Silicon Nitride: A Bioceramic with a Gift | ACS Applied Materials

Jun 28, 2019 - Si3N4 Material and Its Surface Chemistry Characterizations ... were stained with 40 mM Alizarin Red S for 10 min at room temperature. ...
23 downloads 0 Views 9MB Size
Download PDF
Research Article Cite This: ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

www.acsami.org

Silicon Nitride: A Bioceramic with a Gift Giuseppe Pezzotti*,†,‡,§,∥ †

Ceramic Physics Laboratory, Kyoto Institute of Technology, Sakyo-ku, Matsugasaki, Kyoto 606-8585, Japan Department of Orthopedic Surgery, Tokyo Medical University, 6-7-1 Nishi-Shinjuku, Shinjuku-ku, Tokyo 160-0023, Japan § The Center for Advanced Medical Engineering and Informatics, Osaka University, 2-2 Yamadaoka, Suita 565-0854, Osaka, Japan ∥ Department of Immunology, Graduate School of Medical Science, Kyoto Prefectural University of Medicine, Kamigyo-ku, 465 Kajii-cho, Kyoto 602-8566, Japan Downloaded via UNIV OF SOUTHERN INDIANA on July 25, 2019 at 02:40:27 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

‡

ABSTRACT: In the closing decades of the 20th century, silicon nitride (Si3N4) was extensively developed for high-temperature gas turbine applications. Technologists attempted to take advantage of its superior thermal and mechanical properties to improve engine reliability and fuel economy. Yet, this promise was never realized in spite of the worldwide research, which was conducted at that time. Notwithstanding this disappointment, its use in medical applications in the early 21st century has been an unexpected gift. While retaining all of its engineered mechanical properties, it is now recognized for its peculiar surface chemistry. When immersed in an aqueous environment, the slow elution of silicon and nitrogen from its surface enhances healing of soft and osseous tissue, inhibits bacterial proliferation, and eradicates viruses. These benefits permit it to be used in a wide array of different disciplines inside and outside of the human body including orthopedics, dentistry, virology, agronomy, and environmental remediation. Given the global public health threat posed by mutating viruses and bacteria, silicon nitride offers a valid and straightforward alternative approach to fighting these pathogens. However, there is a conundrum behind these recent discoveries: How can this unique bioceramic be both friendly to mammalian cells while concurrently lysing invasive pathogens? This unparalleled characteristic can be explained by the pH-dependent kinetics of two ammonia speciesNH4+ and NH3both of which are leached from the wet Si3N4 surface. KEYWORDS: silicon nitride, bacterial lysis, osteogenesis, osteoblastogenesis, reactive nitrogen species, surface hydrolysis reducing the need of administering the drugs.2,7−9 However, the majority of the abiotic surfaces so far proposed counteract microbes and also inflict, in the medium or long range, damage to cells by altering some of their key structures. It is indeed difficult to envision a solid-state substance whose surface can simultaneously target the weaknesses of bacteria while supporting the physiological activities of eukaryotic cells. Most plastics allow, and even support, bacterial proliferation through the utilization of specific hydrocarbons,10 whereas metal-mediated formation of free radicals equally harms both prokaryotic and eukaryotic cells through modifications to DNA bases, enhanced lipid peroxidation, and changes in calcium and sulfhydryl homeostasis.11 On the other hand, most bioceramics partially reduce bacterial adhesion but remain conspicuously bioinert with very limited interactions between the ceramic surfaces and the surrounding tissues.12 In this context, silicon nitride appears to be an exception. In this paper, the Si3N4 surface is shown to exert a dual action in concurrently killing bacteria and boosting up osteoblast differentiation, proliferation, and bone formation. Through comparative data on different types of cells and bacteria, it is also demonstrated how

