine N-(propyl)aniline (N-ANI) grafted on the surface of SBA-
15. The total mass loss in the temperature range 200–750 ◦ C is 21.7 wt
%, corresponding to an amount of grafted N-ANI groups equal to 1.03
mmol per 1 g of SBA-15. The TGA of curve the sample SBA-15_N-ANI_5-
Fig. 2. XRD patterns of SBA-15, SBA-15_N-ANI and SBA-15_N-ANI_5- FUN containing mixture of loaded drugs (5-FU and NAP) is also pre­
FUN materials. sented in Fig. S4 in the Supplementary Information (brown curve). The
amount of the loaded drugs was calculated from the mass differences
3.3. TEM microscopy observed for SBA-15_N-ANI and SBA-15_N-ANI_5-FUN samples and
represents 9.6 wt%, that corresponds to a total mass of 96 mg of 5-FU
The images from transmission electron microscopy (TEM) are shown and NAP per 1 g of SBA-15_N-ANI solid. Finally, the mass content of
in Fig. S3 in the Supplementary Information. The images show a well- β-CD in the system was also calculated from the SBA-15_N-ANI_5-FUN_
ordered hexagonal mesoporous distribution of uniform channels β-CD TG curve, and a mass of 74.7 mg of β-CD per 1 g of silica
typical for SBA-15 material with 2D hexagonal symmetry. Fig. S3a (0.0658 mmol.g− 1) was found. From the total mass loss observed for the
shows a view obtained with electron beams parallel to the silica chan­ sample SBA-15_N-ANI_5-FUN_ β-CD (Fig. S4, pink curve) the determined
nels, while Fig. S3b shows a view perpendicular to the pore direction. ratio of N-ANI ligands and capping ligand β-CD on the surface of silica
The average pore size determined from the TEM images was 6 nm and was 1:3.
the thickness of the wall of the channel was about 2 nm.
3.5. Adsorption and co-adsorption properties of SBA-15 and SBA-15_N-
ANI
3.4. Thermogravimetric analysis
Adsorption isotherms in liquid phase were measured to understand
Measurements of thermogravimetric analysis (TGA) were performed the adsorption mechanism of the drugs and to quantify the distribution

5
E. Beňová et al. Applied Surface Science 561 (2021) 150011

Fig. 3. (a) Nitrogen adsorption/desorption isotherms of prepared materials and (b) pore size distributions of prepared samples.

the system is far away from saturation and the Henry constant traduces
Table 1
the affinity between the surface and the molecule. Value of Henry
Textural properties of the prepared samples (BET surface area, pore volume and
constant corresponds to:
pore diameter) determined from N2 adsorption/desorption isotherms.
Sample SBET [m2. Mesopore volume Mesopore size dNads
KH = lim (3)
g− 1] [cm3.g− 1] [nm] Ceq →0 dCeq

SBA-15 970 0.96 6.5 Each isotherm was fitted with a Henry-type isotherm (correlation
SBA-15_N-ANI 350 0.53 5.3
SBA-15_N-ANI_5- 260 0.46 5.6
coefficients R2 were >to 0.98) and determined Henry constants are re­
FUN ported in Table 2. As can be seen, whatever the drug and the solid, the
SBA-15_N-ANI_5- 150 0.20 5.0 values of Henry constant are weak suggesting a weak interaction be­
FUN_β-CD tween drugs and solids. Furthermore, the slope observed in the case of
NAP is half compared to the one of 5-FU, evidencing a lower affinity of
of the adsorbate between the liquid phase and solid adsorbent at
equilibrium.
Table 2
The drugs 5-FU and NAP were adsorbed in SBA-15 and SBA-15_N-
Adsorption properties of SBA-15 and SBA-15_N-ANI for 5-fluorouracil (5-FU)
ANI. First, adsorption isotherms of single drugs were measured. The and naproxen (NAP), when each drug was adsorbed alone.
adsorption isotherms of 5-FU and NAP on both, SBA-15 and SBA-15_N-
Sample KH nADS [μmol. nADS [μmol. mADS [mg.
ANI porous solids, determined at 25 ◦ C, are presented in Fig. 4. For both (x1000) g− 1] m− 2] g− 1]
drugs, observed adsorption isotherms are quasi-identical. They are
SBA-15 + 5FU 6.35 ± 322 0.33 41.9
nearly linear corresponding to behavior of Henry isotherms for which a
0.22
linear relation between the adsorbed amount (Nads) and the equilibrium SBA-15_N-ANI + 6.14 ± 325 0.93 42.2
concentration (Ceq) is characteristic: 5FU 0.33
SBA-15 + NAP 3.20 ± 125 0.13 31.6
Nads = KH .Ceq (2) 0.16
SBA-15_N-ANI + 3.25 ± 145 0.41 36.2
where KH is the Henry constant. NAP 0.11
This behavior is frequently observed at low surface coverage when

Fig. 4. Adsorption isotherms of 5-FU (a) and NAP (b) on SBA-15 and SBA-15_N-ANI solids.

