Research Article  Open Access
Effects of Frequency and Acceleration Amplitude on Osteoblast Mechanical Vibration Responses: A Finite Element Study
Abstract
Bone cells are deformed according to mechanical stimulation they receive and their mechanical characteristics. However, how osteoblasts are affected by mechanical vibration frequency and acceleration amplitude remains unclear. By developing 3D osteoblast finite element (FE) models, this study investigated the effect of cell shapes on vibration characteristics and effect of acceleration (vibration intensity) on vibrational responses of cultured osteoblasts. Firstly, the developed FE models predicted natural frequencies of osteoblasts within 6.85–48.69 Hz. Then, three different levels of acceleration of base excitation were selected (0.5, 1, and 2 g) to simulate vibrational responses, and acceleration of base excitation was found to have no influence on natural frequencies of osteoblasts. However, vibration response values of displacement, stress, and strain increased with the increase of acceleration. Finally, stress and stress distributions of osteoblast models under 0.5 g acceleration in Zdirection were investigated further. It was revealed that resonance frequencies can be a monotonic function of cell height or bottom area when cell volumes and material properties were assumed as constants. These findings will be useful in understanding how forces are transferred and influence osteoblast mechanical responses during vibrations and in providing guidance for cell culture and external vibration loading in experimental and clinical osteogenesis studies.
1. Introduction
It is widely accepted that bone is a dynamic tissue, because bone remodelling cells (including bone formation cells (osteoblasts) and bone degrading cells (osteoclasts)) can be activated under mechanical stimuli [1]. To analyse the exterior mechanical stimulation received by bone cells and their cellular responses, various mechanical stimuli have been used in in vitro studies since 1970 [2], for example, strain [3], fluid shear stress [4], and vibration [5]. An in vivo investigation of mice subjected to highfrequency mechanical signals suggested that some diseases or metabolic conditions can be inhibited or attenuated by vibrational stimuli, for example, adiposity [6]. Similarly, bone formation at the implantation sites and thus the osseointegration of boneanchored implants can be enhanced by the vibrational stimuli [7, 8]. One in vitro sineshaped vibration experiment with a displacement amplitude of 25 μm and frequencies of 20–60 Hz applied to the cultured osteoblasts revealed that the vibration with an acceleration amplitude of 0.05 g and frequency of 20 Hz was optimal for cell proliferation and that the vibration with 0.13 g and 60 Hz was optimal for metabolic activity [9]. In a later study, a sinusoidal inertia force (at an acceleration amplitude of 0, 0.125 g, 0.25 g, or 0.5 g and frequency of 50 Hz) applied to cultured osteoblasts caused levels of gene expression of alkaline phosphatase (ALP) (a marker of osteogenic differentiation) to increase with the acceleration amplitude [10]. Furthermore, when MLOY4 osteocytes were exposed to lowmagnitude, highfrequency vibration (0.3 g, 30, 60, and 90 Hz, 1 hour), their promoting effect on the osteoclast formation was inhibited [11]. These biomechanical experimental studies clearly illustrate that mechanical stimuli including vibration can affect bone cell formation and activity.
Human skeleton and bone cells are frequently subjected to vibration force experienced through activities or exercise, and the vibration is described often by frequency and by acceleration (acceleration < 1 g as low intensity and acceleration ≥ 1 g as high intensity) [12, 13]. While the acceleration values of human walking, running, and jumping hurdles are 1 g, 34 g, and 5 g, respectively [14], exercise can be considered as a repetitive vibration force with low frequency (~12 Hz) and high intensity [13]. Although the above and some other studies, using either experimental methods, computational methods, or finite element (FE) methods, have suggested that mechanical vibration can affect osteoblast proliferation, differentiation, and metabolic activity [9], as well as bone formation [7, 8], the biomechanical responses and the mechanisms of their responses of osteoblasts are unclear in responses to vibration stimuli of different acceleration and/or frequency.
In the present work, we aimed to investigate the biomechanical responses (displacement, von Mises stress, and strain) of osteoblasts of various shapes to mechanical vibration of different levels of acceleration. The main objectives of this study were fivefold: (1) to develop the idealized continuum FE models of osteoblasts of six different shapes; (2) to obtain the natural frequencies and mode shapes of all osteoblast FE models; (3) to determine the harmonic responses (like displacement and von Mises stress of nucleus centre) to base excitation vibration of osteoblast FE models; (4) to investigate the effect on osteoblast responses of base excitation under three levels of acceleration, that is, 0.5 g, 1 g, and 2 g, respectively; and (5) to investigate the effect factor on resonance frequency.