1. INTRODUCTION Bacterial attachment and proliferation on abiotic surfaces and the related risk of bacterial contamination of the peri-implant tissue are key factors in lowering the overall success rate in medical implant devices and remain a significant clinical concern for premature failure of artificial prosthesis.1 The excessive and inappropriate use of antibiotics has exacerbated this problem by producing a number of resistant bacterial strains. This issue, which represents a serious threat to public health, is exacerbated by the rapid development of drugresistant microbes2,3 and the high costs involved with the establishment of new drugs.3 As a matter of fact, a number of relevant bacteria have already developed the ability to resist new drugs through the exploitation of concurrent biological mechanisms of resistance.4 The exploited mechanisms might be directed at the antibiotic itself (i.e., through newly synthesized or modified enzymes), target how the drug is transported (i.e., ribosomes, metabolic enzymes, or proteins involved in DNA replication or cell wall synthesis), or alter the intracellular target of the drug (e.g., ribosomes, metabolic enzymes, or proteins involved in DNA replication or cell wall synthesis).5,6 In such a scenario of increasing bacterial resistance to antibiotics, novel approaches have been searched for, which include the possibility of engineering biomaterial surfaces that intrinsically deliver antibacterial agents, thus © XXXX American Chemical Society

Received: May 8, 2019 Accepted: June 28, 2019 Published: June 28, 2019 A

DOI: 10.1021/acsami.9b07997 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

The arithmetic average of the three tests was then plotted to compare differences. The receptor activator of nuclear factor (NF)-kB ligand (RANKL), a membrane-bound protein cleaved into soluble sRANKL by metalloproteinase 14, was used as a probe to test for the propensity to form osteoclasts. The amount of sRANKL was quantified as a function of time from 1 to 7 days in cell-conditioned media using the R&D System ELISA kits MTR00 according to the manufacturer’s instructions. Micrographs of the living cells were also collected timelapse after 8, 16, and 24 h of exposure to the Si3N4 substrates were collected by means of a 3D laser scanning microscope (VK-X200 K Series, Keyence, Osaka, Japan) using a 150× objective lens with a numerical aperture of 0.9. Bone mineralization was assessed by means of the Alizarin Red S (Sigma-Aldrich) method upon staining in time lapse up to 7 days. Cells were washed twice and then fixed in 95% ethanol for 10 min at room temperature. Following washing in distilled water, cells were stained with 40 mM Alizarin Red S for 10 min at room temperature. For the quantification of staining density, concentrations were determined by measuring the absorbance at 562 nm. Also, in this case, tests were repeated three times (n = 3) for each investigated exposure time, and the arithmetic average was used for comparison. Real-time detection of nitric oxide (NO) production in living SaOS-2 cells was performed through fluorescence imaging by means of a membrane permeable fluorescent indicator DAF-2(NO) (Goryo Chemical, Inc., Sapporo, Japan). This indicator consisted of diaminofluorescein-2 diacetate. Once inside the SaOS-2 cells, this substance is deacetylated by intracellular esterases and can be detected with excitation/emission maxima of 495/515 nm. The dyeloaded cells were incubated at 37 °C osteogenic medium (complete medium supplemented with ≫50 μg/mL ascorbic acid, 10 mM βglycerol phosphate, and 100 nM dexamethasone) and observed in situ on the Si3N4 substrate at increasing exposure times between 0 and 24 h at intervals of 8 h. Fluorescent intensities of DAF-2 were determined by confocal microscopy (BZ X710, Keyence, Japan). Dye-loaded cells were excited at 488 nm of a krypton/argon laser for DAF-2, and increases in DAF-2 fluorescence were monitored in situ with the SaOS-2 cells on the Si3N4 substrate. The KUSA-A1 mesenchymal cells were acquired from the Japanese Collection of Research Bioresources Cell Bank (JCRB, Osaka, Japan). KUSA-A1 is a bone marrow stromal stem cell line capable of differentiating into osteoblasts, chondrocytes, and myotubes under inducing conditions.14,15 The KUSA-A1 cells were first cultured and incubated in an osteoblast-inducer medium consisting of 4.5 g/L of glucose DMEM (D-glucose, L-glutamine, phenol red, and sodium pyruvate) supplemented with 10% fetal bovine serum. Cells were allowed to proliferate within Petri dishes for 24 h at 37 °C. The final cell concentration was adjusted at 5 × 105 cells/well. The cultured cells were then deposited on the top surface of the Si3N4 substrates previously sterilized by exposure to UV-C light. During testing, cell seeding took place in an osteogenic medium, which consisted of DMEM supplemented with nominal amounts of the following constituents: 50 μg/mL of ascorbic acid, 10 mM ß-glycerol phosphate, 100 mM hydrocortisone, and 10% fetal bovine serum. Samples were incubated at 37 °C for 14 days. The medium was changed twice during the incubation period. Cell proliferation was assessed using an initial cell density of 5 × 105 cells/mL seeded onto the substrates within conventional Petri dishes. Cells were allowed to proliferate within the Petri dish at 37 °C for up to 9 days. For visualization, the cells were stained for fluorescence microscopy with Phalloidin Alexa Fluor (green; F-actin) and Hoechst 33342 (blue; nuclei) for 1 h. They were then washed three times with 1 mL of TBST solution (mixture of Tris-buffered saline and Tween 20). Fluorescence spectroscopy images were taken at 8, 16, and 24 h with a fluorescence microscope of the type BZ-X700 (Keyence, Osaka, Japan). Insulin-like growth factor 1 (IGF-1) and vascular endothelial growth factor signaling were used as a probe for measuring cell proliferation and differentiation efficiency. IGF-1 is a protein stimulator for differentiation and apatite growth,16 which modulates bone growth through endocrine/paracrine and autocrine mechanisms. Human IGF-1 ELISA (RayBio; RayBiotech, Inc., Norcross, GA) was