6
E. Beňová et al. Applied Surface Science 561 (2021) 150011

NAP towards both solids. As it can be seen from Table 2, the adsorbed Table 3
maximum of 325 μmol⋅g− 1 (42 mg⋅g− 1) was found for 5-FU and nearly Co-adsorption properties of drugs 5-fluorouracil (5-FU) and naproxen (NAP) on
half for NAP, 145 μmol⋅g− 1 (36 mg⋅g− 1). SBA-15 and SBA-15_N-ANI solids.
When 5-FU and NAP drugs are co-adsorbed, the adsorption isotherms Sample KH (x1000) nADS [μmol. nADS [μmol. mADS [mg.
are also found to be linear (see Fig. 5) like in the case of drugs alone. g− 1] m− 2] g− 1]
Nevertheless, during the co-adsorption a difference is observed in the SBA-15 + 5-FU in 3.43 ± 90 0.09 11.7
case of 5-FU as isotherms determined on SBA-15 and SBA-15_N-ANI mix 0.16
solids for 5-FU do not superimpose, and the latter is shifted to higher SBA-15_N-ANI + 5- 5.82 ± 150 0.43 19.6
FU in mix 0.27
adsorbed amounts (see Fig. 5). For NAP, both isotherms superimpose.
SBA-15 + NAP in 10.79 ± 230 0.23 57.1
Therefore, SBA-15_N-ANI silica has a higher loading capacity of drugs mixture 0.36
than pure SBA-15 silica, although SBA-15 has a higher specific surface SBA-15_N-ANI + 11.42 ± 250 0.71 62.5
area (970 m2⋅g− 1) compared to SBA-15_N-ANI (350 m2⋅g− 1). The NAP in mix 0.27
determined Henry constants and adsorbed amounts during the co-
adsorption are reported in Table 3.
Supplementary information). The maximal amounts of both drugs
Furthermore, in the case of co-adsorption, the slope of the NAP
adsorbed alone and from the mixture in different molar ratios are re­
isotherm is double compared to the one of 5-FU (see Fig. 5 and the values
ported in Table ST1 in the Supplementary information.
of Henry constants in Table 3), evidencing a higher affinity of NAP to­
To complete the study of adsorption properties, the microcalori­
wards both solids. The maximum adsorbed amounts for both drugs and
metric measurements were performed for the adsorption of NAP on SBA-
both solids are reported in different units in Table 3. A maximal
15_N-ANI, as a single drug and as a mixture with 5-FU. The corre­
adsorbed amount 150 μmol⋅g− 1 (20 mg⋅g− 1) was found for 5-FU and
sponding integral displacement enthalpies are presented in Fig. 6. The
higher, 250 μmol⋅g− 1 (62 mg⋅g− 1), was found for NAP. Totally 82
overlay of the adsorption enthalpies versus the NAP adsorbed amount
mg⋅g− 1 of drugs was adsorbed and this maximum agrees with the total
adsorbed mass close to 96 mg, determined for 5-FU+NAP using TGA
experiments. All these results suggest a difference in the adsorption
properties of NAP when the drug is adsorbed alone or mixed with 5-FU.
To clearly evidence this difference, isotherms determined for SBA-
15_N-ANI for each drug adsorbed alone and in a mutual mixture are
presented in Fig. S5 in the Supplementary Information. As it is clearly
shown, there are no differences in adsorption properties for 5-FU drug
alone or in mixture; both isotherms, for single 5-FU and in a mixture are
very similar. For NAP, the amount of drug adsorbed when it is mixed
with 5-FU is almost three times larger than when it is adsorbed alone
(see Fig. S5 in the Supplementary Information). The samples SBA-15_N-
ANI has a negative surface charge of − 82 mV at pH 7.4, as it was re­
ported in our previous work [31]. 5-FU has NH groups which are not
deprotonated at pH 7.4 (pKa (5-FU) = 8.02). On the contrary, NAP (the
pKa (NAP) = 4.15) has a polar carboxylate group so we assume that the
adsorption of both drugs is governed by electrostatic interactions and
hydrogen bonding with negatively charged particles.
The maximal amounts of the drugs adsorbed on the SBA-15_N-ANI_5-
FUN_ β-CD sample, which was prepared for biological tests, were 224
μmol⋅g− 1 (29 mg⋅g− 1) and 361 μmol⋅g− 1 (91 mg⋅g− 1) for 5-FU and NAP,
Fig. 6. Evolution of integral displacement enthalpies of NAP adsorbed on SBA-
respectively. These maxima agree with the total mass of 126 mg,
15_N-ANI, as a single drug (green dots) and as a mixture with 5-FU (yel­
determined for 5-FU+NAP using TGA experiments (see Figure S6 in
low dots).

Fig. 5. Adsorption isotherms of co-loaded drugs 5-FU (a) and NAP (b) on SBA-15 and SBA-15_N-ANI.

7
E. Beňová et al. Applied Surface Science 561 (2021) 150011

exhibits the same behavior: at very low coverage, the enthalpies are 5-FU and 11.7% of NAP were released. In total, 88.5% of 5-FU and
close to − 15 kJ⋅mol− 1, then decrease monotonically with the NAP 98.7% of NAP was released in 24 h.
adsorbed amount until reaching an asymptotical value about − 5 It is obvious that the release performance of the investigated drug
kJ⋅mol− 1. These values of enthalpies can be related to the weak affinity delivery system is dependent on the pH. Substantial differences in
observed between NAP and the surface of SBA-15_N-ANI and are usually release profiles at pH = 7.4, when the pores of the carrier are blocked by
met in the case of weak electrostatic interactions. Considering the evo­ β-CD, were observed for both drugs, in comparison with the profiles at
lution of enthalpies versus the coverage, two interpretations can be pH = 5. Indeed, there is no significant release at pH 7.4 and only a
done. First, there is a succession of sites saturated step by step from the release of 16% of 5-FU and 20% of NAP in the initial 3 h was observed.
highest energy to the lowest. Moreover, the identical shape of enthalpies This release could be explained by the fact that not all pores were capped
evolution for a single drug NAP (green dots) and in a mixture with 5-FU properly by β-CD molecules. This hypothesis is also in agreement with
(yellow dots) means that the NAP adsorption sites are not dependent on N2 adsorption/desorption measurements, where some mesoporosity was
the 5-FU presence, in this range of coverage. Furthermore, NAP and 5- still observed after β-CD capping. However, no release of drugs was
FU molecules are not in competition regarding their sites of adsorption. observed after these 3 h, which leads to the conclusion that supramo­
lecular interaction of the couple N-ANI/ β-CD is efficient, and this couple
may serve as a tight gatekeeper.
3.6. Release performance of 5-Fluorouracil and Naproxen
To investigate the effect of co-existence of two drugs on the release
process, the release profiles of single drugs were also measured. Fig. 8
Drugs release studies were performed in two different media, in
shows the release profiles of 5-FU (see Fig. 8a) and NAP (see Fig. 8b), at
simulated body fluid (pH = 7.4) and slightly acidic medium (pH = 5)
pH 5, when they were released as single drugs (black curves) or in their
simulating the pH of tumor affected area. The released amounts of 5-FU
mixture (5-FU – red curve, NAP – orange curve).
and NAP from SBA-15_N-ANI_5-FUN_ β-CD sample at different time in­
The release profile of 5-FU as a single drug (Fig. 8a, black curve)
tervals (0.08 h, 0.33 h, 0.66 h, 1 h, 2 h, 3 h, 4 h, 5 h, 6 h and 24 h) were
shows gradual drug release of 45% in 6 h and 84% of 5-FU in total after
investigated using UV spectroscopy. The 5-FU and NAP concentrations
24 h. When 5-FU was mixed with naproxen, (Fig. 8a, red curve), 68% of
during the release studies were determined using the absorbance at 266
5-FU was released after 6 h, while 89% of 5-FU released in total after 24
nm for 5-FU and 330 nm for NAP. The absorbance of 5-FU at 266 nm and
h. It can be seen, that during the same time interval, especially in the
NAP at 330 nm does not overlay on each other. The calibration curves
first 6 h, a larger amount of 5-FU was released from the system con­
for 5-FU and NAP were obtained from the UV spectra of drug standards
taining a mixture of drugs (even though the starting concentration was
and are shown in Fig. S1 and S2 in the Supplementary Information. The
two times lower 5.10− 2 M and 2.5.10− 2 M for the release of single 5FU in
total amount of drugs loaded in SBA-15_N-ANI_5-FUN_ β-CD sample was
the mixture, respectively).
established from the adsorption isotherms. The loaded amount was 20
The release profile of NAP as a single drug (Fig. 8b, black curve) at
mg of 5-FU per 1 g of silica (0.15 mmol⋅g− 1) and 62 mg of NAP (0.248
pH = 5 shows an initial burst of almost 40% of NAP during the first hour
mmol⋅g− 1) per 1 g of solid. This result was also confirmed by TGA
and almost no drug release afterward. The release profile of NAP from
measurements. These amounts were taken as 100% in the release studies
the system containing a mixture of drugs (Fig. 8b, orange curve) shows a
and used to calculate the percentage of the released amount in the
gradual release of 98.7% of NAP within 24 h with a faster release rate.
respective time intervals. From the UV spectroscopy results the curves of
To sum up, the experiments showed that, while at physiological pH
the time dependencies of the released amounts of 5-FU and NAP from
7.4, the prepared drug delivery system is tight enough and only releases
SBA-15_N-ANI_5-FUN_ β-CD were constructed. The release curves were
16% of 5-FU and 20% of NAP. When pH decreases to 5, the pores of the
determined at pH = 5, i.e. in the open pore configuration and, at pH =
system open and released amounts of drugs are close the maximum
7.4, i.e. in the configuration when pores of the particles are capped by
adsorbed ones, i.e. 89% of 5-FU and 98.7% of NAP.
β-CD. The results are presented in Fig. 7.
As it can be seen from Fig. 7, at pH = 5, a significant part of the drugs
was released in the first 6 h, when the release of 68% of 5-FU (Fig. 7a) 3.7. Release kinetics
and 87% of NAP (Fig. 7b) was observed. In the next 18 h, the sustained
release of drugs continued at a much slower rate and additional 20.5% of Drug release is a phenomenon controlled by diffusion or dissolution

Fig. 7. Release profiles of a) 5-FU and b) NAP from solid SBA-15_N-ANI_5-FU_ β-CD at different two different pH (5 and 7.4).

