Adhesion of human gingival fibroblasts/Streptococcus mitis co-culture on the nanocomposite system Chitlac-nAg

Composite materials are increasingly used as dental restoration. In the field of biomaterials, infections remain the main reason of dental devices failure. Silver, in the form of nanoparticles (AgNPs), ions and salt, well known for its antimicrobial properties, is used in several medical applications in order to avoid bacterial infection. To reduce both bacterial adhesion to dental devices and cytotoxicity against eukaryotic cells, we coated BisGMA/TEGDMA methacrylic thermosets with a new material, Chitlac-nAg, formed by stabilized AgNPs with a polyelectrolyte solution containing Chitlac. Here we analyzed the proliferative and adhesive ability of human gingival fibroblasts (HGFs) on BisGMA/TEGDMA thermosets uncoated and coated with AgNPs in a coculture model system with Streptococcus mitis. After 48 h, HGFs well adhered onto both surfaces, while S. mitis cytotoxic response was higher in the presence of AgNPs coated thermosets. After 24 h thermosets coated with Chitlac as well as those coated with Chitlac-nAg exerted a minimal cytotoxic effect on HGFs, while after 48 h LDH release raised up to 20 %. Moreover the presence of S. mitis reduced this release mainly when HGFs adhered to Chitlac-nAg coated thermosets. The reduced secretion of collagen type I was significant in the presence of both surfaces with the co-culture system even more when saliva is added. Integrin β1 localized closely to cell membranes onto Chitlac-nAg thermosets and PKCα translocated into nuclei. These data confirm that Chitlac-nAg have a promising utilization in the field of restorative dentistry exerting their antimicrobial activity due to AgNPs without cytotoxicity for eukaryotic cells.


Introduction
Composite materials are increasingly used in restorative dentistry, but in the broad field of biomaterials, infections remain the main reason of dental devices failure [1]. The surface of the oral cavity is exposed to a broad variety of microorganisms capable of forming biofilms [2], not only on oral mucosa and teeth surfaces, but also on the components used for dental restoration [3,4]. It is well known that resin-based composites are susceptible to microbial colonization and to biofilm formation in the oral cavity. Oral microbial biofilms produce acid and are involved in the occurrence of periodontitis and dental caries [5]. Recently, in order to avoid the proliferation and the bacterial adhesion on material surface, materials with antimicrobial properties have been set up [6,7]. Silver, in the form of nanoparticles, ions and salt, well known for its antimicrobial properties, is used in several medical and general applications in order to avoid bacterial infection [8,9]. Silver ions and nanoparticles are capable to destroy the bacterial cell wall by means of the reaction with electron donor groups. Silver acts on sulfhydryl groups on transmembrane and outer-membrane proteins, including proteins of the electron transport chain, protruding into the extracellular portion of the membrane [10], while in eukaryotic cells lacking these extracellular binding sites, the potential cytotoxic effect could be due to silver internalization [11]. This nanoparticles diffusion into the cells can lead to their death interfering with several metabolic pathways [10]. During the preparation, to prevent silver nanoparticles aggregation, which leads to the loss of the antimicrobial activity, associated with the nanoscale [12], the polyelectrolyte Chitlac (lactose-modified chitosan) at low concentrations is used to stabilize silver nanoparticles colloidal solutions [11]. Immobilizing silver nanoparticles into a biocompatible polymeric film in order to allow nanoparticles interaction with bacterial membrane without affect eukaryotic cells is necessary to minimize the interaction between eukaryotic cells and silver nanoparticles [11]. In this work the substrate material is a thermoset based on Bisphenol A glycidylmethacrylate (BisGMA) and triethylenglicol dimethacrylate (TEGDMA), widely used for dental devices [13]. Chitlac-nAg coated methacrylic thermosets have been already tested in order to assay the interaction of the coating either with bacteria either with eukaryotic cells [11]. It cannot be forgotten that oral cavity environment consists of eukaryotic gingival epithelial and fibroblasts cells, human oral microbiota and saliva. The oral surface is kept wet by saliva [14]. This fluid is essential for the maintenance of the oral ecosystems by ensuring water, nutrients, adherence and antimicrobial factors. Saliva contains 99 % water and glycoproteins (for example mucins), proteins, hormones, vitamins, urea and several ions. The role of mucins in bacterial adherence is complex. When salivary glycoproteins are adsorbed on solid surface, they may bind to bacteria and promote bacterial adherence [15]. Instead, some of these glycoproteins, when free in saliva, may prevent bacterial colonization by binding to their adhesion proteins or by agglutinating bacteria in saliva [16]. Within the oral microbiota, Streptococci constitute the main population of bacteria, able to foster tight associations with host epithelial tissue and resulting in positive effects on healthy oral status [17]. In particular, Streptococcus mitis represents one of the first colonizers of hard (teeth) as well as mucosal surfaces [18]. Thus, a eukaryotic/prokaryotic co-culture model, a first step to try to mimic ''in vitro'' the ''in vivo'' oral cavity environment, here was analyzed as a biological model [17,19,20].
Integrin a5b1, a heterodimeric transmembrane protein able to bind extracellular ligands as well as intracellular proteins involved in cell signaling, represents the main bidirectional receptor connecting the cytoskeleton and the extracellular matrix (ECM) in HGFs [21]. Numerous data suggest that the binding of various bacterial modulins to their cognate receptors available on different eukaryotic cells leads to the release of inflammatory cytokines playing an important role in periodontal diseases [22]. In addition, previous studies suggest that integrin b1 can also trigger PKCa isoenzyme in response to resin composite monomers released from dental materials [23].
The aim of this study was to investigate the proliferative and adhesive behaviour of HGFs on BisGMA/TEGDMA thermosets uncoated and coated with AgNPs in our coculture model. Then the molecular mechanisms driving such processes were studied in order to clarify biological reactions that can occur between biomaterials, host tissue and microbial environment. for 20 s. The postcuring was performed with a Photopol IR/UV Plus oven (Dentalfarm, Torino, Italy) equipped with eight lamps and two spots operating in the wavelength range 320-550 nm following the procedure: 20 min in light oven (eight lamps), 20 min in light oven (eight lamps) on a rotating plate, 60 min in light oven (eight lamps) under vacuum, and 7 min in light oven (eight lamps plus two spots). The thermosets were then sandpaper polished (granulometry: 1200) [11]. Chitlac (1-deoxylactit-1-yl chitosan, CAS registry number 85941-43-1) sample was prepared according to the procedure elsewhere reported [24]. BisGMA/TEGDMA thermosets were prepared and coated with Chitlac-nAg, as already reported [11].

