Abstract
BACKGROUND:
The adsorption of salivary pellicle proteins onto the material surface is key for denture plaque formation.
OBJECTIVE:
We aimed to investigate the adsorption of bovine serum albumin (BSA) and mucin (MCN) onto denture base metal materials using a 27-MHz quartz crystal microbalance (QCM) method.
METHODS:
A gold (Au), titanium (Ti), and cobalt chromium alloy (Co–Cr) sensors were employed for QCM measurements. Adsorbed amounts of BSA or MCN were calculated by observing the frequency decrease, and the apparent reaction rate, k obs , was obtained by the curve fitting of the frequency shift against the adsorption time.
RESULTS:
The adsorbed amounts of BSA on Ti were significantly lower than those on Au and Co–Cr. For MCN adsorption, Au showed significantly greater amounts of adsorption than Co–Cr. The k obs of Ti for BSA adsorption was significantly smaller than for the Co–Cr. The k obs of Ti, and Co–Cr for MCN adsorption were significantly smaller than for the Au. A clear correlation was not determined between adsorbed amounts of BSA or MCN onto each sensor and the surface topography or contact angles.
CONCLUSIONS:
The difference of denture base metals and the difference of salivary proteins influences the adsorption behavior of salivary proteins.
Introduction
It has been reported that wearing dentures fosters the colonization of oral pathogen bacteria and the formation of biofilm on the denture base materials [1–3]. Biofilm on the denture base materials is called denture plaque. Denture plaque often leads to denture stomatitis and has been associated with aspiration pneumonia [4,5]. Biofilm formation is a multi-stage process [6], the first of which is the attachment or adsorption of salivary pellicle proteins on the surface of the materials. The surface of materials exposed to the oral environment is directly converted by spontaneous adsorption of protein-dominated films, denoted by the acquired pellicle [7,8]. It is suggested that this presence of proteins on the material surface will influence bacterial attachment. The study of the adsorption behavior of salivary proteins onto denture base materials is therefore important for understanding denture plaque formation.
Some metals such as gold (Au), titanium (Ti) and cobalt-chromium (Co–Cr) alloy were compared with polymethyl methacrylate (PMMA) resin as denture base materials. The base metals have several advantages over PMMA, including higher mechanical strengths and higher thermal conductivities. Urushibara et al. conducted a quantitative and qualitative evaluation of the biofilms formed on denture base metals in vitro [9], and found that biofilm formation on each denture metal was different in terms of both quantity and quality, and that some bacteria selectively adhered to each denture metal. However, few studies have investigated the adsorption of salivary proteins on the surface of denture base metals.
There are a number of methods of evaluating protein adsorption onto the materials, such as infrared reflection spectroscopy, ellipsometry, and surface plasmon resonance [10]. Among them, the quartz crystal microbalance (QCM) method is a straightforward technique, which detects the adsorption of protein onto a material’s surface by measuring the difference in the oscillating frequency of the quartz cell [11]. In a QCM system, the decrease in oscillating frequency is related to the amount of protein bound to the crystal surface. The adsorbed amounts of proteins can be estimated by the Sauerbrey equation [12]. Previously, Yoshida et al. investigated the adsorption of salivary proteins onto different materials; that is, gold, silica, and titanium, by using the QCM method, and they identified the differences in the adsorption behaviors of salivary proteins onto each material [13]. It was also found that titanium and stainless steel showed a greater amount of lactoferrin adsorption than zirconia or PMMA [14].
In the present study, we aimed to investigate the adsorption of two salivary glycoproteins onto Au, Ti, and Co–Cr using a QCM method as a pilot study for examining the mechanism of denture plaque formation. As salivary glycoproteins, bovine serum albumin (BSA), a cell adhesion inhibiting protein, and mucin (MCN), the main component of mucous, were evaluated. The null hypothesis tested was that the difference of denture base metals, Au, Ti, and Co–Cr, and that of salivary glycoproteins, BSA and MCN, did not influence the adsorption behaviors to denture base metals.