8
E. Beňová et al. Applied Surface Science 561 (2021) 150011

Fig. 8. Release profiles of 5-FU (a) and NAP (b) at pH 5 performed in their free form and the mixture. The profiles obtained during the release of single drugs are
shown as black curves and profiles obtained during the release in their mixture are shown as red curve (5-FU) and orange curve (NAP).

processes or both factors and in the previous studies, the drug release diffusion. Thus, the Korsmeyer-Peppas fitting confirms the diffusion-
from porous silica matrices was described as a diffusion-controlled controlled mechanism identified by the Higuchi analysis.
process [27,42–45]. To study the mechanism of the 5-FU and NAP
release from SBA-15_N-ANI_5-FUN_ β-CD, drugs release profiles were
3.8. Cytotoxicity of SBA-15_N-ANI loaded with 5-Fluorouracil and
first fitted with a Higuchi model. The Higuchi square root of time model
Naproxen
has commonly been used for modeling diffusion-controlled processes of
drug release. The linear regression analysis of the data was performed
Two different cancer cell lines were used to test the silica-based
and determined correlation coefficients (r2) and rate constants (k) are
delivery system: U87 MG and SKBR3 cells. U87 MG cells being larger
listed in Table ST2 in the Supplementary Information. As correlation
than SKBR3 ones are more suitable for imaging. SKBR3 cells were used
coefficients obtained from the regression analysis for data over the full
in a flow-cytometric study. A drug resistance can be developed in both
release time duration were far from one (see Table ST2), it was under­
selected lines [51,52].
stood that the release of the drugs should involve several steps. The
As it was reported recently, SBA-15_N-ANI_5-FU_β-CD drug delivery
release of drug is often a two-stage process: a fast release in the initial
system induced cytotoxic stress in mitochondria and lysosomes of U87
stage followed by a second stage with a slower release rate. Various
MG cells [31]. For this reason, the cytotoxicity of SBA-15_N-ANI_5-
types of carriers like microspheres, mesoporous materials, and nano­
FUN_β-CD was first assessed in U87 MG cells. The representative bright-
particles exhibit different stages of drug release [46,47]. For example,
field images of U87 MG cells incubated with 5-FU and SBA-15_N-ANI_5-
the release of vancomycin from silica microspheres shows three stages of
FUN_β-CD for 24 h are demonstrated in Fig. 9. The morphology of the
drug release, a first delay phase, a second release phase, followed by a
cells in the presence and absence of 5-FU was not changed at the studied
third stage with a slower release rate [48]. For this reason, the release
concentrations. SBA-15_N-ANI_5-FUN_β-CD (with 5-FU:NAP at molar
profiles of 5-FU and NAP were divided into several stages selected on the
ratio 1:1) particles were observed in the same area as the cells (dark
criterion of the proximity of the fitting correlation coefficients to one. In
particles in Fig. 9a). However, some fragments of the cells surrounded by
this way, the experimental data were divided into three regions for 5-FU
SBA-15_N-ANI_5-FUN_β-CD were detected. Administration of 2 mg/mL
(0–0.3 h; 0.3–5 h; 5–24 h) and into two regions for NAP (0–5 h; 5–24 h)
of SBA-15_N-ANI_5-FUN_β-CD resulted in relatively large cluster
and independently regressed with the Higuchi model. The modeling of
formation.
the first and second stages of both 5-FU and NAP release from SBA-15_N-
The viability of U87 MG cells was evaluated by MTT assay at 24 and
ANI_5-FUN_ β-CD showed a perfect fit with the Higuchi model, as cor­
48 h after the administration of 5-FU and NAP loaded SBA-15_N-ANI_5-
relation coefficients lie in the 0.94–0.96 range. It means drugs were
FUN_β-CD. The possible synergism of free 5-FU and NAP was tested as
released by a diffusion-controlled manner. The regression analyses also
well. First, the solution with the concentration ratios of free drugs 5-FU:
revealed that the “best fit” is for 5-FU, lower for the third stage than the
NAP 5:1, 1:1, 1:2, 1:4 and 1:20 were applied. The application of 5-FU by
two first stage. This observation agrees with the literature reporting that
5-FU loaded SBA-15_N-ANI_5-FUN_β-CD significantly increased the
significant deviations are possible above 50% release and that the “best
formazan production in U87 MG cells detected 24 h after administration.
fit” of diffusion-controlled processes is generally better for the modeling
We have not detected a significant difference in cells treated with 5-FU
of the initial stage of release than for the modeling of the terminal stage
alone, its combinations with NAP, and with NAP alone. The significant
[49,50].
difference of formazan production was observed in cells incubated with
To verify the conclusion drawn from the Higuchi analysis dealing
SBA-15_N-ANI_5-FUN_β-CD at 5-FU:NAP molar ratios 1:1 and 1:2. These
with drugs transport mechanisms, the Korsmeyer-Peppas model was also
significances were found between free 5-FU+NAP and SBA-15_N-ANI_5-
tested with the experimental data. The obtained parameters are listed in
FUN_β-CD, and between both ratios as well. However, the level of for­
Table ST3 in the Supplementary Information. For the first and second
mazan production in control and cells incubated in the presence of SBA-
stages of 5-FU release, the correlation coefficients were equal to 1 and
15_N-ANI_5-FUN_β-CD was not significantly different. It looks like SBA-
0.9363, respectively and for the first stage of NAP release, r2 = 0.9537.
15_N-ANI_5-FUN_β-CD formulation protects the release of 5-FU and NAP
Obtained values evidence, for both drugs, a good fitting between the
from the SBA-15_N-ANI_β-CD towards the cells. Later, 48 h of treatment
model and experimental data. Furthermore, the diffusion exponents (n)
with free 5-FU+NAP (5:1 and 1:4), and SBA-15_N-ANI_5-FUN_β-CD
values that are inferior to ≤0.45 suggest the domination of Fickian
(1:2), the viability (related to formazan production) of U87 MG cells was