2. Materials and Methods
2.1. Geometry Information Used for Osteoblast FE Modelling
The shapes of osteoblasts used for the FE modelling in the current study were from the results of experimental investigation [15] and computation modelling [16]. The osteoblast consists of three components, that is, the nucleus, cytoplasm, and cell membrane. In this geometry, the nucleus is embedded in the cytoplasm, the cytoplasm is between the nucleus and the cell membrane, and the cell membrane is the outermost surface of the whole cell (Figures 1 and 2). The fibrous materials (including microfilament and the microtubules) of the cell were neglected/ignored in all modelling in this present study. The geometry shapes and dimensions are shown in Figure 2 and Table 1.

The geometry of the osteoblasts is selected as part of a sphere. The nucleus is modelled as ellipsoid because the nucleus is normally modelled as sphere [17] or ellipsoid [16, 18, 19]. In this study, the volumes of the cell and nucleus are ~3000 μm^{3} and 104.5 μm^{3}, respectively, which are based on the study of McGarry et al. [16]. The thickness of cell membrane is 6 nm [20, 21], and the endothelial cell membrane thickness also varies between 0.1 and 0.5 μm [17]. The cell height is ~2–20 μm [16, 21–23]. The bottom surface of the cell can be circle [16, 22–24] or ellipse [18, 19, 25] for the idealized models.
2.2. 3D Osteoblast FE Modelling
Here, the 3D osteoblast FE models were developed based on the corresponding geometry data. The FE models of different shapes are shown in Figure 1. Using software ABAQUS 6.14 (SIMULIA, Providence, RI, USA), eightnode hexahedral elements (C3D8) were used for the solid regions, that is, the nucleus and cytoplasm. Also, for the nucleus and cytoplasm, the hexahedral FE mesh was mapped according to the high quality geometries using eightnode C3D8 elements. At the same time, the cell membrane was meshed as shell element S4. The total element numbers of the nucleus, cytoplasm, and membrane were given in Table 1, respectively, for the six different models. To prevent any relative movement during subsequent vibration simulation analyses, in the FE modelling, the tie constraints were used to ensure attachment of the cytoplasm to the membrane and nucleus. In this study, we assumed there are no relative motions between cytoplasm and membrane and between cytoplasm and nucleus.
In this study, the materials were assumed isotropic/linear/elastic for the osteoblast FE models; and their properties and density for the osteoblast are given in Table 2. The density ratio 0.4 : 1 : 1.2 (600 kg/m^{3} : 1500 kg/m^{3} : 1800 kg/m^{3}) of membrane, cytoplasm, and nucleus was assumed [26]. Young’s modulus of membrane, cytoplasm, and nucleus was chosen at 1 kPa, 1.5 kPa, and 6 kPa, respectively, and Poisson’s ratio was 0.3, 0.37, and 0.37 for membrane, cytoplasm, and nucleus, respectively.

For endothelial cells, Young’s modulus of membrane, cytoplasm, and nucleus was set at 775 Pa, 775 Pa, and 5.1 kPa, respectively. Poisson’s ratio of membrane, cytoplasm, and nucleus was set at 0.33 [27]. Normally, the elastic modulus of the cytoplasm is only onequarter that of the nucleus [28]. It was reported that Young’s modulus of the cytoplasm and nucleus was chosen at 100 Pa and 400 Pa, respectively, and Poisson’s ratio was 0.37 for the cytoplasm and nucleus [16, 19, 29]. Previously, Young’s modulus of 6.5 kPa and Poisson’s ratio of 0.5 were assigned to the cytoplasm of an osteoblastic cell [22]. Elastic modulus of cytoplasm of osteocytelike MLOY4 cell was set at 1.5 kPa [26]. Similarly, based on the measurements by atomic force microscopy, Young’s modulus of 6 kPa was assigned to the osteoblast nucleus [30]. The elastic modulus of 1 kPa and Poisson’s ratio of 0.3 were selected for the membrane of the adherent eukaryotic cell [16, 26].
Density is an important parameter in vibration simulation. The initial cellular density was assumed as 1000 kg/m^{3} [31], and 1250 kg/m^{3} was used as the density of cytoplasm, nucleus, and membrane in the endothelial cell [27]. Previously, the density of an osteoblast was assumed as 125 kg/m^{3} [19], and, for osteocytelike MLOY4 cell, 1500 kg/m^{3}, 1800 kg/m^{3}, and 600 kg/m^{3} were set as the densities of cytoplasm, nucleus, and membrane, respectively [26].
2.3. Modal Analyses
2.3.1. Natural Frequency Extraction
Natural frequency extraction is an eigenvalue analysis procedure, which determines the natural frequencies and shapes of mode for a structure. In this study, software ABAQUS was used to conduct the natural frequency extraction. The governing dynamic equation of the response in ABAQUS can be expressed as follows [32]:where , , and (symmetric and positive definite) are the mass matrix, damping coefficient matrix, and spring stiffness matrix in the system, respectively. is harmonic loading, and , , and are the acceleration vector, velocity vector, and displacement vector, respectively.