these concomitant properties arise from nitrogen chemistry and the kinetics it develops in an aqueous environment through the formation of reactive nitrogen species (RNS) at the biomolecular interface with cells. Exogenous RNS molecules activate endogenous signaling that activates osteoblast differentiation/proliferation while acting as a potent antibiotic against both Gram-positive and Gram-negative bacteria. Upon elucidating here the unique surface chemistry and the friendly nitrogen kinetics of silicon nitride bioceramics, a path is opened to new opportunities in applicative fields including orthopedics, dentistry, virology, agronomy, and environmental remediation.

2. METHODS 2.1. Si3N4 Material and Its Surface Chemistry Characterizations. The Si3N4 substrate samples used in the present study were produced by SINTX (Salt Lake City, UT, USA) using conventional sintering techniques. The material contained minor fractions of yttria and alumina as sintering aids. The microstructure of the dense sintered samples exhibited a bimodal granular population, consisting of a minor fraction of relatively large acicular ß-Si3N4 grains dispersed in a finer matrix of equiaxed grains. The sintering additives resulted in thin grain boundaries and multiple grain junctions composed of partly crystallized silicon−yttrium−aluminum−oxynitride phases. More details on the processing of the Si3N4 substrate samples have been published elsewhere.13 A Shimadzu ICPE-9820 simultaneous inductively coupled plasma (ICP) optical emission spectrometer was used to measure the concentration of Si eluted from the Si3N4 bulk sample as a function of pH and time. In addition to the axial view, this apparatus provides radial view plasma observation in the perpendicular direction. Such a dual view capability enabled Si elemental analyses across a broad concentration range. A mini-torch system reduced the consumption and required purity (99.95%) of the argon gas. The measurement conditions were set as follows: radio frequency power, plasma gas flow rate, auxiliary gas flow rate, and carrier gas flow rate were equal to 1.2 kW, 10 L/min, 0.6 L/min, and 0.7 L/min, respectively. A calibration curve was built as an internal standard for the Si element by dissolving 0−0.5 mM/L sodium silicate (Wako Chemicals, Ltd., Kyoto, Japan) into distilled water. The Colorimetric Ammonia Assay Kit (ab83360; Abcam) was used to measure the amount of ammonium ions or ammonia eluted from the Si3N4 bulk sample as a function of pH and time. The assay used conversion of ammonia or ammonium ions into a product that reacts with the OxiRed probe to generate color (λmax = 570 nm), which was then quantified by a plate reader. The detection limit of the kit was 1 nmol of ammonia or ammonium ions. In order to build a standard calibration curve, 100 μL of 1 mM ammonium chloride standard was prepared first by adding 10 μL of the 10 mM ammonium chloride standard to 90 μL of double-distilled H2O. Then, a 1 mM standard was used to prepare standard curve dilution as described by the manufacturer’s instructions. 2.2. Cell Cultures and Related Biological Assay Characterizations. The SaOS-2 human osteosarcoma cell line was cultured and incubated in an osteoblast-inducer medium consisting of 4.5 g/L of glucose Dulbecco’s modified Eagle’s medium (DMEM) (D-glucose, Lglutamine, phenol red, and sodium pyruvate) supplemented with 10% fetal bovine serum. They were allowed to proliferate within Petri dishes for 24 h at 37 °C. The final SaOS-2 concentration was adjusted at 5 × 105 cell/mL. The cultured cells were then deposited on the top surface of each of Si3N4 substrates previously sterilized by exposure to UV light. In the cell proliferation test, cell seeding was conducted in an osteogenic medium, which consisted of DMEM supplemented with the following nominal amounts: 50 μg/mL of ascorbic acid, 10 mM β-glycerol phosphate, 100 mM hydrocortisone, and 10% fetal bovine calf serum. All samples were incubated up to 7 days at 37 °C. The medium was changed twice during the incubation period. Tests were repeated three times (n = 3) for each investigated exposure time. B