9
E. Beňová et al. Applied Surface Science 561 (2021) 150011

Fig. 9. Representative bright-field microscopic images of U87 MG cells in the absence and presence of 5-FU and SBA-15_N-ANI_5-FUN_β-CD. The images were
collected 24 h post-administration at pH = 7.3 (a) and 6.3 (b). The MTT assays were performed in cells 24 and 48 h incubated with 5-FU, NAP, the combination of 5-
FU + NAP and SBA-15_N-ANI_5-FUN_β-CD at different concentration ratios at pH = 7.3 (a) and 6.3 (b). The histograms represent the average values of at least three
independent experiments and the error bars represent standard deviations from these values. The level of significance was evaluated by the Student t-test with the *p
< 0.05, **p < 0.01 and ***p < 0.001.

significantly inhibited. It can be seen from these results that the syner­ The apoptosis induced in SKBR3 cells by SBA-15_N-ANI_5-FUN_β-CD
gistic effect of 5-FU and NAP is concentration depended. It should be (1:1 and 1:2) was studied with flow-cytometry, 48 h after the adminis­
noted that the concentration of the 5-FU was constant at 5:1, 1:1 and 1:2, tration. The representative bright-field images of SKBR3 cells in the
and the NAP concentration was constant at 1:2, 1:4 and 1:20. Impor­ absence and presence of 5-FU, NAP, and SBA-15_N-ANI_5-FUN_β-CD
tantly, cell viability at SBA-15_N-ANI_5-FUN_β-CD (1:2) was signifi­ (1:1 and 1:2) are displayed in Fig. 10a. In these images, the presence of
cantly different from the control, free 5-FU+NAP (1:2) and SBA-15_N- SBA-15_N-ANI_5-FUN_β-CD particles are clearly evidenced (black ar­
ANI_5-FUN_β-CD (1:1). These results suggest that SBA-15_N-ANI_5- rows in the zoomed images). These particles were localized in the
FUN_β-CD (1:2) delivers more effectively 5-FU into the cells with the extracellular space, close to the plasma membrane of the cells. We as­
synergistic effect induced by NAP. sume that the adsorption of SBA-15_N-ANI_5-FUN_β-CD (1:1 and 1:2)
The synergistic effect of 5-FU co-administered with another anti- into the plasma membrane occurred.
cancer active molecule was demonstrated in hepatocellular carcinoma The particles adsorbed on the cell surface (plasma membrane) may
and colon cancer cells [53,54]. The application of both molecules change the optical parameters of detected system. The scattered light
resulted in apoptosis, an increase of reactive oxygen species production, was measured with flow cytometry. The FSC and SSC parameters
and dissipation of mitochondrial membrane potential. Moreover, mes­ characterize the size and the density of cells. From Fig. 10b, it can be
oporous silica nanoparticles were previously demonstrated to be able to seen that the size of the cells (FSC value) is similar in all studied con­
deliver drugs with high loading capacity and increase the cytotoxicity of ditions. The granularity and the cell density increased after the admin­
the system based on paclitaxel, tetrandrine, 5-FU, coumarin, p53 gene, istration of SBA-15_N-ANI_5-FUN_β-CD (1:1 and 1:2). The shift in the cell
quercetin, and doxorubicin [9,55–57]. In our study, we aimed to in­ density was demonstrated with the black arrow in Fig. 10b. In the next
crease the cytotoxicity of 5-FU at low acidic pH. step, we have studied if the SBA-15_N-ANI_5-FUN_β-CD (1:1 and 1:2) can
The effect of the cell culture medium pH on the cytotoxicity of SBA- trigger the apoptosis, and elevate oxidative stress level in the mito­
15_N-ANI_5-FUN_β-CD was tested in U87 MG cells. The conditions were chondria. The AnnexinV/FITC labelled phosphatidylserine expressed in
maintained at pH = 6.3. The representative bright-field images of U87 SKBR3 cells was detected and cross-correlated with MTO, a sensor of
MG cells in the presence and absence of SBA-15_N-ANI_5-FUN_β-CD reactive oxygen species (Fig. 10c). The percentage of cells related to
(1:1) are demonstrated in Fig. 9b. living cells (* in the image), apoptosis and cells with the dissipated
There was no difference in the morphological structures of the cells. mitochondrial membrane potential (ΔΨm) were denoted in histograms.
Similarly, as demonstrated in Fig. 9a by MTT assay, the application of 5- We can recognize the reduction of living cells, the increase of
FU significantly increased the formazan production at both pHs: the apoptotic populations and the decrease of MTO fluorescence in cells
administration of SBA-15_N-ANI_5-FUN_β-CD (1:1) at pH = 7.3 resulted subjected to SBA-15_N-ANI_5-FUN_β-CD (1:1 and 1:2). The decrease of
in the same viability as the control U87 MG cells, whereas the viability of MTO fluorescence suggests the generation of oxidative stress in cells.
U87 MG cells at pH = 6.3 was significantly reduced at 24 h and 48 h after This oxidative stress could be the source that triggered the apoptosis
the administration of SBA-15_N-ANI_5-FUN_β-CD (1:1) (Fig. 9b). We observed. The same level of apoptotic cells was observed at both molar
assume that at the longer (48 h) incubation time cells consumed nutri­ ratios of SBA-15_N-ANI_5-FUN_β-CD (1:1 and 1:2). From these results,
tion and partially decreases pH that enabled the release of 5-FU and NAP we can assume that the SBA-15_N-ANI_5-FUN_β-CD is more effective
from SBA-15_N-ANI_5-FUN_β-CD (Fig. 9a). The released 5-FU and NAP than free 5-FU and NAP.
increased cytotoxicity of the system in U87 GM cells. This effect was
confirmed when the cells were incubated at lower pH (Fig. 9b).

10
E. Beňová et al. Applied Surface Science 561 (2021) 150011

Fig. 10. Representative bright-field microscopic images a) of SKBR3 cells detected (48 h) in the presence and absence of 5-FU, NAP, and SBA-15_N-ANI_5-FUN_β-CD
(1:1 and 1:2). b) Representative FSC vs. SSC correlation plots of the same samples detected by flow cytometry. c) Apoptosis (AnnexinV/FITC) and mitochondrial
membrane potential (MTO) detected in SKBR3 cells at the same condition as in a) and b) by flow-cytometry. The number of events detected per population was
plotted in a histogram. The number of events/cells in the plots were colour-coded (blue-minimum, red-maximum). The black arrows point to SBA-15_N-ANI_5-FUN_β-
CD particles.