The free vibration structure without damping may be represented as
The solution of may be expressed as
Then, (2) may be rewritten aswhere is the frequency and is the eigenvector (vibration mode).
Here, the natural frequencies of the six osteoblast models (Model I, Model II, Model III, Model IV, Model V, and Model VI) were obtained through the natural frequency extraction, and the corresponding mode shapes of the six osteoblast models were presented by the FE analysis.
2.3.2. Harmonic Vibration
In this study, the free vibration of the system was considered with only one degree of freedom. In the simulation, the different levels of acceleration of base excitation were applied to analyse the effect of acceleration on the cell, that is, 0.5 g, 1 g, and 2 g (g = 9.8 m/s^{2}), respectively. It is well known that acceleration (g forces, g = 9.8 m/s^{2}) is the best term to describe vibration intensity [12]. In addition, the translational directions of bottom surface of FE models were constrained, which means that zero displacement was applied for the bottom surface of the cell model due to the fixed boundary condition.
2.3.3. Effect Factor on Resonance Frequency
A previous human vibration test was conducted under the acceleration of 0.04–19.3 g, and it was found that the resonant frequency values of ankle, knee, hip, and spine are 10–40 Hz, 10–25 Hz, 10–20 Hz, and 10 Hz, respectively [33]. In this study, the effect of geometry on resonance frequency was investigated. The relationship between resonance frequency and the effect factor can be expressed through a fitted formula. The relationships of resonance frequency with cell height and with bottom area were analysed when the volume and density of the cell were assumed as constants. Cell height or cell bottom area can be considered as the effect factor, and the relationships of resonance frequency with cell height or with cell bottom area can be expressed as the formula based on the data of Figures 12(a) and 12(b).
3. Results
3.1. Natural Frequency Extraction
The natural frequencies and vibration mode shapes of the six osteoblast models were obtained after completing the FE modal analyses. Figure 3 gives the resonant frequencies of the first ten modes for the six models. The first natural frequency values of osteoblasts for Model I–Model VI were 6.85, 11.35, 17.42, 21.40, 24.96, and 28.57 Hz, respectively. The corresponding mode shapes of Model VI are presented in Figure 4, and the natural frequencies for the first ten modes of Model VI were found at 28.57 Hz, 28.61 Hz, 33.79 Hz, 41.13 Hz, 41.22 Hz, 42.60 Hz, 42.71 Hz, 43.53 Hz, 48.68 Hz, and 48.69 Hz, respectively.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
(j)
3.2. Responses to Harmonic Vibration
3.2.1. The Displacement Response
Based on the natural frequencies and the assumed uniform acceleration of base excitation within the frequency range between 1 and 50 Hz, the harmonic responses of the six FE models were computed. To investigate the responses of the different directions (, , and directions) to the harmonic vibration, the displacement values at the centre of the nucleus (Figure 1) were plotted in Figure 5 for the different levels of acceleration (0.5 g, 1 g, and 2 g). The frequency values of the peak displacement in direction for Model I–Model VI are 6.85 Hz, 11.35 Hz, 17.42 Hz, 21.40 Hz, 24.96 Hz, and 28.57 Hz, respectively. The frequency values of the peak displacement in direction for Model IModel VI are 6.88 Hz, 11.38 Hz, 17.46 Hz, 21.44 Hz, 25.00 Hz, and 28.61 Hz, respectively. The frequency values of the peak displacement in direction for Model I–Model VI are 17.69 Hz, 23.38 Hz, 30.59 Hz, 34.97 Hz, 38.99 Hz, and 43.51 Hz, respectively.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
The mode shapes of the FE models at the peak frequency under 0.5 g acceleration in direction, direction, and direction are presented in Figures 6(a), 6(b), and 6(c), respectively. In addition, the displacement values of the models under the different levels of base excitation acceleration in direction, direction, and direction are given in Figure 7. The order of displacement of the centre of the nucleus for the different acceleration in direction, direction, and direction is 2 g > 1 g > 0.5 g. The order of displacement of the centre of the nucleus for the different models in direction, direction, and direction is Model I > Model II > Model III > Model IV > Model V > Model VI. The lowest displacement values of the centre of the nucleus for Model VI under 0.5 g acceleration in direction, direction, and direction are 0.565 μm, 0.565 μm, and 0.146 μm, respectively. The largest displacement values of the centre of the nucleus under 2 g acceleration in , , and direction are 12.749 μm, 12.756 μm, and 2.188 μm, respectively.