DOI: 10.1021/acsami.9b07997 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces employed to measure IGF-1 concentration in the KUSA-A1 cell culture supernatants. Samples for IGF-1 measurements were incubated for 1 h at room temperature. Then, 100 μL of streptavidin solution was added, followed by 45 min of incubation at room temperature. This was in turn followed by the addition of 100 μL of 3,3′,5,5′-tetramethylbenzidine (TMB) one-step substrate reagent and incubation for 30 min at room temperature. Finally, 50 μL of stop solution was added. Optical densities at 450 nm were read immediately after these preparation procedures. IGF-1 tests were repeated in time-lapse fashion on four individual samples (n = 4) for each investigated exposure time, and their arithmetic averages were plotted and compared. The Mouse Gla-Osteocalcin High Sensitive EIA Kit (Cat. #MK127; Takara Bio, Inc., Kusatsu, Shiga Prefecture, Japan) enables to recognize the Gla-type osteocalcin produced by mature osteoblasts. This kit included a monoclonal antibody that specifically recognized the γ-carboxyglutamate (Gla) residues of osteocalcin. The Mouse GlaOsteocalcin High Sensitive EIA Kit was selected because it does not cross-react with bovine antigens. Accordingly, it was directly used with cell culture media containing fetal bovine serum. According to the manufacturer’s recommendations for applying the Gla assay, the Kit’s reagent (100 μL) and samples were prepared in advance in separate well plates and quickly added to the antibody-coated microtiter plate at room temperature for 1 h. The reaction mixture was subsequently discarded, and the samples were rinsed three times with 0.1% wash buffer (Tween 20/PBS). Next, 100 μL of the peroxidase-labeled antibody solution was added to each well using an eight-channel pipette and allowed to react for 1 h at room temperature. After discarding the reaction mixture and washing four times with 0.1% wash Tween 20/PBS, 100 μL of substrate solution (TMB) was added to each well using an eight-channel pipette and reacted at room temperature for 15 min. The reaction was stopped by intimately mixing 100 μL of highly viscous stop solution to each well. The reaction was considered finished after about 1 h by the formation of a stable color. Absorbance at 450 nm was then measured by setting zero with distilled water and using standard plots to determine the concentrations of Gla-osteocalcin. 2.3. Bacteria Cultures and Microbial Viability Assays. Freezedried pellets of Staphylococcus epidermidis (ATCCTM 14990) (S. epidermidis, henceforth) were hydrated in heart infusion (HI) broth (Nissui, Tokyo, Japan) and incubated at 37 °C for 18 h in brain HI (BHI) agar (Nissui). The mixture was subsequently assayed for colony-forming units (cfus) and diluted to a concentration of 1 × 108 cfu mL−1 using phosphate-buffered saline (PBS). An aliquot of 100 μL of the bacterial suspension was spread onto individual BHI agar plates. The Si3N4 substrate samples (previously UV-sterilized) were then placed in contact with the agar for inoculation purposes, followed by incubation at 37 °C under aerobic conditions for 12, 24, and 48 h. At each time point, the test and control samples were observed by fluorescence microscopy (BZ-X700; Keyence, Osaka, Japan). For visualization, bacteria were stained with different solutions: (i) 4′,6-diamidino-2-phenylindole (DAPI) which binds to and stains DNA blue, thereby imaging the nucleus location; (ii) 5(6)carboxyfluorescein diacetate (CFDA; Dojindo, Kumamoto, Japan); and (iii) propidium iodide (PI; Dojindo, Kumamoto, Japan). PI’s red color highlighted dead or injured bacteria; CFDA’s green color revealed living bacteria. The staining protocol consisted of adding 1 μL of DAPI, the phosphatidylinositol (PI) solution, and 15 μL of CFDA solution to the samples and then incubating for 5 min at 37 °C. After removing the buffer, the cells were stained and analyzed under the fluorescence microscope. The exposure time for each image was 5 s under an 80 W metal halide lamp. Quantitative assessments of the presence of living bacteria on the Si3N4 substrates were then directly obtained from automatic image analyses of the above micrographs by using >12 different fluorescence micrographs for each sample and n = 4 samples for each exposure time condition. NO real-time activity within living bacteria was monitored by fluorescence imaging using a membrane permeable fluorescent indicator DAF-2(NO) (i.e., diaminofluorescein-2 diacetate, Goryo Chemical, Inc., Sapporo, Japan). Once inside the bacteria, this