3.9. In vivo biocompatibility – CAM assay after the administration (Fig. 11). Later, at 24 h after the administration,
the applied SBA-15_N-ANI_5-FUN_β-CD created bulky whitish clusters
The CAM assay, in this study, was used to evaluate the biocompati­ that did not penetrate through the ectoderm. The histological sections
bility of SBA-15_N-ANI_5-FUN_β-CD with the live system, the quail CAM were taken at the end of the observation (24 h after the administration of
model of the microcirculatory system. This model is well established, the solutions).
cost-less, reproducible, and represents a simple model of a living cell The H&E staining of histological sections of CAM treated with 5-FU
membrane system with blood vasculature suitable for drug and delivery and low concentration of SBA-15_N-ANI_5-FUN_β-CD displays similar
system studies. It is widely used as a preclinical model of biocompati­ patterns as the control (last row in Fig. 11). The histological section of
bility and toxicity studies of drug/molecule/transport systems [58–60]. the highest applied concentration of SBA-15_N-ANI_5-FUN_β-CD showed
The importance of drug release in this study is in the release of drug from that it induced damage to the CAM ectoderm.
the transport system toward the CAM surface cells. The ectoderm (Ec) of The massive haemorrhages were observed in this sample and were
CAM represents the first barrier for drug penetration into the embryo. In denoted by the black arrow in the image. At our studied conditions, the
case of high toxicity, the membrane would be damaged, and embryo release of the drugs (5-FU and NAP) from the nanoparticles towards
would not survive. Therefore, the main objective of this study was to test CAM is very probable. However, the applied concentrations of drugs
the biocompatibility of the prepared system with health – non-cancerous were below therapeutically applied doses related to CAM embryo.
tissue. In our study, the viability of the treated CAM on the second day was
The representative photographic images of CAM treated with 5-FU, 100%. This suggests that the SBA-15_N-ANI_5-FUN_β-CD are biocom­
NAP and SBA-15_N-ANI_5-FUN_β-CD at different concentrations are patible with CAMs. The local damage of the CAM at the high applied
demonstrated in Fig. 11. The solutions were topically applied to the area concentrations could be explained as the result of the drug cytotoxicity
of silicone ring at 30 µL of the volumes. It can be recognized that the in the area related to the whitish SBA-15_N-ANI_5-FUN_β-CD aggregates.
application of 5-FU and NAP did not induce significant damage. A Similarly, but elsewhere, the rotigotine release from the polymeric
similar effect was observed at low concentrations (0.3 and 0.4 mg/mL) nanoparticles induced the edema in the injection site [61]. To prevent
of SBA-15_N-ANI_5-FUN_β-CD applied on the CAM. However, the high the Ec damage, the SBA-15_N-ANI_5-FUN_β-CD could be attached to
concentrations (3 and 4 mg/mL) of SBA-15_N-ANI_5-FUN_β-CD created some biocompatible scaffold to enable homogeneous distribution.
homogenous filter-like liquid surface (white colour), detected at 1 min Another explanation of the CAM damage could be that the created filter-

11
E. Beňová et al. Applied Surface Science 561 (2021) 150011

Fig. 11. Representative photographic images of quail CAM detected before and after administration of 5-FU, NAP and SBA-15_N-ANI_5FUN_β-CD at the concen­
tration ratio 1:1 and 1:2. The 30 µL of the solutions were applied to the silicone ring to homogeneously cover the CAM surface. The H&E staining was performed 24 h
after the administration (Ec – ectoderm, M – mesoderm, En - endoderm). The black arrow points to Ec damage at higher concentration of SBA-15_N-ANI_5FUN_β-CD.

like surface prevented the oxygenation of the tissue, resulting in hyp­ adsorption/desorption measurements, where some mesoporosity was
oxia, and massive damage. A very similar effect was observed in the still observed after β-CD capping. However, no release of drugs was
study of nanoceria with CAM-assay [60]. observed after these 3 h. At pH = 5, the β-CD molecules were loosened,
pores were opened and 89% of 5-FU and 99% of NAP drug amounts
4. Conclusions released.
Biocompatibility and cytotoxicity studies were performed using U87
An advanced nanoporous system for pH driven drug delivery, based MG and SKBR3 cells. The MTT assay showed that, at physiological pH
on N-(propyl)aniline modified mesoporous silica and β-cyclodextrin (7.4), SBA-15_N-ANI_5-FUN_β-CD formulation protects the release of 5-
(β-CD) was prepared. The idea of the work was driven by the fact, that in FU and NAP from the SBA-15_N-ANI_β-CD towards the cells. A synergic
the tumor tissues, the pH is slightly lower compared to the healthy cells effect induced by NAP was observed, and it was shown that 5-FU can be
represent by the standard physiological environment. For this reason, delivered using SBA-15_N-ANI_5-FUN_β-CD more effectively than 5-FU
the system was designed to have pores closed at physiological pH = 7.4 itself. The effect of pH on the cytotoxicity of SBA-15_N-ANI_5-FUN_β-
and opened at a lower one, pH = 5. The system was tested for the co- CD was tested in U87 MG cells at pH = 7.3 and 6.3. The administration
delivery of two different drugs: an anticancer agent (5-fluorouracil, 5- of SBA-15_N-ANI_5-FUN_β-CD at pH = 7.3 resulted in the same viability
FU) and anti-inflammatory drug (naproxen, NAP). as the control U87 MG cells. The viability of U87 MG cells at pH = 6.3
Before release experiments, a detailed study of drugs adsorption was significantly reduced at the 24 and 48 h after the administration of
(adsorption of 5-FU and NAP each alone), and the co-adsorption of both SBA-15_N-ANI_5-FUN_β-CD. Tests of in vivo biocompatibility in CAM
drugs from mutual mixture was performed. It was evidenced that assay showed the CAM ectoderm damage only at the highest applied
adsorbed amounts of NAP in the mixture were almost three times larger concentration of SBA-15_N-ANI_5-FUN_β-CD.
compared to the adsorption of a single NAP. In the case of 5-FU, no
significant changes were observed. CRediT authorship contribution statement
It was shown that, at physiological pH = 7.4, the prepared drug
delivery system was tight and only 16% of 5-FU and 20% of NAP were Eva Benova: Investigation, Data curation, Formal analysis, Visuali­
released in first 3 h. This release could be explained by the fact that not zation, Writing - original draft. Virginie Hornebecq: Conceptualiza­
all pores were capped properly by β-CD molecules, which agrees with N2 tion, Methodology, Formal analysis, Writing - original draft, Writing -