(a)
(b)
(c)
(a)
(b)
(c)
3.2.2. The von Mises Stress and the Strain
In addition, the von Mises stress values of the centre of the nucleus under the different levels of base excitation in , , and directions are given in Figure 8. The lowest von Mises stress values of Model I under 0.5 g in , , and directions are 1 for all of them. The order of von Mises stress values of the centre of the nucleus for the different base excitation in , , and directions is 2 g > 1 g > 0.5 g. The order of von Mises stress values of the centre of the nucleus for the different models in , , and directions is Model I > Model II > Model III > Model IV > Model V > Model VI. The lowest von Mises stress values of the centre of the nucleus for Model VI under 0.5 g in , , and directions are 0.145 kPa, 0.16 kPa, and 0.094 kPa, respectively. The largest von Mises stress values of the centre of the nucleus for Model I under 2 g in , , and directions are 3.05 kPa, 3.33 kPa, and 1.57 kPa, respectively.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
Moreover, the von Mises stress contours of the FE models under 0.5 g base excitation in direction at the frequency of 1 Hz, peak value, and 80 Hz are shown in Figure 9. The peak frequency values in direction for Model I–Model VI are 17.69 Hz, 23.38 Hz, 30.59 Hz, 34.97 Hz, 38.99 Hz, and 43.51 Hz, respectively. The maximum von Mises stress values of the cell at the peak frequency for Model I–Model VI under base excitation in direction are 590.4 Pa, 327 Pa, 197.7 Pa, 174.7 Pa, 158.9 Pa, and 129.5 Pa, respectively. At 80 Hz, the maximum von Mises stress values are 1.229 Pa, 1.255 Pa, 1.363 Pa, 1.675 Pa, 1.99 Pa, and 2.175 Pa, respectively.
(a)
(b)
(c)
(d)
(e)
(f)
Similarly, the strain values of the centre of the nucleus under the different base excitation in , , and directions are given in Figure 10. The lowest strain values for Model I under 0.5 g in direction, direction, and direction are 1 for all of them. The order of strain values of the centre of the nucleus for the different base excitation in direction, direction, and direction is 2 g > 1 g > 0.5 g. The order of strain values of the centre of the nucleus for the different models in , , and directions is Model I > Model II > Model III > Model IV > Model V > Model VI. The lowest strain values of the centre of the nucleus for Model VI under 0.5 g in , , and directions are 19095.4, 21096.5, and 14827.6 microstrains, respectively. The largest strain values of the centre of the nucleus for Model I under 2 g in , , and directions are 401550, 438575, and 252340 microstrains, respectively.
(a)
(b)
(c)
(d)
(e)
(f)
(g)
(h)
(i)
Furthermore, the strain contours of the 6 FE models under 0.5 g base excitation in direction at the frequency of 1 Hz, peak value, and 80 Hz are given in Figure 11. The maximum strain values of the cell for Model I–Model VI in direction at the peak frequency are 327300, 225700, 101200, 90750, 83890, and 67010 microstrains, respectively. There is a fluctuation of the maximum strain value of the cell at the peak frequency. At 80 Hz, the maximum strain values are 680.3, 866, 689.7, 860.1, 1056, and 1128 microstrains, respectively.
(a)
(b)
(c)
(d)
(e)
(f)
3.3. Effects of Cell Height and Bottom Area on Resonance Frequency
The relationships of resonance frequency with cell height and with bottom area were analysed when the volume and density of the cell were assumed as constants. The results are shown in Figures 12(a) and 12(b), respectively. Moreover, the relationship between frequency and different models was also analysed and is shown in Figure 12(c). These results suggest that the resonance frequency depends on cell height and bottom area. If (cell height) or (bottom area) (Figure 2) is an independent variable, the value (resonance frequency) is the function of (cell height) or (bottom area) and can be expressed through a fitted formula in (5) or (6) based on the data of Figures 12(a) and 12(b) as follows:where is the cell height and the values of are coefficients that are given in Table 3. where is the bottom area and the values of are coefficients that are given in Table 4.


4. Discussion
There have been many previous studies using various models that were developed to reveal bone cell mechanical behaviours. Some numerical computational models of cultured cells have been developed to examine the responses of cultured cells to various mechanical stimulations. For example, to study the universal dynamic behaviours of osteoblasts, a threedimensional (3D) soft matter cell model was developed using the multiscale moving contact line theory [31, 34]. On the other hand, the finite element (FE) method has been more widely used to analyse the biomechanical behaviours of osteoblasts or other cell types, using the biorealistic or idealized cell models. Based on confocal microscopy data, a 3D cellspecific FE model was created to simulate the cellular mechanics tests such as large deformation [35]. For the idealized FE models, both the continuum model [17, 23, 36] and the tensegrity model [16, 24, 29, 37] have been developed. Although the above prior studies have made significant advancements in revealing bone cell mechanical behaviours, it still remains a great challenge to understand the mechanisms of biomechanical behaviours of osteoblasts in response to vibration signals and their characteristics of deformations under various mechanical stimuli.
