Biointerfacial cell/protein–polymer interactions investigated by quartz crystal microbalance with dissipation

Abstract

Understanding cell-surface interactions is required for the development of novel and functional biomaterials. The biocompatibility and in vivo performance of these biomaterials heavily depend on processes such as protein adsorption and subsequent cell adhesion on the material surface. Common experimental approaches used to assess these processes typically involve end-point assays. However, these assays often require cell fixation or disruption and pre-or post-labeling of the cells, potentially affecting cell physiology and leading to the loss of valuable information. Furthermore, these methods cannot distinguish interfacial interactions occurring at nanometer scales between cells and the surface. The physicochemical properties of the material surface also significantly impact the performance of potential biomaterials. The interactions taking place at the interface between cells/proteins and materials are intricate and must be comprehensively understood and carefully designed to meet specific application requirements. In this context, Quartz Crystal Microbalance with Dissipation (QCM-D) emerges as an alternative and complementary method. QCM-D serves as a powerful, noninvasive technique that enables real-time and label-free monitoring of cell-surface interactions at the nanoscale. QCM-D provides distinct data regarding specific interactions at the cell - material interface, thereby offering new insights into the cell adhesion / protein adsorption behaviors. The aim of the thesis is to investigate cell-polymer interactions and to monitor the entire process in real-time using QCM-D system. For this purpose, two commonly employed polymers in the biomaterials field, Polycaprolactone (PCL) and Chitosan (CH), as well as their blends (75:25 and 25:75), were employed to investigate real-time cell adhesion behavior. As surface topography, chemical composition and wettability have significantly influence on cell adhesion process, it is important to analyze cell adhesion on well-characterized surfaces. Two types of cell lines (hFOB and 3T3) were employed to monitor cell interactions. Complementary cell culture assays were also conducted to validate the outcomes obtained from QCM-D. In the first part of the thesis, the preparation and characterization of thin films on silicon substrates and silica sensor surfaces were completed. In order to achieve homogeneous films, various parameters were investigated, i.e., polymer ratio, solvent type, substrate surface characteristics (activated with oxygen plasma or hydrophobic treatment), polymer molecular weight, and polymer ratio. The homogeneity of the films was assessed using Atomic Force Microscopy (AFM). It was founded that blends prepared with a constant amount of chitosan yielded homogeneous coatings on the silicon substrate. The chemical composition of the constructed surface was further analyzed using Attenuated Total Reflection Fourier Transform Infrared Spectroscopy (ATR-FTIR). The spectra exhibited distinct peaks at 1725 cm-1 for PCL and 1645 cm-1 and 1584 cm-1 for chitosan, confirming the successful coating of both polymers onto the surface. Analysis of AFM images over a scanned area of 100 µm2 revealed that pure polymers produced morphologically homogeneous films, whereas the blend surfaces displayed visible domains. Particularly, 75:25 PCL/CH blend exhibited micrometer-scale domains. Comparatively, thin films prepared with pure polymers displayed smoother surfaces compared to the blends. In contrast, the blends prepared with a ratio of 75:25 exhibited the highest roughness value. In terms of zeta potential, pure PCL films exhibited the highest negative value (-88 mV), whereas pure chitosan films displayed the lowest negative value (-15 mV) at pH 7.4. Blend film zeta potential values were between these two extremes. Film thickness analysis revealed that pure chitosan films had the smallest thickness (ca. 10 nm), whereas 75:25 PCL/CH blend film had the greatest thickness (ca. 55 nm). The pure PCL film and the 25:75 blend film had thicknesses of approximately 38 nm and 14 nm, respectively. In addition, in situ spectroscopic ellipsometry was employed to determine swollen polymer thicknesses. It was observed that PCL, being a hydrophobic polymer, did not swell much in aqueous solutions. Chitosan thin films exhibited the highest degree of swelling. The blend films exhibited swelling degrees between those of the pure polymer films, while higher PCL amount (75:25) resulted in reduced swelling, as expected. Before the cell adhesion studies, protein adsorption studies onto the constructed films was conducted. Bovine Serum Albumin (BSA) adsorption was monitored in real-time using both QCM-D and spectroscopic ellipsometry at various pH values at room temperature. In the case of pure PCL film, BSA adsorption onto pure PCL film showed a consistent frequency change upon adsorption with QCM-D for all