substance was deacetylated via intracellular esterase, and it was detected with an excitation/emission maximum of 495/515 nm. Experiments using DAF-2 were performed in a dark room because the dye is light-sensitive. Then, the samples were transferred into a chamber on the stage of the fluorescence microscope, and intensities of DAF-2 were determined by in situ confocal microscopy using a 488 nm krypton/argon laser as the excitation source. Escherichia coli (25922 ATCC) (denoted as E. coli henceforth) was cultured at Kyoto Prefectural University of Medicine at 37 °C using BHI agar (Nissui, Tokyo, Japan). Starting from an initial 1.0 × 109 cfu/mL, the concentration was diluted with PBS at physiological pH and ionic strength. Subsequently, 100 μL of the bacterial suspension at a density of 1 × 108 cfu/mL was spread onto a BHI agar plate. The ceramic substrate samples were sterilized by UV and pressed into the bacteria on BHI agar for inoculation. The incubation took place at 37 °C under aerobic conditions for 12, 24, and 48 h. The conditions for staining and fluorescence microscopy observation were the same as those described above for S. epidermidis colonies. 2.4. High-Resolution TEM. Transmission electron microscopy (TEM) was carried out at Oak Ridge National Laboratory (Oak Ridge, TN, USA). TEM images were acquired on sample cross sections obtained by a focused electron beam (FIB) procedure without any fixing procedure or manipulation. A Hitachi HF-3300 FETEM was used in scanning TEM mode to record high-resolution images of the cell-grown bony tissue. Energy-dispersive X-ray spectroscopy (EDX) analyses were carried out at the Kyoto University Microstructural Characterization Platform (Uji, Kyoto, Japan). In these latter experiments, cross-sectional samples were prepared with a focused ion beam machining apparatus (JEM9310FIB, JEOL, Tokyo, Japan) equipped with a Ga-ion source operating at 5−30 kV. EDX analyses were carried out with a monochromated atomic resolution analytical electron microscope (JEM-ARM200F Double Wien filter, JEOL, Tokyo, Japan) operating at 200 kV. This equipment was capable of high-sensitivity EDX elemental mapping with atomic resolution. 2.5. In Situ Time-Lapse Raman Spectroscopy. Time-lapse Raman spectra were collected on living cells and bacteria as cultured (on silica glass; Silanized Slides; DAKO, Denmark) or exposed to Si3N4 substrates. Highly spectrally resolved spectra were collected using a confocal Raman microscope (LabRAM HR800; Horiba/Jobin Ivon, Kyoto, Japan), which employed a single monochromator connected to an air-cooled CCD detector (Andor DV420-OE322; 1024 × 256 pixels). The excitation radiation was provided by a coherent helium−neon lamp emitting at 532 nm with a power of 10 mW. A 600 gr/mm grating and a D.03 holographic notch filter were employed with a 100× objective lens. A 100 mm cross-slit and a 200 μm confocal pinhole were applied throughout all the experiments. The Raman spectra were acquired with an exposure time of 15 s in selected spectral ranges. For each condition examined, average Raman spectra were calculated from 10 measurements at different arbitrary locations. The signal from a neon lamp was systematically collected together with each spectrum recorded and used as a reference frequency. According to preliminary calibrations, this procedure allowed reaching a spectral resolution better than ±0.2 cm−1. Raman spectral acquisition and preprocessing of raw data (i.e., baseline subtraction, smoothing, normalization, and fitting) were carried out using commercially available software (LabSpec, Horiba/Jobin-Yvon, Kyoto, Japan and Origin 8.5, OriginLab Co., Northampton, MA, USA). Gaussian−Lorentzian weighted hybrid functions were used for fitting, with a Gaussian weight higher than 0.75 for all bands. A Savitzky−Golay second derivative filter has been used for signal smoothing on windows of 8, 12, or 16 points depending on the quality of the data. Single spectra were normalized after baseline removal. In situ Raman mapping experiments on osteoblasts were conducted using a dedicated instrument (RAMANtouch, Nanophoton Co., Osaka, Japan) with a 20× immersion-type optical lens. This spectroscope allowed ultrafast imaging of up to 400 spectra simultaneously, thus collecting maps in a time faster than the movement of the cells. The excitation source was at 785 nm, and the C

