Osteogenic cells differentiation on topological surfaces under ultrasound stimulation
- 186 Downloads
The current trends in bone tissue engineering aim to fasten the cells osteogenic differentiation by mechanical stimulation. To date, several approaches have proved efficient for this purpose. One is related to changing the shape of the cells nuclei using topological surfaces with appropriate dimensions and stiffness. Another successful method is by low-intensity pulsed ultrasound stimulation (LIPUS) of the cells. The goal of this proof-of-concept study is to introduce and validate, for the first time, the synergistic effect of topological surfaces and LIPUS for improving the osteogenic differentiation of osteoblast-like cells. Cells were grown on topological surfaces consisting of vertical microtubes fabricated by laser direct writing. The flexibility of the topological surfaces was tuned by varying the microtubes’ height. The spatial arrangement and dimensions of the microtubes limited the cell–cell interactions and allowed us to observe individual cells. A finite element model simulation was proposed for explaining the cell–surface interaction details. We monitored the cells nuclei deformations in response to the topological surfaces in conjunction with LIPUS. The topological surfaces alone induced dramatic changes of the shape of the cells nuclei that wrapped around the microtubes. The nuclei deformation was further increased by LIPUS. This synergy between the topological surfaces and LIPUS allowed us to obtain an increase of up to 200% in the cells osteogenic differentiation, as determined by ALP activity and osteocalcin secretion measurements, in comparison with flat surfaces in static regime. A causal relationship between the nuclei deformation and the cells osteogenic differentiation was established.
In present, the major challenge in tissue engineering is to fasten the process of tissue repair by stimulating the cells with stimuli of appropriate type, intensity and duration that improve the cellular metabolism and phenotypic adaptation . An efficient approach relies on the fact that the cells sense the mechanical influences from their surroundings and convert biophysical and biochemical cues into intracellular signals [2, 3, 4, 5, 6]. Micro- and nanostructured surfaces, mostly in the form of vertically aligned pillar structures, have been developed for mechanical stimulation of various cell types [7, 8, 9, 10, 11]. Another efficient mechanical stimulation method is by low-intensity ultrasound stimulation (LIPUS) of the cells, a non-invasive and easy-to-handle approach approved in 1994 by the Food and Drug Administration for the treatment of bone fractures [12, 13].
The nucleus is the largest and stiffest mechanosensitive organelle in a cell, with important roles in regulating the cell mechanics. External forces from the surroundings are transmitted from the adhesion molecules dispersed at the cell surface to the cell nucleus and induce changes of the nucleus structure and functions, which further influence the sensitivity of the whole cell to mechanical forces . To date, several attempts have been made to fabricate patterned cell culture surfaces, with high spatial accuracy and reproducibility, able to deform mechanically the cells nuclei [2, 11]. Nuclei deformation triggered by topological surfaces with micropillar arrays of appropriate dimensions and stiffness showed great potential for tissue regeneration [2, 10, 11]. A particular interest is focused to the development of engineered bone tissue as better alternative to the conventional use of bone grafts . Within this context, the ability of topological surfaces to trigger biochemical signaling that improve the cells osteogenic differentiation emerges as a promising approach.
In this study, we developed flexible topological surfaces to enhance the osteogenesis by the deformation of the cells nuclei in conjunction with ultrasound stimulation. The topological surfaces were fabricated by laser direct writing via two-photon polymerization (LDW via TPP) of IP-L780 biocompatible photopolymer, in the form of arrays of vertical microtubes. These topological surfaces allowed us to explore single-cell behaviors at array interfaces. We monitored how the cells interact with microtubes of varying heights, and we particularly focused on the nuclei deformation induced by these structures. We then correlated the experimental results with numerical simulations of the mechanical deformations induced by the cell–surface interaction. We further amplified the cells nuclei deformation by low-intensity ultrasound stimulation (LIPUS), which is a non-invasive and easy-to-handle approach approved in 1994 by the Food and Drug Administration for the treatment of bone fractures [12, 13].