pH values investigated. However, for the other investigated films, the presence of chitosan led to pH dependent adsorption behavior. At pH 4.5, both BSA and CH were positively charged, resulting in adsorption under repulsive conditions. At pH 6.0, the electrostatic attraction between the polymer chains and BSA led to higher adsorption on films containing chitosan. The lowest frequency decrease, i.e., mass load, was observed at pH 7.4 compared to pH 4.5 and 6.0. These findings indicate that blend composition, pH and ion presence in the environment have a substantial influence on protein adsorption. To compare the adsorbed protein amounts determined by QCM-D and ellipsometry methods, diverse models were applied. When two methods are assessed, the protein quantity derived from QCM-D data was consistently higher than that obtained by ellipsometry. The amount of protein calculated from ellipsometric data was similar for the blend films and pure chitosan films for all pH values investigated. However, higher values were evident in QCM-D method due to the inclusion of coupled water in the calculations. In addition, fibrinogen adsorption presented composition dependent behavior on thin films. The highest adsorbed fibrinogen amount was monitored on pure PCL films. In contrast, no significant protein adsorption was monitored on pure chitosan films. Consequently, the adsorbed amount of fibrinogen decreased with an increasing percentage of chitosan in the films, which predominantly showed an inverse correlation with the surface hydrophilicity. Following the comprehensive characterization of the films and conduction of the protein adsorption experiments, the cell adhesion behavior of two cell lines, human fetal osteoblastic (hFOB) and mouse fibroblast (NIH/3T3), was monitored on constructed films using QCM-D. For this purpose, the cells were introduced into the QCM-D chamber and allowed to flow for 1 hour. Initial cell sedimentation after 1 h resulted in reduced cell deposition as the chitosan ratio increased in the film. This trend was consistent for the both cell lines in the first hour. Subsequently, changes in frequency and dissipation were monitored over an 18-hour period. Complementary cell culture assays were performed to validate the observations of QCM-D. For this purpose, fluorescence images and live cell images at various time intervals were captured. Distinct QCM-D signal patterns were found for the investigated cell lines, indicating the influence of the varying interfacial properties on cell adhesion, which is also dependent on the specific cell type. In the case of hFOB cells, fully spreading was observed on pure PCL films, with elongated morphologies as confirmed by fluorescence microscopy and scanning electron microscopy (SEM). Corresponding QCM-D signals showed the highest frequency drop and the highest dissipation. Blend films supported hFOB cell adhesion, but with lower dissipation values compared to the PCL film. This might be attributed to higher rigidity at the hFOB cell−blend interface, because these cells did not progress to the further stages of spreading after secretion of their extracellular matrix (ECM) proteins. Variations in the QCM-D data obtained from the blend films could be attributed to differences in the morphology of the films. Pure chitosan films showed limited hFOB cell adhesion, accompanied by low frequency drop and low dissipation. The initial sedimentation of 3T3 cells onto the constructed surfaces similarly showed dependence on the surface composition. Unlike the behavior of hFOB cells, 3T3 cell lines did not adhere to pure chitosan surfaces, evident from consistent positive frequency signals. The highest frequency change was observed on reference silica surface, with dissipation gradually decreasing. This behavior indicated an average number of cells remaining in ECM remodeling stage. The ΔD signal shape was similar for 75:25 PCL/CH blend to the reference silica surface, however a slight decrease in the frequency was observed after 10 h. This suggests the stronger attachment to the surface while cells lacked further spreading stages on 75:25 PCL/CH blend. 3T3 cells on 25:75 PCL/CH blend showed substantial frequency drop after 10 h, which accompanied by an increase in dissipation. This behavior corresponded to the later stages of cell adhesion, implying that cells probably underwent actin remodeling and fully spreading on the surface. In conclusion, distinct QCM-D signal patterns were evident in the adhesion of hFOB and 3T3 cell lines. These distinctive patterns were attributed to the variations in the strength of cell adhesion, which are influenced by both cell type and surface chemical properties. The real-time and label-free data collected through QCM-D gave us a more profound comprehension of the dynamic adhesion behavior of the cells on constructed thin films. This knowledge and understanding holds the potential to provide valuable insights for the design of novel biomaterials tailored to diverse applications.