DOI: 10.1021/acsami.9b07997 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces spectral resolution was 1.2 cm−1 (spectral pixel resolution equal to 0.3 cm−1/pixel). The software attached to this equipment automatically provided average spectra on selected areas. 2.6. Statistical Analysis. The experimental data were analyzed with respect to their statistical meaning by computing their mean value ± one standard deviation. Student’s t-test was also applied with p values 10.24−26 Because of the different sensitivity of the methods used to measure NH3/NH4+ and Si ions, it was not possible to detect the elution of Si ions at short times. Moreover, extrapolations of Si-ion concentrations at significantly shorter times are not easy to obtain because of the unknown time dependence. With measurements made over different time spans, a direct comparison between the concentrations of the two different eluted species is thus difficult. NH3/NH4+ data at 16 h from ref 23 indicated concentrations of 0.33 and 0.44 mMol/dm3 at pH 4.7 and 8.5, respectively. These data are comparable with those obtained by extrapolating the NH3/NH4+ concentration dependences given in Figure 2a to the same time. On the other hand, the Si-ion concentration detected at a similar time appears larger by approximately a factor of 3 (cf. Figure 2b). Ionized ammonia NH4+ can enter the cytoplasmic space of cells in controlled concentrations and through specific transporters.27,28 NH4+

The nitrogen ions that elute from Si3N4 predominately become NH4+ and NH3 at low and high pH values, respectively. Their release leaves N-vacancies at the surface and generates free electrons. In turn, these electrons induce splitting of the surrounding water molecules as shown in Figure 1a. The chemical equations associated with these