Finally, we evaluated if the changes of the shape of the cells nuclei influence the cells osteogenic differentiation, by in vitro assays of ALP activity and osteocalcin secretion. For the first time to our knowledge, we demonstrate that the topological surfaces and LIPUS treatment exert a synergic action on the changes of the cells nuclei shape, which significantly improves the cells osteogenic differentiation.
Materials and methods
Fabrication of the topological surfaces
The topological surfaces consisted of vertical microtubes arrays fabricated by laser direct writing via two-photon polymerization (LDW via TPP) of IP-L 780 photopolymer (Nanoscribe GmbH) [14, 15, 16]. LDW via TPP is a method for multiphoton laser direct writing with resolutions below the diffraction limit. The IP-L780 is a biocompatible UV-curable liquid solution optimized for TPP technology. The irradiation was performed using Nanoscribe Photonic Professional. The laser source was a fiber laser that delivers 120 fs pulses at a frequency of 80 MHz, with a central wavelength at 780 nm. The photopolymer has high transparency for the incident wavelength. Polymerization was initiated for the second harmonic, λ = 395 nm, obtained through two-photon absorption. We used an average laser power of 30 mW for all samples. This results in a voxel of ~ 2 µm diameter and 4 µm height. The focusing was realized using an inverted microscope, equipped with a 63 × objective. Sample positioning and processing were achieved using a hybrid system comprised of a set of high accuracy translation stages, completed with a set of piezoelectric stages. The microtubes had inner diameter of 5 µm and various heights (5, 10, 20 µm). They were distributed in a hexagonal lattice, with a center-to-center distance of 10 µm between neighboring tubes. The structures were designed to be fabricated in a spiral-like fashion, in order to optimize the fabrication time, as well as obtain homogeneously polymerized walls. The distance between two consecutive spirals, in the Z-axis direction, was set to 1 µm. Thus, considering the voxel height (~ 4 µm), this results in an overlap of 3 consecutive spirals. This contributes to the degree of polymerization and structural integrity at mesoscopic level. The samples were fabricated using a scanning velocity of 100 µm/s. After the laser writing process, the samples were developed by immersion in propylene glycol methyl ether acetate (PGMEA) for 15 min. After the samples were pulled out of the solvent, they were let to dry at room temperature.
The topological surfaces were investigated by scanning electron microscopy (SEM, FEI InspectS model). Prior to SEM examination, the samples were coated with ∼ 10 nm of gold. For observing the seeded cells, we performed fixation and dehydration protocols described in the next section.
Finite element model simulation
The topological surfaces with microtubes having the same dimensions and spatial arrangement as the imprinted ones were designed in SolidWorks®. The microtubes’ deformation at the contact with the cells was evaluated by finite element model simulation of the interaction between the cells and the microtubes underneath.
Cell seeding and culture
MG-63 osteoblast-like cell line purchased from the European Collection of Cell Cultures (ECACC, UK) was used for the experiments. The cells were plated into 25 cm2 culture flasks and cultured in growth medium (MEM, Biochrom) supplemented with 10% fetal bovine serum (Biochrom), 2 mML-glutamine, 1% (v/v) non-essential amino acids and 100 IUml−1 of penicillin/streptomycin. The cells were incubated in 5% CO2 atmosphere, at a temperature of 37 °C. After reaching confluence, the cells were trypsinized and seeded on the topological surfaces. Prior to cell seeding, these were sterilized for 2 h under UV light in the cell culture hood. All chemicals were purchased from Sigma-Aldrich, unless otherwise specified. Each sample was placed in a 3-cm-diameter Petri Dish and seeded with 2500 cells/sample. The incubation was done in standard conditions of atmosphere, humidity and temperature (5% CO2, 37 °C and 96%, respectively). The cells were monitored for a period of 4 weeks. Cells were also cultured on flat glass surfaces and used for comparison.
We employed an ultrasonic system (Sharpertek). The distance between the transducer and cells was about 10 cm. We employed low-frequency, low-intensity ultrasound at 40 kHz provided continuously at intensities of about 60 mW/cm2 SATA (surface averaged time averaged), for 20 min per day, during 5 consecutive days. No thermal effect was observed during the experiments. The selection of the stimulation parameters relies on previous studies reporting the benefits of continuous [17, 18], low-frequency LIPUS  on the process of bone regeneration.