12
E. Beňová et al. Applied Surface Science 561 (2021) 150011

review & editing, Supervision, Project administration. Vladimír Zele­ delivery of macromolecules, J. Control. Release 326 (2020) 544–555, https://doi.
org/10.1021/acs.bioconjchem.8b00416.
nak: Conceptualization, Methodology, Formal analysis, Visualization,
[12] R.M. Pallares, P. Agbo, X. Liu, D.D. An, S.S. Gauny, S.E. Zeltmann, A.M. Minor, R.
Writing - original draft, Writing - review & editing, Supervision, Project J. Abergel, Engineering mesoporous silica nanoparticles for targeted alpha therapy
administration, Funding acquisition. Veronika Huntosova: Investiga­ against breast cancer, ACS Appl. Mater. Interfaces 12 (2020) 40078–40084,
tion, Data curation, Formal analysis, Visualization, Writing - original https://doi.org/10.1021/acsami.0c11051.
[13] R.R. Castillo, D. Lozano, B. González, M. Manzano, I. Izquierdo-Barba, M. Vallet-
draft. Miroslav Almasi: Formal analysis, Visualization, Writing - orig­ Regí, Advances in mesoporous silica nanoparticles for targeted stimuli-responsive
inal draft. Mariana Macajova: Investigation, Data curation. David drug delivery: an update, Expert Opinion on Drug Delivery, Expert Opinion on
Berge-Lefranc: Investigation, Data curation. Drug Deliv. 16 (2019) 415–439, https://doi.org/10.1080/
17425247.2019.1598375.
[14] L. Huang, M. Liu, L. Mao, D. Xu, Q. Wan, G. Zeng, Y. Shi, Y. Wen, X. Zhang, Y. Wei,
Declaration of Competing Interest Preparation and controlled drug delivery applications of mesoporous silica
polymer nanocomposites through the visible light induced surface-initiated ATRP,
Appl. Surf. Sci. 412 (2017) 571–577, https://doi.org/10.1016/j.
The authors declare that they have no known competing financial apsusc.2017.04.026.
interests or personal relationships that could have appeared to influence [15] Y. Wang, Y. Cui, J. Huang, D. Di, Y. Dong, X. Zhang, Q. Zhao, N. Han, Y. Gao,
T. Jiang, S. Wang, Redox and pH dual-responsive mesoporous silica nanoparticles
the work reported in this paper.
for site-specific drug delivery, Appl. Surf. Sci. 356 (2015) 1282–1288, https://doi.
org/10.1016/j.apsusc.2015.07.151.
Acknowledgements [16] T. Zhao, L. Chen, Q. Li, X. Li, Near-infrared light triggered drug release from
mesoporous silica nanoparticles, J. Mater. Chem. B 6 (2018) 7112–7121. https://
doi.org/10.1016/j.jconrel.2020.07.021.
This work was supported by the Slovak Research and Development [17] F. Liu, P. Huang, D. Huang, S. Liu, Q. Cao, X. Dong, H. Zhang, F. Ko, W. Zhou,
Agency under the contract APVV-15-0520. E.B thanks French Govern­ Smart “on-off” responsive drug delivery nanosystems for potential imaging
diagnosis and targeted tumor therapy, Chem. Eng. Journal 365 (2019) 358–368,
ment for the financial support linked to the implementation of French-
https://doi.org/10.1016/j.cej.2019.02.037.
Slovak doctorate. V.Z. thanks the Ministry of Education, Science, [18] T. Numpilai, T. Witoon, M. Chareonpanich, J. Limtrakul, Impact of
Research and Sport of the Slovak Republic and the Accreditation Com­ physicochemical properties of porous silica materials conjugated with
mission for the financial support of the TRIANGEL team in the frame of dexamethasone via pH-responsive hydrazone bond on drug loading and release
behavior, Appl. Surf. Sci. 396 (2017) 504–514, https://doi.org/10.1016/j.
the scheme “Top Research Teams in Slovakia”. This publication is also apsusc.2016.10.183.
the result of the project implementation: “Open scientific community for [19] Y. Gao, Y. Xiao, K. Mao, X. Qin, Y. Zhang, D. Li, Y. Zhang, J. Li, H. Wan, S. He,
modern interdisciplinary research in medicine (OPENMED)”, Thermoresponsive polymer-encapsulated hollow mesoporous silica nanoparticles
and their application in insecticide delivery, Chem. Eng. J. 383 (2020) 123169,
ITMS2014+: 313011V455 supported by the Operational Programme https://doi.org/10.1016/j.cej.2019.123169.
Integrated Infrastructure, funded by the ERDF. [20] P. Joyce, H. Ulmefors, S. Maghrebi, S. Subramaniam, A. Wignall, S. Jõemetsa,
F. Höök, C.A. Prestidge, Enhancing the cellular uptake and antibacterial activity of
rifampicin through encapsulation in mesoporous silica nanoparticles,
Appendix A. Supplementary data Nanomaterials 10 (2020) 815, https://doi.org/10.3390/nano10040815.
[21] M. Vallet-Regi, A. Rámila, R.P. del Real, J. Pérez-Pariente, A new property of MCM-
Supplementary data to this article can be found online at https://doi. 41: drug delivery system, Chem. Mater. 13 (2001) 308–311, https://doi.org/
10.1021/cm0011559.
org/10.1016/j.apsusc.2021.150011. [22] G. Marinescu, D.C. Culita, C. Romanitan, S. Somacescu, C.D. Ene, V. Marinescu, D.
G. Negreanu, C. Maxim, M. Popa, L. Marutescu, M. Stan, C. Chifiriuc, Novel hybrid
References materials based on heteroleptic Ru(III) complexes immobilized on SBA-15
mesoporous silica as highly potent antimicrobial and cytotoxic agents, Appl. Surf.
Sci. 520 (2020) 146379, https://doi.org/10.1016/j.apsusc.2020.146379.
[1] P. Borst, Looking back at multidrug resistance (MDR) research and ten mistakes to
[23] L. Hajiaghababaei, M. Eslambolipour, A. Badiei, M.R. Ganjali, G.M. Ziarani,