In this present study, to investigate the vibrational responses of the different shapes of an osteoblast subjected to vibration of base excitation, six idealized FE models of an osteoblast were created. Firstly, cell geometry was created and FE models were developed accordingly. For these models, the initial volumes of the cell were basically the same and the densities were constant in the simulation. Secondly, natural frequency (resonance frequency) of the different models was extracted by using FE analysis. Then, harmonic vibration of the osteoblast models was analysed with three different base excitation acceleration values, namely, 0.5 g, 1 g, and 2 g. The response results were obtained for the harmonic vibration including displacement, von Mises stress, and strain of the centre of the nucleus under different acceleration values. Finally, the effects of cell height and bottom area on resonance frequency were analysed, and the fitted curves of resonance frequency versus cell height and resonance frequency versus bottom area were obtained.
Based on the previous studies, the vibration frequency is crucial for bone cells to complete the bone resorption and bone formation [1, 11]. Using the FE cell models developed, the natural frequency of osteoblasts was predicted in the current study, with the first ten resonant frequencies of the cells being in the range ~6.85–48.69 Hz. Previously, the vibration frequency was selected as 5–100 Hz for in vitro bone cell studies [5, 9]. Similar natural frequency values were computed by other FE studies; for example, the first ten natural frequency values of bone cells were predicted at ~9.95–211.05 Hz for the first ten modes [18] and at ~18.11–21.05 Hz for the first five modes [19]. In comparison with the experimental data and FE studies, the natural frequency of our FE models was within the range in the literatures; the difference may be caused by some factors in FE modelling, for example, the density, material property, and spreading shape. Thus, the FE models developed in the current study can be validated by the data of the literatures.
In this study, the osteoblast models were assumed as onedegreeoffreedom vibrational system. The resonance phenomenon can occur at some natural frequency and can be observed from vibrational responses like displacement, von Mises stress, and strain. The cell biomechanical responses (e.g., displacement, von Mises stress, and strain of the centre of the nucleus models of osteoblasts) were obtained when the models were subjected to three different acceleration values of base excitation vibration (0.5 g, 1 g, and 2 g). The current study observed that the resonance frequency does not change with the acceleration, which suggests that the natural frequency of the bone cell is determined by the intrinsic factors and is not affected by the external factors. The current study has also examined the vibration responses of different FE models of the bone cell under 0.5 g acceleration in , , and directions. It was found that the peak response frequency is the same with the first mode frequency for the acceleration in direction, and the peak response frequency is around at the second mode frequency for the acceleration in direction. While the peak response frequency varies with the models for the acceleration in direction, the peak frequency occurs round at the fourth mode frequency for Model I to Model IV, sixth mode frequency for Model V, and eighth mode frequency for Model VI. A similar resonance phenomenon was also found by one previous study [19]; it found that the resonance of the continuum model of bone cell occurred at mode 1, mode 2, and mode 3 for direction, direction, and direction, respectively. Furthermore, it can be seen from the curves of Figures 6, 9, and 11 that the response peak values at that frequency are remarkably larger than those at the rest. Thus, this phenomenon is typical resonance.
The current study has investigated the displacement, von Mises stress, and strain responses of bone cell models under different acceleration values and in different directions. The results showed that the values of displacement, von Mises stress, and strain increased with the acceleration. The values in direction are slightly larger than those in direction, and the values in direction are smaller than those in  and directions. Figure 7 presents the displacement response of the different models in , , and directions under different acceleration values. Figure 7 further indicates the effect of the acceleration on the displacement response of bone cell for the external vibration.
Our FE models also indicate that von Mises stress is concentrated in the nucleus and strain is basically concentrated around the nucleus for the harmonic response under 0.5 g acceleration in direction. By comparing Model I and Model VI at the peak frequency, the maximum von Mises stress value changed from 590.4 Pa (Model I) to 129.5 Pa (Model VI), and the maximum strain value changed from 327300 microstrains (Model I) to 67010 microstrains (Model VI). Therefore, the maximum von Mises stress and the maximum strain values reduce by 78.07% and 78.92%, respectively. They reduce with the decrease of cell height or with the increase of bottom area, and thus the cell shape has a great influence on the harmonic response.