After the LIPUS treatment, the cells were investigated in terms of morphology, viability, ALP activity and osteocalcin secretion, according to the below protocols.
Cells fluorescent staining and morphological observations
Cells staining was performed with F-Actin (Sigma) to evidence the cytoskeleton and with Hoechst (Molecular Probes) to mark the cells nuclei. For this, the cells were seeded on the topological surfaces and left to grow for 48 h, at 37◦C and 5% CO2. Then, they were washed three times with PBS, fixed in 3.7% paraformaldehyde for 10 min and finally rinsed with PBS. After being permeabilizated for 20 min with 0.1% Triton X, a solution containing Texas Red®-X phalloidin (TexasRed™-X Phalloidin, ThermoFisher scientific) was added and incubated for 24 h. Finally, the samples were washed three times with PBS and stained for 10 min with 2 mg/ml of Hoechst in PBS. The stained cells were visualized under an Olympus BX51 fluorescence microscope.
Scanning electron microscopy (SEM) analysis
For morphological investigations using SEM, the samples were washed with PBS and fixed for 1 h and at 37 °C with 2.5% glutaraldehyde prepared in PBS. Then, the samples were washed again with PBS and dehydrated in ethanol (EtOH) solutions, according to the following successive steps: 2 × 15 min wash with EtOH 70%, 2 × 15 min wash with EtOH 90% and 2 × 15 min wash with EtOH 100%. Next, the samples were washed in EtOH-HMDS solutions prepared in 50%:50%; 25%:75% and 0%:100% ratios, for 3 min in each solution, then left to dry and gold-sputtered. For closer relationship with the physical system, we also simulated in SolidWorks® the interaction of osteoblast-like cells with the microtubes arrays.
The cells viability was determined using the trypan blue dye exclusion method. The samples were placed in 12-well plates, and cells were seeded at 2500 cells/well. After incubation for 48 h, the samples were washed with PBS and the adherent cells were stained with trypan blue 0.4%. Non-viable cells were stained blue, and the cell viability was calculated as the number of viable cells divided by the total number of cells.
Alkaline phosphatase (ALP) activity
ALP activity was quantified by measuring the absorbance at 405 nm in the cell lysate, using a PerkinElmer UV–VIS Spectrophotometer and p-Nitrophenyl Phosphate Liquid Substrate System (N7653, Sigma). The results were expressed as units per milligram of protein in cell lysate. Protein was assayed by Bradford method (B6916, Sigma), using serum bovine albumin as the standard.
After 14, 21 and 28 days of cell culture, the medium was harvested and prepared for osteocalcin determination according to the producer’s specifications [Quantikine®ELISA Human Osteocalcin Immunoassay Catalog Number DSTCN0 (R&D SYSTEMS)]. The standard curve for osteocalcin calibration was obtained using standard osteocalcin solution. 50 μl from the harvested supernatant was added in a well, together with 100 μl of Assay Diluent. After 2 h of shacked incubation, the samples were washed 3 times with the washing buffer and 200 μl of conjugate was added in each well. The samples were shacked for 2 h at room temperature and washed 4 times with the washing buffer. 200 μl of substrate solution was added in each well, followed by 30 min of incubation, in dark. Finally, 50 μl of stop solution was added on each sample. The absorbance was read spectrophotometrically at 450 nm, with a correction at 570 nm, using the Mithras-Berthold Technologies plate reader.
For comparison purposes, all biological assays were also performed on cells cultivated on flat (glass) and on topological surfaces not exposed to LIPUS.
The experimental data were represented as the mean ± standard deviation of 5 different experiments. The statistical analysis was done using the two-tailed Student’s test. The data were considered statistically significant for p ≤ 0.05.