be avoided when writing about ABC transporters in MDR, FEBS Lett. 594 (2020)
Controlled release of anticancer drug using o-phenylenediamine functionalized
4001–40011, https://doi.org/10.1002/1873-3468.13972.
SBA-15 as a novel nanocarrier, Chem. Papers 75 (2021) 1841–1850, https://doi.
[2] A. Wolfensberger, S.P. Kuster, M. Marchesi, R. Zbinden, M. Hombach, The effect of
org/10.1007/s11696-020-01422-9.
varying multidrug-resistence (MDR) definitions on rates of MDR gram-negative
[24] B. Malfait, N.T. Correia, C. Ciotonea, J. Dhainaut, J.P. Dacquin, S. Royer,
rods, Antimicrobial Resistance & Infection Control 8 (2019) 193. https://doi.
N. Tabary, Y. Guinet, A. Hédoux, Manipulating the physical states of confined
org/10.1186/s13756-019-0614-3.
ibuprofen in SBA-15 based drug delivery systems obtained by solid-state loading:
[3] R.R. Castillo, M. Colilla, M. Vallet-Regí, Advances in mesoporous silica-based
Impact of the loading degree, J. Chem. Phys. 153 (2020) 154506, https://doi.org/
nanocarriers for co-delivery and combination therapy against cancer, Expert Opin.
10.1063/5.0020992.
Drug Deliv. 14 (2017) 229–243, https://doi.org/10.1080/
[25] S. Kingchok, S. Pornsuwan, Comparison of spherical and rod-like morphologies of
17425247.2016.1211637.
SBA-15 for enzyme immobilization, J. Porous Mater. 27 (2020) 1547–1557,
[4] J.L. Paris, M. Vallet-Regí, Mesoporous silica nanoparticles for co-delivery of drugs
https://doi.org/10.1007/s10934-020-00932-x.
and nucleic acids in oncology: a review, Pharmaceutics 12 (2020) 526, https://doi.
[26] F. Khanmohammadi, M.A. Molina, R.M. Blanco, S.N. Azizi, C. Marquez-Alvarez,
org/10.3390/pharmaceutics12060526.
I. Diaz, Microp. Mesop. Mater. 309 (2020) 1387–1811, https://doi.org/10.1016/j.
[5] Z.Y. Li, Y. Liu, X.Q. Wang, L.H. Liu, J.J. Hu, G.F. Luo, W.H. Chen, L. Rong, X.
micromeso.2020.110527.
Z. Zhang, One-pot construction of functional mesoporous silica nanoparticles for
[27] V. Zeleňák, D. Halamová, M. Almáši, L. Žid, A. Zeleňáková, O. Kapusta, Ordered
the tumor-acidity-activated synergistic chemotherapy of glioblastoma, ACS Appl.
cubic nanoporous silica support MCM-48 for delivery of poorly soluble drug
Mater. Interfaces 5 (2013) 7995–8001, https://doi.org/10.1021/am402082d.
indomethacin, Appl. Surf. Sci. 443 (2018) 525–534, https://doi.org/10.1016/j.
[6] Q. Liu, J. Zhang, W. Sun, Q. Reuben Xie, W. Xia, H. Gu, Delivering hydrophilic and
apsusc.2018.02.260.
hydrophobic chemotherapeutics simultaneously, Int. J. Nanomed. 7 (2012)
[28] M. Almáši, E. Beňová, V. Zeleňák, B. Madaj, V. Huntošová, J. Brus, M. Urbanová,
999–1013, https://doi.org/10.2147/IJN.S28088.
J. Bednarčík, V. Hornebecq, Cytotoxicity study and influence of SBA-15 surface
[7] X. Li, M. Wu, L. Pan, J. Shi, Tumor vascular-targeted co-delivery of anti-
polarity and pH on adsorption and release properties of anticancer agent
angiogenesis and chemotherapeutic agents by mesoporous silica nanoparticle-
pemetrexed, Mater. Sci. Eng. C 109 (2020) 110552, https://doi.org/10.1016/j.
based drug delivery system for synergetic therapy of tumor, Int. J. Nanomed. 11
msec.2019.110552.
(2015) 93–105, https://doi.org/10.2147/IJN.S81156.
[29] V. Zeleňák, E. Beňová, M. Almáši, D. Halamová, V. Hornebecq, V. Hronský, Photo-
[8] Z. Gounani, M.A. Asadollahi, J.N. Pedersen, J. Lyngsø, J. Skov Pedersen,
switchable nanoporous silica supports for controlled drug delivery, New J. Chem.
A. Arpanaei, R.L. Meyer, Mesoporous silica nanoparticles carrying multiple
42 (2018) 13263–13271, https://doi.org/10.1039/C8NJ00267C.
antibiotics provide enhanced synergistic effect and improved biocompatibility,
[30] E. Beňová, V. Zeleňák, D. Halamová, M. Almáši, V. Petrul’ová, M. Psotka, A.
Colloids Surf. B Biointerfaces 175 (2019) 498–508, https://doi.org/10.1016/j.
Zeleňáková, M. Bačkor, V. Hornebecq, A drug delivery system based on switchable
colsurfb.2018.12.035.
photo-controlled p-coumaric acid derivatives anchored on mesoporous silica, J.
[9] L. Jia, Z. Li, J. Shen, D. Zheng, X. Tian, H. Guo, P. Chang, Multifunctional
Mater. Chem. B 5 (2017) 817–825. https://doi.org/10.1039/C6TB02040B.
mesoporous silica nanoparticles mediated co-delivery of paclitaxel and tetrandrine
[31] E. Beňová, D. Bergé-Lefranc, V. Zeleňák, M. Almáši, V. Huntošová, V. Hornebecq,
for overcoming multidrug resistance. Int. J. Pharm. 489 (2015) 318–330. https://
Adsorption properties, the pH-sensitive release of 5-fluorouracil and cytotoxicity
doi.org/10.1016/j.ijpharm.2015.05.010.
studies of mesoporous silica drug delivery matrix, Appl. Surf. Sci. 504 (2020)
[10] M. Vallet-Regí, M. Colilla, I. Izquierdo-Barba, M. Manzano, Mesoporous silica
144028, https://doi.org/10.1016/j.apsusc.2019.144028.
nanoparticles for drug delivery: current insights, Molecules 23 (2017) 47, https://
[32] D.B.G. Williams, M. Lawton, Drying of organic solvents: quantitative evaluation of
doi.org/10.3390/molecules23010047.
the efficiency of several desiccants, J. Org. Chem. 75 (2010) 8351–8354, https://
[11] M.M. Abeer, P. Rewatkar, Z. Qu, M. Talekar, F. Kleitz, R. Schmid, M. Lindén,
doi.org/10.1021/jo101589h.
T. Kumeria, A. Popat, Silica nanoparticles: a promising platform for enhanced oral