Based on the simulation results analysis, there is a relationship between resonance frequency and cell height or between resonance frequency and bottom area. The resonance frequency can be expressed as the function of cell height (h) or bottom area (s) (see (5)). The resonance frequency is a monotonic decreasing function of cell height (h) and is a monotonic increasing function of bottom area (s). In other words, the cell volume and material properties do not vary and h or s is the independent variable, and the resonance frequency will decrease with the increase of h or increase with the increase of s. The resonance frequencies in direction are much higher than those in  and direction, and the resonance frequencies in direction and direction are very close. That is because the bottom areas are assumed as a circle, and the differences of resonance frequencies between  and direction would be larger if the bottom areas are not circle but an ellipse.
It must be noted that, in this study, the geometry of the bone cell was assumed to be comprised of three components, that is, membrane, cytoplasm, and nucleus. In the previous studies, the tensegrity structure was used to simulate the fibres of cells like microtubules and microfilaments [29, 37]. In those studies, the nucleus was in the centre of the cell, and the nucleus and the cell membrane were connected by the tensegrity structure (microtubules and microfilaments). However, the effect of microtubules and microfilaments on the harmonic response was not large [19]. For the present FE models, membrane, cytoplasm, and nucleus were assumed to be the linear elastic homogeneous materials. Although the cell has viscoelastic characteristics, which can lead to the damping effect, the viscoelastic material property was ignored in all simulations in this study as the effects of damping on the osteoblast were found to be small [19]. Furthermore, it is well accepted that the stiffness ratio of cytoplasm and nucleus of the cell is 1 : 4 and the elastic modulus of cell membrane is 1 kPa. Further studies using the nonlinear material property and biorealistic geometries are being considered in further investigations.
5. Conclusion
In the current study, different osteoblast FE models were developed and were used to extract the natural frequencies and to analyse the harmonic responses under different acceleration values (0.5 g, 1 g, and 2 g). It was found that the natural frequencies do not change with the variation of acceleration of base excitation. The response values of displacement, von Mises stress, and strain increase with the increase of acceleration, and the response values in direction are much higher than those in the other directions ( and direction). Moreover, the model simulation predicted that the von Mises stress is concentrated in the nucleus and strain is basically concentrated around the nucleus. This study also found that the resonance frequencies can be a monotonic function of cell height or bottom area when the cell volume and material properties are assumed as constants. Therefore, the cell shape has a great influence on the vibrational characteristics of the osteoblast. Therefore, the FE simulations of osteoblast harmonic vibration presented in this study have provided adequate description of vibrational responses of osteoblasts in vitro. These findings could be helpful for guiding the in vitro cell culture biomechanical research and will help in understanding the deformation of osteoblasts under various mechanical stimuli.
Competing Interests
The authors declare that they have no competing interests.
Authors’ Contributions
Liping Wang and Cory J. Xian were responsible for study design. Cory J. Xian conducted the study. Data collection was the responsibility of Liping Wang. Data analysis was carried out by Liping Wang and HungYao Hsu. Data interpretation was done by Liping Wang and HungYao Hsu. Liping Wang and Cory J. Xian drafted the manuscript. And Liping Wang, Xu Li, and Cory J. Xian revised the manuscript content. All authors have read and approved the final submitted manuscript.
Acknowledgments
Liping Wang is supported by the Australian National Health and Medical Research Council (NHMRC) Postgraduate Research Scholarship grant, and Cory J. Xian is supported by the NHMRC Senior Research Fellowship. This work was supported by the National Natural Science Foundation of China (Project Grant no. 81671928).