Methods
QCM device and sensors
A 27-MHz QCM (AFFINIX QNμ, ULVAC, Inc., Kanagawa, Japan) with 550-μL sensor cells was used. The temperature was maintained at 25 ± 1 °C by a temperature control system, and the solution in the cells was stirred while obtaining the measurements. AT-cut quartz crystal was sandwiched by gold electrode on both sides. The surface area of the Au electrode was 4.9 mm2.
Au, Ti, and Co–Cr sensors were employed in this study. The Ti and Co–Cr sensors were prepared by sputter coating onto the Au electrode by using sputtering deposition equipment (CS200, ULVAC, Inc.). The sputtering conditions are listed in Table 1. Each sensor was irradiated with ultraviolet radiation (BioForce Nanosciences Holdings Inc., US) for 20 min before QCM measurement. UV irradiation of 15 mW/cm2 was emitted perpendicularly to the sensors from a distance of 20 mm at 𝜆 = 254 nm.
Sputtering conditions of Ti and Co–Cr sensors
Sputtering conditions of Ti and Co–Cr sensors
An atomic force microscope (AFM; Nanosurf Easyscan 2, Nanosurf, AG, Switzerland) observation identified the surface condition and surface roughness of the Au, Ti, and Co–Cr sensors; the AFM images were captured in air. Tapping mode silicon probes (Tap190Al-G, force contact 48 N/m, Budget sensors, Bulgaria) with resonance frequencies of approximately 190 kHz were used for imaging. The AFM images were obtained for an area of 2 × 2 μm2. As a surface roughness parameter, three-dimensional arithmetic height (Sa) value was obtained. All measurements were performed five times.
Contact angle measurement of Au, Ti and Co–Cr sensors
The contact angles of the Au, Ti, and Co–Cr sensor surfaces with respect to double-distilled water were measured using a contact angle meter (DMe-201, Kyowa Interface Science Co. Ltd., Tokyo, Japan) after ultraviolet irradiation of each sensor. The water drop volume was maintained at 0.5 μL, and three measurements for 10 s each were made for each surface type. Measurements were performed five times at the same room temperature and humidity.
EPMA mapping analysis of Ti and Co–Cr sensors
The Ti and Co–Cr sensor surfaces after ultraviolet irradiation were evaluated by electron probe micro-analysis (EPMA, JXA-8900R, JEOL Ltd., Tokyo, Japan) at an accelerating voltage of 20 kV by detecting the X-ray intensities of Ti–K𝛼, Co–K𝛼, and Cr–K𝛼. The presence of Ti, Co, and Cr was confirmed by elementary mapping.
X-ray photoelectron spectroscopic (XPS) measurements of Co–Cr sensor
The surface elemental analysis of the Co–Cr sensor was performed using an X-ray photoelectron spectroscope (XPS; M-Probe, Seiko Instruments Inc., Chiba, Japan). The peaks for Co2p, Cr2p and O1s were analyzed. The binding energy of each spectrum was calibrated against the C1s peak at 284.8 eV. C1s was the contaminated peak derived from vacuum oil.
QCM measurement of BSA and MCN adsorption
Bovine serum albumin (BSA, MW = 6–7 kDa, Wako Pure Chemical Inds. Ltd., Osaka, Japan) and pig stomach mucin (MCN, MW = 1,000–10,000 kDa, Wako Pure Chemical Industries, Ltd., Osaka, Japan) were dissolved in a phosphate-buffered saline (PBS) solution of pH = 7.4 at a concentration of 1.0 mg/mL. Before QCM measurement, each sensor was cleaned by ultraviolet irradiation as mentioned above.
The procedure for QCM measurement is illustrated in Fig. 1. After mounting the sensor cell, 500 μL of PBS was injected into the cell. After stabilization of the frequency shift, 5 μL of BSA or MCN solution was injected into the PBS solution in the cell. The frequency decrease was monitored until 120 min after the protein injection. The amount of BSA or MCN adsorbed onto each surface (Δm) at 120 min after the injection was calculated by following Sauerbrey’s equation [12].