13
E. Beňová et al. Applied Surface Science 561 (2021) 150011

[33] D. Zhao, J. Feng, Q. Huo, N. Melosh, G.H. Fredrickson, B.F. Chmelka, G.D. Stucky, [49] T. Higuchi, Mechanism of sustained-action medication. Theoretical analysis of rate
Triblock Copolymer Syntheses of Mesoporous Silica with Periodic 50 to 300 of release of solid drugs dispersed in solid matrices, J. Pharm. Sci. 52 (1963)
Angstrom Pores, Science 279 (1998) 548–552, https://doi.org/10.1126/ 1145–1149, https://doi.org/10.1002/jps.2600521210.
science.279.5350.548. [50] S. Kockisch, G.D. Rees, J. Tsibouklis, J.D. Smart, Mucoadhesive triclosan-loaded
[34] https://www.vanderbilt.edu/viibre/CellCultureBasicsEU.pdf, last accessed 7.1. polymer microspheres for application to the oral cavity: preparation and controlled
2021. release characteristics, Eur. J. Pharm. Biopharm. 59 (2005) 207–216, https://doi.
[35] J. Rouquerol, F. Rouquerol, P. Llewellyn, G. Maurin, K. Sing, Adsorption by org/10.1016/j.ejpb.2004.07.007.
powders and porous solids, first ed., Academic Press, Cambridge, Massachusetts, [51] J. Han, Y. Jun, S.H. Kim, H.H. Hoang, Y. Jung, S. Kim, J. Kim, R.H. Austin, S. Lee,
2012. S. Park, Rapid emergence and mechanisms of resistance by U87 glioblastoma cells
[36] C.A. Cheng, T. Deng, F.C. Lin, Y. Cai, J.I, Zink, Supramolecular nanomachines as to doxorubicin in an in-vitro tumor microfluidic ecology, Proc. Natl. Acad. Sci. 113
stimuli-responsive gatekeepers on mesoporous silica nanoparticles for antibiotic (2016) 14283–14288, https://doi.org/10.1073/pnas.1614898113.
and cancer drug delivery, Theranostics, 9 (2019), 3341–3364. https://doi.org [52] V. Němcová-Fürstová, D. Kopperová, K. Balušíková, M. Ehrlichová, V. Brynychová,
/10.7150/thno.34576. D. Václavíková, P. Daniel, P. Souček, J. Kovář, Characterization of acquired
[37] Z. Li, D.L. Clemens, B.Y. Lee, B.J. Dillon, M.A. Horwitz, J.I. Zink, Mesoporous silica paclitaxel resistance of breast cancer cells and involvement of ABC transporters,
nanoparticles with pH-sensitive nanovalves for delivery of moxifloxacin provide Toxicol. Appl. Pharmacol. 310 (2016) 215–228, https://doi.org/10.1016/j.
improved treatment of lethal pneumonic tularemia, ACS Nano 24 (2015) taap.2016.09.020.
10778–10789, https://doi.org/10.1021/acsnano.5b04306. [53] X.Y. Hu, J.Y. Liang, X.J. Guo, L. Liu, Y.B. Guo, 5-Fluorouracil combined with
[38] H. Meng, M. Xue, T. Xia, Y.-L. Zhao, F. Tamanoi, J.F. Stoddart, J.I. Zink, A.E. Nel, apigenin enhances anticancer activity through mitochondrial membrane potential
Autonomous in vitro anticancer drug release from mesoporous silica nanoparticles (ΔΨm)-mediated apoptosis in hepatocellular carcinoma, Clin. Exp. Pharmacol.
by pH-sensitive nanovalves, J. Am. Chem. Soc. 132 (2010) 12690–12697, https:// Physiol. 42 (2015) 146–153, https://doi.org/10.1111/1440-1681.12333.
doi.org/10.1021/ja104501a. [54] H. Zhao, Q. Liu, S. Wang, F. Dai, X. Cheng, W. Chen, M. Zhang, D. Chen, In vitro
[39] L. Bai, Q. Zhao, J. Wang, Y. Gao, Z. Sha, D. Di, N. Han, Y. Wang, J. Zhang, S. Wang, additive antitumor effects of dimethoxycurcumin and 5-fluorouracil in colon
Mechanism study on pH-responsive cyclodextrin capped mesoporous silica: effect cancer cells, Cancer Med. 6 (2017) 1698–1706, https://doi.org/10.1002/
of different stalk densities and the type of cyclodextrin, Nanotechnology 26 (2015) cam4.1114.
165704, https://doi.org/10.1088/0957-4484/26/16/165704. [55] W. Liu, Y. Zhu, F. Wang, X. Li, X. Liu, J. Pang, W. Pan, Galactosylated chitosan-
[40] Z. Luan, J.A. Fournier, J.B. Wooten, D. Miser, Preparation and characterization of functionalized mesoporous silica nanoparticles for efficient colon cancer cell-
(3-aminopropyl)triethoxysilane-modified mesoporous SBA-15 silica molecular targeted drug delivery, R. Soc. Open Sci. 5 (2018), https://doi.org/10.1098/
sieves, Microporous Mesoporous Mater. 83 (2005) 150–158, https://doi.org/ rsos.181027.
10.1016/j.micromeso.2005.04.006. [56] S. Zhou, C. Ding, C. Wang, J. Fu, UV-light cross-linked and pH de-cross-linked
[41] M. Thommes, K. Kaneko, A.V. Neimark, J.P. Olivier, F. Rodriguez-Reinoso, coumarin-decorated cationic copolymer grafted mesoporous silica nanoparticles
J. Rouquerol, K.S.W. Sing, Physisorption of gases, with special reference to the for drug and gene co-delivery in vitro, Mater. Sci. Eng. C 108 (2020) 110469,
evaluation of surface area and pore size distribution (IUPAC technical report), Pure https://doi.org/10.1016/j.msec.2019.110469.
Appl. Chem. 87 (2015) 1051–1069, https://doi.org/10.1515/pac-2014-1117. [57] J. Fang, S. Zhang, X. Xue, X. Zhu, S. Song, B. Wang, L. Jiang, M. Qin, H. Liang,
[42] J.C. Doadrio, E.M.B. Sousa, I. Izquierdo-Barba, A.L. Doadrio, J. Perez-Pariente, L. Gao, Quercetin and doxorubicin co-delivery using mesoporous silica
M. Vallet-Regí, Functionalization of mesoporous materials with long alkyl chains as nanoparticles enhance the efficacy of gastric carcinoma chemotherapy, Int. J.
a strategy for controlling drug delivery pattern, J. Mater. Chem. 16 (2006) Nanomed. 13 (2018) 5113–5126, https://doi.org/10.2147/IJN.S170862.
462–466, https://doi.org/10.1039/B510101H. [58] L. Blascakova, D. Horvath, D. Belej, G. Wagnieres, P. Miskovsky, D. Jancura,
[43] L. Polo, N. Gómez-Cerezo, A. García-Fernández, E. Aznar, J.L. Vivancos, D. Arcos, V. Huntosova, Hypericin can cross barriers in the chicken’s chorioallantoic
M. Vallet-Regí, R. Martínez-Máñez, Mesoporous bioactive glasses equipped with membrane model when delivered in low-density lipoproteins, Photodiagnosis
stimuli-responsive molecular gates for controlled delivery of levofloxacin against Photodyn. Ther. 23 (2018) 306–313, https://doi.org/10.1016/j.
bacteria, Chem. Eur. J. 24 (2018) 18944–18951, https://doi.org/10.1002/ pdpdt.2018.07.009.
chem.201803301. [59] L. Lenkavska, L. Blascakova, Z. Jurasekova, M. Macajova, B. Bilcik, I. Cavarga,
[44] S. Pathan, P. Solanki, A. Patel, Functionalized SBA-15 for controlled release of P. Miskovsky, V. Huntosova, Benefits of hypericin transport and delivery by low-
poorly soluble drug erythromycin, Microporous Mesoporous Mater. 258 (2018) and high-density lipoproteins to cancer cells: from in vitro to ex ovo,
114–121, https://doi.org/10.1016/j.micromeso.2017.09.012. Photodiagnosis Photodyn. Ther. 25 (2019) 214–224, https://doi.org/10.1016/j.
[45] M. Hamzehloo, J. Karimi, K. Aghapoor, H. Sayahi, H.R. Darabi, The synergistic pdpdt.2018.12.013.
cooperation between MCM-41 and azithromycin: a pH responsive system for drug [60] K. Siposova, V. Huntosova, Y. Shlapa, L. Lenkavska, M. Macajova, A. Belous,
adsorption and release, J. Porous Mater. 25 (2018) 1275–1285, https://doi.org/ A. Musatov, Advances in the study of cerium oxide nanoparticles: new insights into
10.1007/s10934-017-0538-3. antiamyloidogenic activity, ACS Appl. Bio Mater. 2 (2019) 1884–1896, https://doi.
[46] A.M. El-Kady, M.M. Farag, A.M.I. El-Rashedi, Bioactive glass nanoparticles org/10.1021/acsabm.8b00816.
designed for multiple deliveries of lithium ions and drugs: curative and restorative [61] R.W. Moreadith, T.X. Viegas, M.D. Bentley, J.M. Harris, Z. Fang, K. Yoon, B.
bone treatment, Eur. J. Pharm. Sci. 91 (2016) 243–250, https://doi.org/10.1016/j. Dizman, R. Weimer, B.P. Rae, X. Li, C. Rader, D. Standaert, W. Olanow, Clinical
ejps.2016.05.004. development of a poly(2-oxazoline) (POZ) polymer therapeutic for the treatment of
[47] N. Zirak, A. Bolandparvaz Jahromi, E. Salahinejad, Vancomycin release kinetics parkinson’s disease – proof of concept of POZ as a versatile polymer platform for
from Mg–Ca silicate porous microspheres developed for controlled drug delivery, drug development in multiple therapeutic indications, Eur. Polym. J. 88 (2017)
Ceram. Int. 46 (2020) 508–512, https://doi.org/10.1016/j.ceramint.2019.08.290. 524–552. https://doi.org/10.1016/j.eurpolymj.2016.09.052.
[48] S. Radin, T. Chen, P. Ducheyne, The controlled release of drugs from emulsified sol-
gel processed silica microspheres, Biomaterials 30 (2009) 850–858, https://doi.
org/10.1016/j.biomaterials.2008.09.066.

14