Culture of human gingival fibroblasts
Human gingival fibroblasts (HGFs) were obtained at the Dental Clinic of the Department of Medical, Oral and Biotechnological Sciences of the University G. d'Annunzio. The tissue fragments were treated as previously described [25]. Signed informed consent was obtained from the donors. None of the authors participated to the sample collection and samples were anonymized before the authors received them. Cells were used from passage 5 up to 12.

Bacterial strains and growth condition, saliva
collection and co-culture set up The clinical strain S. mitis DS12 from a saliva sample was used in the present study. The strain was cultured as previously described [17]. Pooled unstimulated saliva was obtained from laboratory healthy staff (after no drinking or eating for two hours) by spitting method in polypropylene tube. The saliva was treated as previously described [25].
The co-culture assay was performed as reported elsewhere [25]. The experimental design was carried out for at least three independent experiments.

Scanning electron microscopy (SEM) analysis of cell and bacteria
To evaluate HGFs adhesion on thermosets (coated with Chitlac or Chitlac-nAg) and cells morphology, SEM analysis was carried out. To better assess the interaction between eukaryotic, prokaryotic cells and devices, thermosets were observed with adherent HGFs, with bacteria, or with the co-culture in the presence or absence of saliva. HGFs (12.5 9 10 3 /thermoset) were seeded on each thermoset placed in a 24-well microtiter plate. Where required, thermosets were pre-treated with saliva and HGFs were allowed adhering for 3 h.