The procedure for QCM measurements.
By curve fitting for the ΔF curve against the adsorption time, the apparent reaction rate, k
obs
, in the following equation was obtained.
ΔF ∞ is the frequency shift at infinite time. k obs is calculated from the bonding and dissociation rate and is the reciprocal number of the relaxation time. Relaxation time means the time for approximately 63% (=1 − e−1) adsorption of saturated adsorption quantity. Namely, larger relaxation time corresponds to slower adsorption rate. Thus, a larger value of k obs indicates a more rapid rate for adsorption.
Three runs of QCM measurements for BSA or MCN for each sensor were performed during 10 min of injection, respectively.
The data of surface roughness, contact angles, adsorbed amounts, and k obs were evaluated by a one-way analysis of variance (ANOVA) and the Bonferroni test for multiple comparisons among the means at P = 0.05 with SPSS Version 25 Statics (IBM Japan, Ltd., Tokyo, Japan).
Results
Characterization of each sensor
Contact angles and surface roughness (Sa) of each sensor are listed in Table 2. Au and Co–Cr showed significantly greater contact angles than Ti (p < 0.05). There are significant differences in surface roughness among the three sensors (p < 0.05), with Ti having the most significantly rough surface (p < 0.05).
Contact angle and surface roughness of Au, Ti and Co–Cr sensors before salivary protein adsorption (n = 5)
Contact angle and surface roughness of Au, Ti and Co–Cr sensors before salivary protein adsorption (n = 5)
Values in brackets are SD. Same alphabets (a–e) indicate no significantly differences (p > 0.05).
AFM images of each sensor before protein adsorption are shown in Fig. 2. There are no distinct differences in AFM images, and spherical particles with dimeters of 0.1–0.2 μm were observed on each surface. Figure 3 show the elementary distribution of Ti on the Ti sensor and of Co & Cr on the Co–Cr sensor, respectively. Homogeneous sputter coating of Ti, Co, and Cr on each sensor surface was confirmed.

AFM images of QCM sensors before protein adsorption. (a) Au, (b) Ti, (c) Co–Cr.

EPMA analysis of surface of Ti sensors and Co–Cr sensors. (a) Ti: titanium, (b) Co: cobalt, (c) Cr: chromium.
The XPS spectrum for Co–Cr sensor was displayed in Fig. 4. Co–Cr sensor showed O1s peak at about 530 eV besides Co2p and Cr2p. The presence of oxide passive layer could be confirmed on Co–Cr sensor.

XPS spectrum for Co–Cr sensor.
Figure 5 shows the typical ΔF curves for the adsorption of BSA and MCN to each sensor by QCM measurements. The decrease in ΔF immediately after the protein injection could be monitored. The greater degree of the decrease in ΔF corresponds to a greater degree of adsorption of proteins to each sensor. Each sensor showed a different degree of ΔF. The Co–Cr sensor showed the greatest degree of decrease in frequency for BSA adsorption, while Au showed the greatest degree of decrease in frequency for MCN adsorption.

Typical ΔF for the adsorption of BSA and MCN to each sensor by QCM measurements. (a) BSA, (b) MCN.
Figure 6 shows the amounts of BSA and MCN adsorbed on each sensor 120 min after injection, estimated by the Sauerbrey equation [12]. The adsorbed amounts of BSA on Ti were significantly lower than the amounts observed on Au and Co–Cr (p < 0.05). There were no significant differences in the amounts of BSA adsorbed between Au and Co–Cr (p > 0.05). For MCN adsorption, Au showed significantly greater amounts of adsorption than Co–Cr (p < 0.05). There were no significant differences between Au and Ti or Ti and Co–Cr in adsorbed amounts of MCN (p < 0.05).