References
 L. Tirkkonen, H. Halonen, J. Hyttinen et al., “The effects of vibration loading on adipose stem cell number, viability and differentiation towards boneforming cells,” Journal of the Royal Society Interface, vol. 8, no. 65, pp. 1736–1747, 2011. View at: Publisher Site  Google Scholar
 E. A. Evans, “New membrane concept applied to the analysis of fluid shear and micropipette deformed red blood cells,” Biophysical Journal, vol. 13, no. 9, pp. 941–954, 1973. View at: Publisher Site  Google Scholar
 L. Wang, J. Dong, and C. J. Xian, “Strain amplification analysis of an osteocyte under static and cyclic loading: a finite element study,” BioMed Research International, vol. 2015, Article ID 376474, 14 pages, 2015. View at: Publisher Site  Google Scholar
 S. Kapur, D. J. Baylink, and K.H. W. Lau, “Fluid flow shear stress stimulates human osteoblast proliferation and differentiation through multiple interacting and competing signal transduction pathways,” Bone, vol. 32, no. 3, pp. 241–251, 2003. View at: Publisher Site  Google Scholar
 R. G. Bacabac, T. H. Smit, J. J. W. A. Van Loon, B. Z. Doulabi, M. Helder, and J. KleinNulend, “Bone cell responses to highfrequency vibration stress: does the nucleus oscillate within the cytoplasm?” The FASEB Journal, vol. 20, no. 7, pp. 858–864, 2006. View at: Publisher Site  Google Scholar
 C. T. Rubin, E. Capilla, Y. K. Luu et al., “Adipogenesis is inhibited by brief, daily exposure to highfrequency, extremely lowmagnitude mechanical signals,” Proceedings of the National Academy of Sciences of the United States of America, vol. 104, no. 45, pp. 17879–17884, 2007. View at: Publisher Site  Google Scholar
 A. Torcasio, G. H. Van Lenthe, and H. Van Oosterwyck, “The importance of loading frequency, rate and vibration for enhancing bone adaptation and implant osseointegration,” European Cells and Materials, vol. 16, pp. 56–68, 2008. View at: Google Scholar
 E. De Smet, S. V. Jaecques, J. J. Jansen, F. Walboomers, J. Vander Sloten, and I. E. Naert, “Effect of strain at lowfrequency loading on periimplant bone (re)modelling: a guineapig experimental study,” Clinical Oral Implants Research, vol. 19, no. 8, pp. 733–739, 2008. View at: Publisher Site  Google Scholar
 N. Rosenberg, M. Levy, and M. Francis, “Experimental model for stimulation of cultured human osteoblastlike cells by high frequency vibration,” Cytotechnology, vol. 39, no. 3, pp. 125–130, 2002. View at: Publisher Site  Google Scholar
 T. Shikata, T. Shiraishi, S. Morishita, R. Takeuchi, and T. Saito, “Effects of amplitude and frequency of mechanical vibration stimulation on cultured osteoblasts,” Journal of System Design and Dynamics, vol. 2, no. 1, pp. 382–388, 2008. View at: Publisher Site  Google Scholar
 E. Lau, S. AlDujaili, A. Guenther, D. Liu, L. Wang, and L. You, “Effect of lowmagnitude, highfrequency vibration on osteocytes in the regulation of osteoclasts,” Bone, vol. 46, no. 6, pp. 1508–1515, 2010. View at: Publisher Site  Google Scholar
 B. R. Beck, “Vibration therapy to prevent bone loss and falls: mechanisms and efficacy,” Current Osteoporosis Reports, vol. 13, no. 6, pp. 381–389, 2015. View at: Publisher Site  Google Scholar
 M. P. Nagaraja and H. Jo, “The role of mechanical stimulation in recovery of bone losshigh versus low magnitude and frequency of force,” Life, vol. 4, no. 2, pp. 117–130, 2014. View at: Publisher Site  Google Scholar
 R. Marcus, “Exercise: moving in the right direction,” Journal of Bone and Mineral Research, vol. 13, no. 12, pp. 1793–1796, 1998. View at: Publisher Site  Google Scholar
 T. Frisch and O. Thoumine, “Predicting the kinetics of cell spreading,” Journal of Biomechanics, vol. 35, no. 8, pp. 1137–1141, 2002. View at: Publisher Site  Google Scholar
 J. G. McGarry, P. J. Prendergast, B. Ashton, and J. KleinNulend, “A threedimensional finite element model of an adherent eukaryotic cell,” European Cells and Materials, vol. 7, pp. 27–34, 2004. View at: Google Scholar
 S. Deguchi, H. Fukamachi, K. Hashimoto, K. Iio, and K. Tsujioka, “Measurement and finite element modeling of the force balance in the vertical section of adhering vascular endothelial cells,” Journal of the Mechanical Behavior of Biomedical Materials, vol. 2, no. 2, pp. 173–185, 2009. View at: Publisher Site  Google Scholar
 H. Wee and A. Voloshin, “Modal analysis of a spreading osteoblast cell in culturing,” in Proceedings of the 38th Annual Northeast Bioengineering Conference (NEBEC '12), pp. 167–168, March 2012. View at: Publisher Site  Google Scholar
 H. Wee and A. Voloshin, “Dynamic analysis of a spread cell using finite element method,” in Mechanics of Biological Systems and Materials, vol. 4, pp. 135–140, Springer, Berlin, Germany, 2014. View at: Google Scholar
 N. Abolfathi, G. Karami, and M. Ziejewski, “Biomechanical cell modelling under impact loading,” International Journal of Modelling and Simulation, vol. 28, no. 4, pp. 470–476, 2008. View at: Google Scholar
 R. Kamm, A. McVittie, and M. Bathe, On the Role of Continuum Models in Mechanobiology, ASME Applied Mechanics DivisionPublicationsAmd, 2000.