Antibodies staining
To assess integrin b 1 and PKCa expression and localization in HGFs seeded on both thermosets (coated and uncoated with silver nanoparticles), immunofluorescence staining was performed. Cells (12500/thermoset) were seeded on each thermoset (where required pre-treated with saliva) and allowed to adhere for at least 3 h. Bacteria were added where required. After 48 h, cells were fixed for 10 min with PBS 1X/3 % paraformaldehyde supplemented with 2 % sucrose. Then cells membranes were permeabilized for 5 min at room temperature, with a pH 7.6 solution containing 0.5 % Triton X100/20 mM HEPES, 300 lM sucrose, 50 mM NaCl, 3 mM MgCl 2 . After membrane permeabilization, cells were incubated with PBS 1X/BSA 10 % for 30 min at room temperature, followed by 4 h incubation at 37°C with antiintegrin b1 mouse monoclonal antibody (Santa Cruz Biotechnology, CA, USA) 1:20 in 1 % BSA/PBS, and with anti-PKCa rabbit monoclonal antibody (Abcam, Cambridge, UK) 1:100 in 1 % BSA/PBS. Alexa-568 goat antimouse and Alexa-488 goat anti rabbit (both 1:50 in 1 % BSA/PBS) secondary antibodies were added and samples were incubated for 2 h at 37°C.

Confocal microscopy analysis
Images were acquired using a confocal microscope (LSM510 META, Carl Zeiss, Germany), equipped with an inverted microscope (Axio Observer) and an EC-Plan Neofluar 63X NA 1.25 and 100X NA 1.3 OIL lens. The excitation was obtained with an Argon laser line (488 nm) and a HeNe laser line (543 nm). The emission was collected through a primary dichroic Filter (HTF) 488/543 nm and separate with a secondary dichroic filter NTF545. In Integrin (Green)/PKC (Red) and Live (Green)/death (Red) staining, emission was recorded selecting a BP 505-530 nm for Green and BP 585-615 nm for Red. To prevent fluorescence emissions overlaps, images were acquired sequentially (Multitrack mode). Ten fields, for each slide randomly chosen were examined. Densitometry and the counts of viable adhering bacteria on thermoset or on HGF were carried out by using image analysis software (LEICA QWin, Germany).
Microscopic observations were repeated for three independent experiments.

Statistics
Statistical analysis was performed using the t-student test. Results are expressed as the mean ± SD. Values of P \ 0.05 were considered statistically significant.

Confocal 3D surface rendering and SEM
observation and quantification Figure 1 shows the 3D surface rendering of HGFs growth onto Chitlac and Chitlac-nAg thermosets stained for b1 integrin. These images allow to visualize HGFs monolayer formed on both surfaces. In order to assess the ability of HGFs, S. mitis and the co-culture to adhere onto BisGMA/ TEGDMA thermosets, SEM observations and quantifications were run out. Figure 2a shows SEM images of the different two surfaces with HGFs or S. mitis. These images show that cells uniformly proliferate creating a monolayer which adheres onto the substrate, both in presence with Chitlac-nAg (AgH) or Chitlac alone (ThH). As regard S. mitis, the percentage of cells growing onto Chitlac thermosets is significantly higher (96 % of the total surface analysed) with respect to those growing on Chitlac-nAg (6 %). Figure 2b exhibits HGFs co-cultured with S. mitis growing onto thermosets. It should be noted that bacteria percentage decreased in the presence of AgNPs (ThMH 33.06 %; AgMH 17.3 %), while when saliva is added, a greater bacterial aggregation onto the surfaces is shown (ThMHS 48.34 %, AgMHS 33.74 %). In Fig. 2c the sessile microbial proliferation with cells embedded in the extracellular matrix formation is shown.

Live/dead staining
Live/Dead assay was performed and the fluorescence intensity (FI) of the staining was measured on bacteria growing onto Chitlac and Chitlac-nAg BisGMA/TEGDMA thermosets, in order to visualize and quantify the viability and the proliferation rate of S. mitis in the different experimental situations (Fig. 3a, b). This analysis confirms SEM quantification. An enormous amount of living cells is evidenced onto Chitlac thermosets (FI live = 16.18; FI dead = 1.8), while these values decrease when S. mitis is cultured onto thermosets coated with Chitlac-nAg (FI live = 4.68; FI dead = 6.38) (Fig. 3b). When saliva is added to the culture, the amount of detectable cells through Live/Dead staining decreases, both in Chitlac and Chitlac-nAg thermosets. In Fig. 3a, images referred to ThMS and AgMS show live adherent bacterial aggregates (biofilm) stained in green. Due to bacterial aggregation, the staining overlaps and therefore a lower FI values are detected.