From the fluorescence images, we quantified the deformation of the cells nuclei. For this, we built a code in MATLAB starting from the “distance transform” function, which provides a metric or calculates the separation points between objects (cells nuclei in our case) from each group image. This function individually identifies the nuclei when they are far apart and create thin bridges between objects when some nuclei are close to each other. These bridges appeared mainly in the images for flat surfaces; in the images where cells were seeded on the topological surfaces, the nuclei images were better separated by the microrelief. To vanish the bridges, we introduced a method based on standard deviation to select separately each nucleus in a given image and detect their edges, centroids, radius, eccentricity, spread area and perimeter. To find the values for these parameters for each nucleus, we employed the regionprops function. From the list of these values, we excluded nuclei with extreme values by a threshold adapted separately for each image. The parameters considered significant for this type of analysis are: spread area S, perimeter P and eccentricity, for each nucleus separately. The eccentricity is the ratio between the foci of the object and its major axis. The shape index (SI) was defined as 4πS/P2 . 60 ± 5 nuclei of cells were selected and analyzed from each fluorescence image.
To investigate the vibration modes of the microtubes, we made a simulation in SolidWorks®. The solving method was FFEPlus. The detailed description of the parameters of the simulation and the results were added in the Supplementary information. For all microtubes heights, the normal vibration modes were found at frequencies up to 102–103 higher, i.e., 2–3 orders of magnitude higher than 40 kHz. For example, the lowest resonance frequency was 6964 kHz and was obtained for the tallest microtubes, i.e., 20 μm in height. Based on these results, one can conclude that all microtubes, regardless of their height, did not enter in a resonance state.
Figure 4 shows scanning electron micrographs of MG-63 osteoblast-like cells cultivated in static conditions on flat and topological surfaces. The cells from the topological surfaces looked less spread than on the flat ones, but they preserved their native polygonal morphology. Moreover, it can be observed that the cells exerted traction forces on the microtubes, which bended toward the center of the cell, as indicated by the red arrows from Fig. 4b–d. Furthermore, given that the flexibility increases with the microtubes’ height , we investigated if this has any influence on the interaction with the cells. For this, we cultivated the cells on topological surfaces with microtubes heights of 5, 10 and 20 μm, respectively. The highest microtubes (Fig. 4d) strongly bended under the action of the cells traction forces, while the shortest microtubes had much less conformational changes (Fig. 4b).
The microtubes’ displacement (bending) is displayed below each figure.
To investigate the vibration modes of the cells on the microtubes, we made a simulation in SolidWorks® and we used the FFEPlus solving method. For better reproducibility of the real physical system, we took into consideration that the osteoblasts have different shapes, specifically: triangular, rhombic and polygonal. The parameters of the simulation and the results are included in the Supplementary information. For all cell shapes, the normal vibration modes of the cells on the microtubes were found at frequencies ~ 103 higher than 40 kHz. One may therefore conclude that the cells on the microtubes do not enter in a resonance state with LIPUS.
Fluorescence microscopy analysis
For the same spread area, a nucleus with more corrugated edges will exhibit a small value for SI. With increasing SI toward 1, the shape of the nucleus is closer to a rounded native shape, with smooth edges. In the static regime, most of the cells nuclei from the flat surface had SI higher than 0.8, as they preserve their native, ovoidal shape, with smooth edges (Fig. 8a). The topological surface induced only slight nuclei deformations, with SI maximum lowered down to 0.6 for MT20 surface. On all topological surfaces, the LIPUS treatment induced deformations of the cells nuclei, with SI centered at lower values than in the static regime. This effect increased with increasing microtubes’ height. For example, for MT20 surface, most of the nuclei had SI values below 0.5, indicating that they were strongly deformed and the edges were corrugated.
The eccentricity (ECC) values provide complementary information regarding the changes in the shape of the cells nuclei as a whole. For a nucleus with a shape close to the circle, the eccentricity value is close to 0. Increasing eccentricity toward 1 indicates dramatic changes of the nuclei shape as a whole. The eccentricity of cells nuclei from flat and topological surfaces are displayed in Fig. 8b. For flat surfaces, most nuclei have eccentricity of 0.6, whereas the nuclei from the topological surfaces had an eccentricity up to 0.8. In agreement with the fluorescence images and the SI distributions, this result indicates a more pronounced nuclei deformation on the topological surfaces in comparison with the flat ones.