Figure 1. (a) Schematic diagram of Si3N4 surfaces exposed to aqueous solution, resulting in leaching of nitrogen and the concurrent emission of free electrons; (b) pH buffering effect at the surface of a bulk Si3N4 sample embedded in acidic environment21 (upper plot) and pH dependence of the NH3/(NH3+NH4+) concentration ratio20 (lower plot). transient off-stoichiometry reactions are provided at the bottom of Figure 1a. Regardless of the prevailing nitrogen species, these reactions result in robust pH buffering, which stabilizes at a value around 8.75. As shown in Figure 1b, an experiment using a pH microscope demonstrated that this change occurs within minutes after exposure to a Si3N4 substrate in an acidic gel.21,22 The two forms of eluted nitrogen obey a steep pH dependence with a threshold value at pH = 9.25. At homeostatic pH values of about 7.3−7.5, ∼99% of the D

DOI: 10.1021/acsami.9b07997 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 3. Comparison of time-lapse results of biological assays (with statistics) on (a) SaOS-2 and (b) KUSA-A1 cell lines exposed to Si3N4 and control (silica glass) substrates; in the inset, laser and fluorescence micrographs of the cell lines, respectively, as taken at different exposure times. is a nutrient used to synthesize building block proteins for enzymes and genetic compounds, thus sustaining cell differentiation and proliferation.29 Together with the leaching of orthosilicic acid and related compounds, NH4+ promotes osteoblast synthesis of bone tissue,30,31 stimulates collagen type 1 synthesis in human osteoblasts,31−33 and fosters skin fibroblast proliferation.34 Noticeably, NH4+ is a nutrient to both eukaryotic and prokaryotic cells; both assimilate it through the glutamine synthetase/glutamate synthase pathway. Conversely, highly volatile ammonia (NH3) can freely penetrate the external membrane and directly target the stability of DNA/RNA structures in both mammalian and bacterial cells.35,36 The cellular toxicity of ammonia has been widely reported in a variety of contexts.37−40 The release of unpaired electrons upon the liberation of nitrogen is a key step in the successive cascade of reactions, which starts with NH3 oxidation into hydroxylamine NH2OH (ammonia monooxygenase) along with an additional reductant contribution, leading to further oxidation into NO2− nitrite through a process of hydroxylamine oxidoreductase.36,41−43 This latter process involves nitric oxide NO formation. This reaction is associated with the release of four additional electrons, two of which partly sustain further ammonia oxidation and the remaining two support the generation of proton gradients and bacterial metabolism. From a chemistry viewpoint, the Si3N4 surface plays a role similar to that of a family of catalyzing enzymes which are referred to as inducible nitric oxide synthase (iNOS) and endothelial nitric oxide synthase (eNOS). Similar to these NOS enzymes, which generate NO from the amino acid Larginine, when Si3N4 is exposed to an aqueous solution, it produces NO from the solid state in an oxidative environment. Further reactions with superoxides generate peroxynitrite which leads to antibacterial effects similar to the oxidative iNOS burst by macrophages.44 The presence of NO also induces apoptosis in osteoclast progenitors and inhibits the resorptive activity of mature osteoclasts.45 Through these combined actions, NO production results in a potent inhibitory effect on bone resorption.41,45 Similar to the eNOS pathway, the nitrogen chemistry at the surface of Si3N4 boosts endothelial NO production, which subsequently regulates osteoblast activity and stimulates bone formation.41,46−51 It has been previously shown that the presence of Si3N4 substrates within an in vitro biological environment promotes the proliferation of mesenchymal progenitor cells and stimulates osteoblast differentiation.51 Si3N4 spinal implants were also found to accelerate regeneration of osseous tissue within the human body.52 In the following sections, the results of a series of in vitro experiments are presented on different types of both cells and bacteria. Such series of experiments was designed to evaluate the chemical interactions at the biomolecular interface between cells and the Si3N4 substrate and to validate the hypothesis of exogenously activated eNOS-/iNOS-like actions as promoted by the ionic species eluted from the ceramic substrate in aqueous solution. 3.2. Concurrent Antibacterial and Osteogenic Actions. Figure 3a shows data obtained on SaOS-2 cells exposed to Si3N4