Estimated amounts of BSA and MCN adsorbed on each sensor at 120 min after injection estimated by the Sauerbrey equation. (a) BSA, (b) MCN.
Table 3 lists the k obs values for the adsorption of the proteins to each sensor during 30 min. A larger value of k obs indicates a more rapid adsorption rate. No significant differences existed between Au and Ti, nor between Au and Co–Cr (p > 0.05) in BSA adsorption. For MCN adsorption, there was no significant difference between Ti and Co–Cr (p < 0.05).
k obs values for the adsorption of BSA or MCN onto each sensor during 30 min of adsorption
Values in brackets are SD. Same alphabets (a–d) indicate no significantly differences (p > 0.05).
In this study, we evaluated the adsorption behavior of BSA and MCN onto Au, Ti, and Co–Cr using a 27-MHz QCM. Au, Ti, and Co–Cr are used as denture base metals. This study revealed that the difference of denture base metals, Au, Ti, and Co–Cr, and the difference of salivary glycoproteins, BSA and MCN, influenced the adsorption behaviors of salivary glycoproteins onto denture base metals. Thus, the null hypothesis was rejected.
In the sputtering process, atoms or molecules are ejected from the target material by the plasma or gas bombardment. Ejected atoms or molecules are deposited on the substrate. In the present study, pure Ti and Co–Cr alloy were used as a target for each sensor. It is suggested that the cohesive force of ejected atoms or molecules, adhesion strength between substrate and deposited film etc will influence the formation of the deposited film. Sputtering conditions and the differences of element species will also influence the distribution of each element. Moreover, the ejected speed and amounts of Ti were different from Co and Cr. We speculated these factors will influence the distribution of Ti, Co and Cr as shown in Fig. 3.
BSA and MCN are both components of human saliva, and are components of acquired pellicle [15]. It is generally recognized that the hydrophilicity and the surface roughness of the materials exposed to the oral environment influences protein adsorption. Hydrophobic interactions play an important role as a driving force in pellicle formation, as higher amounts of salivary proteins adsorb onto hydrophobic surfaces [16]. Generally, a roughened surface enhances protein adsorption onto the material surface [17].
However, the present results were complicated. In this study, Au and Co–Cr were more hydrophobic than Ti, and Ti had a more roughened surface than Co–Cr and Au. Wei et al. observed that BSA showed greater adsorption to hydrophobic surfaces [18]. Present results confirmed that BSA adsorbed to the more hydrophobic Au and Co–Cr. In the series of protein adsorption, first, water molecules contact with the material surface and may bind to surface by hydrogen bonding. Then, proteins reach the hydrated surface [19]. The balance between the binding affinity of proteins to material surface and that of water will control the protein adsorption. More binding affinity of proteins to the material surface than that of water will be needed to replace the water for protein adsorption [20]. Although the details of binding affinity of BSA and mucin is not clear, it is presumed that BSA adsorption is more difficult on the hydrophilic surface due to the less binding affinity of BSA on hydrophilic surface.
Ti possessed the roughest surface, but the adsorbed amounts of BSA and MCN on Ti were less than those on Au. Surface roughness was therefore not the dominant factor influencing adsorption in the present study. The reason of the difference in surface roughness of each sensor was not clear. Sputtering conditions and the differences of element species will influence the distribution of each element on the QCM sensor. Moreover, the ejected speed and amounts of Ti were different from Co and Cr. It is presumed that these factors may influence the difference of surface roughness.