 J. S. Milner, M. W. Grol, K. L. Beaucage, S. J. Dixon, and D. W. Holdsworth, “Finiteelement modeling of viscoelastic cells during highfrequency cyclic strain,” Journal of Functional Biomaterials, vol. 3, no. 1, pp. 209–224, 2012. View at: Publisher Site  Google Scholar
 S. Vichare, M. M. Inamdar, and S. Sen, “Influence of cell spreading and contractility on stiffness measurements using AFM,” Soft Matter, vol. 8, no. 40, pp. 10464–10471, 2012. View at: Publisher Site  Google Scholar
 J. Bursa and V. Fuis, “Finite element simulation of mechanical tests of individual cells,” in World Congress on Medical Physics and Biomedical Engineering, September 7–12, 2009, Munich, Germany: Vol. 25/10 Biomaterials, Cellular and Tussue Engineering, Artificial Organs, vol. 25/10 of IFMBE Proceedings, pp. 16–19, Springer, Berlin, Germany, 2009. View at: Publisher Site  Google Scholar
 W. Ronan, V. S. Deshpande, R. M. McMeeking, and J. P. McGarry, “Numerical investigation of the active role of the actin cytoskeleton in the compression resistance of cells,” Journal of the Mechanical Behavior of Biomedical Materials, vol. 14, pp. 143–157, 2012. View at: Publisher Site  Google Scholar
 G. Uzer, S. Pongkitwitoon, C. Ian et al., “Gap junctional communication in osteocytes is amplified by low intensity vibrations in vitro,” PLoS ONE, vol. 9, no. 3, Article ID e90840, 2014. View at: Publisher Site  Google Scholar
 M. C. Ferko, A. Bhatnagar, M. B. Garcia, and P. J. Butler, “Finiteelement stress analysis of a multicomponent model of sheared and focallyadhered endothelial cells,” Annals of Biomedical Engineering, vol. 35, no. 2, pp. 208–223, 2007. View at: Publisher Site  Google Scholar
 F. Guilak, J. R. Tedrow, and R. Burgkart, “Viscoelastic properties of the cell nucleus,” Biochemical and Biophysical Research Communications, vol. 269, no. 3, pp. 781–786, 2000. View at: Publisher Site  Google Scholar
 J. G. McGarry, J. KleinNulend, M. G. Mullender, and P. J. Prendergast, “A comparison of strain and fluid shear stress in stimulating bone cell responses—a computational and experimental study,” The FASEB Journal, vol. 19, no. 3, pp. 482–484, 2005. View at: Publisher Site  Google Scholar
 E. M. Darling, M. Topel, S. Zauscher, T. P. Vail, and F. Guilak, “Viscoelastic properties of human mesenchymallyderived stem cells and primary osteoblasts, chondrocytes, and adipocytes,” Journal of Biomechanics, vol. 41, no. 2, pp. 454–464, 2008. View at: Publisher Site  Google Scholar
 X. Zeng and S. Li, “A three dimensional soft matter cell model for mechanotransduction,” Soft Matter, vol. 8, no. 21, pp. 5765–5776, 2012. View at: Publisher Site  Google Scholar
 V. Abaqus, 6.14 Documentation, Dassault Systemes Simulia Corp, Providence, RI, USA, 2014.
 J. Kiiski, A. Heinonen, T. L. Järvinen, P. Kannus, and H. Sievänen, “Transmission of vertical whole body vibration to the human body,” Journal of Bone and Mineral Research, vol. 23, no. 8, pp. 1318–1325, 2008. View at: Publisher Site  Google Scholar
 H. Fan and S. Li, “Modeling universal dynamics of cell spreading on elastic substrates,” Biomechanics and Modeling in Mechanobiology, vol. 14, no. 6, pp. 1265–1280, 2015. View at: Publisher Site  Google Scholar
 N. Slomka and A. Gefen, “Confocal microscopybased threedimensional cellspecific modeling for large deformation analyses in cellular mechanics,” Journal of Biomechanics, vol. 43, no. 9, pp. 1806–1816, 2010. View at: Publisher Site  Google Scholar
 R. Katzengold, N. Shoham, D. Benayahu, and A. Gefen, “Simulating single cell experiments in mechanical testing of adipocytes,” Biomechanics and Modeling in Mechanobiology, vol. 14, no. 3, pp. 537–547, 2015. View at: Publisher Site  Google Scholar
 D. Kardas, U. Nackenhorst, and D. Balzani, “Computational model for the cellmechanical response of the osteocyte cytoskeleton based on selfstabilizing tensegrity structures,” Biomechanics and Modeling in Mechanobiology, vol. 12, no. 1, pp. 167–183, 2013. View at: Publisher Site  Google Scholar
Copyright
Copyright © 2016 Liping Wang et al. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.