LDH cytotoxicity assay
Chitlac and Chitlac-nAg exert only a minimal cytotoxic effect on HGFs at 24 h (8.9 and 7.8 fold higher, respectively), while LDH release at 48 h rises up to 27.9 and 22.0  (Fig. 4a). Moreover, in the presence of S. mitis, a greater reduction of LDH release by HGFs is detected in the presence of Chitlac-nAg coated thermosets (indicated in figure as AgMH, 1.0 fold) with Chitlac alone (indicated as ThMH, 9.0 fold) after 48 h. The addition of saliva in the culture medium even more reduces LDH release after 48 h (1.9 fold Chitlac thermoset, 1.4 fold Chitlac-nAg thermoset) (Fig. 4a).

Quantification of human collagen type 1 in culture media (ELISA assay)
In order to investigate the amount of collagen secreted into the culture medium by HGFs alone or in co-culture with S. mitis in the presence of BisGMA/TEGDMA thermosets, ELISA assay was performed (Fig. 4b). After 48 h of treatment, collagen secretion rises in the presence of Chitlac coated thermosets with respect to HGFs alone (3.157 lg/ml vs 2.318 lg/ml), as well as in the presence of Chitlac-nAg thermosets (3.09 lg/ml) The co-culture system shows a decrease in collagen production dramatically evident in the presence of saliva (0.806 lg/ml ThMHS; 1.104 lg/ml AgMHS).

Confocal microscopy of integrin b1 and protein kinase Ca (PKC a)
Immunofluorescence staining images and quantification of the fluorescence intensity (FI) of integrin b1 expression are shown in Fig. 5a. As regard Chitlac thermosets, in the presence of S. mitis and saliva the amount of protein does not undergo any modifications. When HGFs are co-cultured with S. mitis onto Chitlac-nAg thermosets, the expression of total b1 integrin increases (AgMH FI = 17.94 vs AgH FI = 9.89) as well as when saliva is added (FI = 14.53). White arrows show integrin b1 localized closely to cell membranes in HGFs cultured onto Chitlac-nAg thermosets. In Fig. 5b, HGFs are stained for PKC a and analyzed through a confocal microscope. In the panel on the left, representative images of HGFs in the different experimental conditions are shown. While the amount of total protein does not undergo significant modifications, with the exception of AgMHS (FI = 13.39 vs 7.95), images show nuclear translocation of PKC a when HGFs are co-cultured with S. mitis onto both thermosets (coated with Chitlac or Chitlac-nAg) in the presence of saliva (ThMHS and AgMHS).