Electrostatic interaction is also another important factor for protein adsorption onto the materials. The isoelectric points (IEP) of BSA and MCN are very similar, at 4.7 and 4.8–5.1, respectively [20,21]. Thus, BSA and MCN were negatively charged under this study’s condition of pH = 7.4. The surfaces of Ti and Co–Cr were oxidized. IEP of TiO2 and Cr2O3 are 5–5.5 and 6.5–7.4, respectively [23–26]. Regarding Au, IEP of Au is reported to be approximately 4.1 [27]. Zeta potential is another relevant electronic surface characteristic of metals. The zeta potential of bulk metals may be measured using streaming potential methods [28]. Zeta potentials of Au, Ti, and Co–Cr determined by streaming potential method are reported to be approximately −20 mV, −87 mV and −26 mV at pH = 7.4, respectively [14,29,30]. These results indicated that the surface of Au, Ti, and Co–Cr are also negatively charged at the present conditions. Thus, electric repulsion occurred between the proteins and the metals. However, distinct differences in adsorption amounts and adsorption rates were recognized between BSA and MCN, although IEP of BSA and MCN are almost the same. It is suggested that other factors besides electrostatic interaction influence the BSA and MCN adsorption.
Yan et al. investigated the adsorption of BSA to cobalt–chromium–molybdenum (CoCrMo) alloy at different pH values and found that maximum adsorption of BSA to CoCrMo alloy occurred at pH = 4.7 which corresponded to the IEP of BSA [31]. They speculated that repulsion of charged BSA molecules and conformation changes in BSA molecules influenced the adsorption behavior. Hermmersan et al. observed the conformational differences in fibronectin, a well-known cell adhesive protein, when adsorbed onto Au and Ti [32]. The adsorption of fibronectin on Au was greater than that of Ti in their study. They discovered that fibronectin is adsorbed in a more compact form on Ti, whereas on Au it is adsorbed in an upright, elongated form. Repulsion of molecules and/or conformational changes of BSA and MCN may explain the complex results for adsorption in this study. More detailed studies testing adsorption at different pH values or with different concentrations of BSA and MCN will elucidate the mechanism of adsorption behaviors.
Our previous study revealed that Ti sensor had an oxidized passive layer on the surface [33]. Present analysis also confirmed the presence of an oxidized passive layer on the Co–Cr sensor surface. In contrast to Ti and Co–Cr, Au had no oxidized passive layer. For MCN adsorption, greater amounts of adsorption and a more rapid adsorption rate onto Au may be due to Au lacking an oxidized layer on its surface. The presence of an oxidized layer on Ti and Co–Cr may have interfered with the adsorption of MCN onto their surfaces, although the reason for this is not clear.
This study revealed that adsorption behaviors of salivary proteins onto denture base metals were different depending on both the kinds of proteins and the denture base metals. The compositions of acquired pellicles on enamel and dentin are different [34]. It is therefore presumed that the formation of bacterial colonies on denture base metals will vary with each metal, and the compositions or components of denture plaque on denture base metals may likewise vary. It is suggested that designated denture detergents be available for cleaning each denture base metal material in dental clinics.
In this study, we used a 27-MHz QCM, which enabled measurements with high sensitivity and low noise [35,36]. Generally, higher fundamental frequency QCM produces higher sensitivity. The present 27-MHz QCM has a very low noise level of ±0.05 Hz and can monitor the protein adsorption onto the materials at the nano-scale. The advantages of the QCM method are that it is simple, and it offers direct in situ detection of the chemical reaction or adsorption behavior. The QCM method was revealed to be a useful tool for monitoring the adsorption behaviors of proteins onto the surfaces of materials.
Conclusion
This study revealed that the adsorption behaviors of salivary protein such as BSA or MCN onto denture base materials were influenced by the difference of denture base metals—gold, titanium, and cobalt-chromium alloy—and the difference of salivary proteins, bovine serum albumin and mucin. The QCM method can directly monitor the adsorption behavior of proteins on the materials surfaces.
Footnotes
Acknowledgements
The authors are grateful to E. Yoshida for his help with QCM measurement. This work was supported by Grants-in-Aid for Young Scientists (B) (JP17K17225) and Grants-in-Aid for Scientific Research (C) (JP17K11815) from the Japan Society for the Promotion of Science.
Conflict of interest
The authors declare that there is no conflict of interest associated with this study.