substrates as a function of time up to 7 days in comparison with control samples exposed to the silica glass sample under exactly the same conditions. The number of cells per unit area was monitored in time-lapse since the beginning of their exposure to the substrates. The laser micrographs on the right side of the figure show a significant cell proliferation after short-term (8−24 h) exposures to the Si3N4 substrate. Standard Alizarin Red stain photometric tests of optical density (cf. data shown in the plot) were used to assess osteoconductivity. These tests quantified the process of deposition of bony apatite by SaOS-2 cells, thus showing their osteogenic activity as a function of exposure time to the Si3N4 substrate. An exponentially increasing dependence in optical density was recorded as a function of time, which greatly differed from the low and only slightly increasing trend of the control cell sample (cf. control data in the plot). Formation of hydroxyapatite appeared to be substantiated by statistically meaningful repeatability. An additional plot in Figure 3a refers to free sRANKL assessments. RANKL is a receptor activator of NF-κB ligand, a membrane-bound protein that can be cleaved into soluble sRANKL by metalloproteinase 14. It is a key molecular regulation system for bone remodeling and the main stimulatory factor for the formation of mature osteoclasts. Its concentration thus gives a measure of the propensity of SaOS-2 osteoblasts to convert into osteoclasts. The very low levels of free sRANKL detected at any tested time in comparison with the control sample (SaOS-2 cells on the silica glass substrate) clearly indicate that the Si3N4 substrate represents an environment in which SaOS-2 cells tend to stably assume the characteristics of mature osteoblasts. Figure 3b summarizes data from biological assays applied to KUSAA1 mesenchymal cells. The fluorescence micrographs on the right side (nuclei and F-actin in blue and green, respectively) confirmed that the KUSA-A1 cells were able to proliferate on the Si3N4 substrate after exposure times as short as 24 h. The plots in Figure 3b quantitatively assess the metabolism of mesenchymal KUSA-A1 cells exposed for different times (up to 9 days) to the Si3N4 substrate. IGF-1 signaling was taken as a probe for cell proliferation and differentiation efficiency. This protein is a stimulator for both differentiation and apatite growth and a modulator of bone growth through endocrine/ paracrine and autocrine mechanisms.53,54 Because higher IGF-1 concentrations stimulate cells to fast proliferation and differentiation, the amount of IGF-1 within the supernatant quantitatively probed the metabolic propensity for the KUSA-A1 cells to differentiate into osteoblasts and subsequently generate native apatite. The results of the IGF-1 assay in the plot of Figure 3b show an exponential increase upon time with a threshold between 5 and 7 days. This trend indicates that the KUSA-A1 cells exposed to the Si3N4 substrate produced a high IGF-1 concentration, more than 300% higher than the control cell sample cultured on silica glass under the same conditions. A similar trend was observed for Gla-osteocalcin as a function of time (cf. plot in Figure 3b). Again, a significant enhancement in concentration was observed between 5 and 7 days exposure to the Si3N4 substrate. The concentration reached at 7 and 9 days was more than 1 order of magnitude larger than that recorded on a control E

DOI: 10.1021/acsami.9b07997 ACS Appl. Mater. Interfaces XXXX, XXX, XXX−XXX

Research Article

ACS Applied Materials & Interfaces

Figure 4. Time-lapse results of biological assays (with statistics) on (a) S. epidermidis and (b) E. coli bacterial strains exposed to the Si3N4 substrate; in the inset, DAPI blue (nuclei), CFDA green (live), and PI red (dead) stained fluorescence micrographs of the two strains at 48 h exposure time. Note the formation of biofilms and the isolated colonies in S. epidermidis and E. coli, respectively.

Figure 5. (a) Optical image of living SaOS-2 cells cultivated for 24 h on a Si3N4 substrate and the relative location of NO molecules upon staining with DAF green; the plot gives the time dependences of cell population and NO concentration as a function of time (with statistics); (b) DAPI blue image of S. epidermidis bacteria exposed for 24 h on a Si3N4 substrate and the relative location of NO molecules upon staining with DAF green; the plot gives the time dependences of bacteria population and NO concentration as a function of time (with statistics). The threshold for statistical relevance, p, was set at