Discussion
Dental implant materials interact with several cell types, including osteoblasts, periodontal ligament fibroblasts, monocytes and human gingival fibroblasts [27]. Implantassociated infections are the major complications in dental surgery which could lead to implant failure. To date dental research is focused on silver nanoparticle technology for the well-known antimicrobial properties and the resulting findings are encouraging [6]. However, there are controversial results with respect to the possible adverse effects and toxicity of silver nanoparticles for eukaryotic cells [28]. Several reports focused on the ability of positively charged polyelectrolytes, chitosan for instance, to stimulate cell adhesion [26,29]. In our experimental model, coating thermosets with Chitlac and Chitlac-nAg allows HGFs forming a monolayer on both surfaces (Figs. 1, 2). SEM observations confirm the good adhesion of HGFs on both thermosets (Fig. 2a). In order to test the antimicrobial activity of our devices we used a commensal bacterial strain colonizing the oral cavity, S. mitis. Microscopic images show that in the presence of silver nanoparticles the amount of viable bacteria dramatically decrease (Fig. 2a). In order to mimic interactions occurring between eukaryotic/prokaryotic cells and composites, HGFs were co-cultured with S. mitis in the presence of saliva as previously described [25]. When co-cultured, the presence of silver nanoparticles does not affect HGFs adherence on thermosets, while the amount of viable Streptococci decreases even more in the presence of saliva. To protect themselves from a hostile environment, bacteria often form surface-attached communities described as bacterial biofilms. Biofilms are ubiquitous in natural, industrial, and clinical environments and have been shown to play a critical role in many chronic infections. Biofilms are usually composed of multiple bacterial species. For example, dental biofilms (i.e., dental plaque) contain more than 500 different bacterial species. Biofilms consist of viable microbial cells along with dead cells and a wide range of self-generated extracellular polymeric substances (EPS) including polysaccharides, nucleic acids (extracellular DNA from bacteria), and proteins [30]. The increased aggregation of S. mitis forming biofilm with saliva is confirmed by SEM observation in Fig. 2c. In order to evaluate the antimicrobial activity of Chitlac-nAg coated thermosets, Live/Dead staining was performed (Fig. 3). This confirms SEM observations with the lowest amount of viable bacterial cells observed on thermosets coated with silver nanoparticles. The staining overlaps exactly when saliva is added showing biofilm formation. Although the antibacterial mechanism is not totally understood, the main molecular targets of silver nanoparticles are supposed to be the thiol groups (-SH) of proteins exposed to the extracellular portion of the bacterial membrane; conversely, eukaryotic cells lack these exterior binding sites, so the nanoparticles are supposed to interact with them only upon metal internalization. The lack of physical barriers to nanoparticle diffusion into cells represents a risk of uptake by eukaryotic cells, which can lead to their death [10,11]. LDH release after 24 and 48 h increases on both surface when HGFs are cultivated alone (Fig. 4a). In the presence of S. mitis the amount of LDH decreases even more when saliva is added, confirming a protective role of bacteria on HGFs as previously described [25,31]. Collagen type I is the principal biopolymer in the extracellular matrix of both vertebrates and invertebrates. It is prevalent in connective tissues, and the arrangement of collagen determines the mechanical response and the adhesion properties [32]. HGFs collagen type I secretion is not affected by the presence of Chitlac and Chitlac-nAg thermosets, being even more increased when they are cultured on both surfaces, confirming again SEM observations (Fig. 4b). When HGFs are co-cultured with S. mitis, the amount of collagen is comparable with the amount in control sample, while the presence of saliva decreases the secretion of collagen type I. Previous studies demonstrated that oral tissues, such as cementum, root, and periodontal ligament fibers, rich in collagen type I, if exposed to the oral environment, may be vulnerable to attachment and colonization by organisms equipped with collagen-binding adhesions, leading to the formation of biofilm [33]. In addition to specific adhesions, the presence of immobilized fibril-forming collagens could promote the formation of dense biofilm. The induction of a dense biofilm is also observed in the presence of collagen type I in solution [34]. It is plausible to assume that the decreased collagen secretion in the supernatant detected by ELISA assay is due to the biofilm formation and the consequent digestion of collagen type I by S. mitis.
Gingival fibroblasts produce type I collagen which represents a scaffold for numerous associated proteins in Arrows indicate PKC a nuclear localization. Graph outlines the statistical quantitative expression of PKC a. * AgMHS vs AgH P \ 0.05 the extracellular matrix, but also produce integrin b1, a protein involved in the regulation of cell migration, proliferation, survival/apoptosis and differentiation. Among the intracellular signaling pathways activated by integrins, PKC isoenzymes are included [23,31]. Chitlac coated thermosets (Th) do not particularly affect the amount of total expression of integrin b1, confirming that HGFs adhere well on these surfaces (Fig. 5a). In the presence of Chitlac-nAg thermosets, the amount of integrin b1 increases and more important, it appears localized closely to the cell membrane, confirming that HGFs, notwithstanding the presence of silver nanoparticles which could disturb cell attachment, adhere well on Chitlac-nAg thermosets (Fig. 5a, white arrows). In parallel, a PKC a activation as nuclear translocation can be visualized in Fig. 5b indicated by white arrows.
Based on the above considerations, a key role of the oral environment in preventing cellular toxic events occuring in human gingival fibroblasts can be highlighted. Moreover cells are well attached on Chitlac and Chitlac-nAg and the amount of bacteria decreases in the presence of silver nanoparticles, confirming the antimicrobial properties of this coating and its biocompatibility. In conclusion, the nanocomposite layer formed by the polysaccharide Chitlac and Chitlac-nAg has a considerable potential as a good performing surface coating for biomaterials usable in restorative dentistry.