After LIPUS treatment, for both flat and topological surfaces, a growing number of nuclei shifted toward higher values of eccentricity, indicating stronger nuclei deformations as compared to the static regime. For the flat surfaces, the nuclei were less affected and thus most of them picked at eccentricity 0.7. The effect increased with microtube’ height. For MT20 surface the maximum of the eccentricity distribution reached 0.9 owing to dramatic changes of the cells nuclei.
These results confirm the synergic effects of topological surfaces and LIPUS treatment for inducing conformational changes in the cells nuclei. More specifically, the nuclei deformation was more severe in the LIPUS-treated samples and increased with increasing microtubes’ height.
The ALP activity was investigated as early osteoblastic marker with relevance to the gene expression of other osteoblastic differentiation markers and with important role in initiating bone mineralization . For all samples, the ALP activity reached a maximum at 21 days of cell culture, followed by a decrease in the next days, indicating the end of the differentiation process and the starting point of cells mineralization (Fig. 9b). The ALP activity related to surface topology increased significantly (p ≤ 0.05), up to 150%. At every time point, the ALP activity increased with microtubes’ height by 20–40%. The LIPUS treatment induced up to 50% increase in the ALP activity. The ALP activity enhancement owing to the LIPUS treatment only was less sensitive to microtubes height.
The osteocalcin secretion was further measured as late marker of differentiation that increases as mineral is deposited . The topological surfaces induced a statistically significant (p ≤ 0.05) increase of 50% in osteocalcin secretion in comparison with the flat surfaces (Fig. 9c). The increase related to the topological surfaces only was not sensitive to the microtubes’ height. On the contrary, the increase in the osteocalcin secretion related to LIPUS treatment strongly increased with increasing microtubes’ height. For example, the topological surfaces having the highest microtubes (20 µm) induced a 100% increase in osteocalcin secretion after the LIPUS treatment.
According to the experimental results and numerical calculations from Figs. 7 and 8, the strongest nuclei deformation occurred on topological surfaces with the highest microtubes, under LIPUS treatment. This provides evidence that the strongest increase in ALP activity and osteocalcin secretion in these samples directly relates to the extent of nuclei deformation.
Changes at cellular and subcellular level can be induced by the interaction between cells and topological surfaces [3, 4, 5, 6, 7, 10, 11]. For tissue engineering, microstructured surfaces with aligned pillars emerged as ideal platform to induce cytoskeleton and nuclei deformation that further trigger biochemical signaling . Topological surfaces with different pillar spacing’s, sizes and heights were developed to investigate the nuclei deformation [10, 11]. Cellular functions can be also modulated through the mechanical stiffness of the cellular environment . Patterned micropillars with various heights, where increasing height corresponds to decrease in stiffness, have been used to manipulate the mechanical regulation of human MSC differentiation .
Another efficient approach for bone healing, including for clinical trials, concerns the LIPUS effects performed at pulsed 1–3 MHz ultrasound regime [20, 29]. As an alternative, the effects generated by continuous kHz ultrasound therapy proved to be comparable . The positive influence of continuous wave ultrasound on the acceleration of bone healing of fresh fractures was confirmed in the early ‘50 s . To date, LIPUS studies have been performed at the kHz to MHz frequencies range, 0.005 to 1 W/cm2 intensities levels, in continuous or burst modes and with exposure times from 1 to 20 min per day . It is noteworthy that both MHz and kHz ultrasound proved to be equally effective in promoting osteoblast proliferation [30, 31, 32]. Adding the benefit of greater tissue penetration and wider exposure, low-frequency, low-intensity ultrasound is an effective alternative therapeutic tool. Recently, a clinical study proved the efficiency of low-frequency LIPUS for the treatment of metatarsal fracture .
In the present study, we enhanced the cells osteogenic differentiation on pillar-like topological surfaces consisting of vertical microtubes arrays and we amplified their osteogenic effect by low-frequency LIPUS. We found evidence that the topological surfaces induce dramatic changes in the nuclei shape, the cells nuclei wrapping around the microtubes. The nuclei deformation increased with increasing microtubes height, indicating that surface flexibility plays an important role. We also performed a finite element model simulation of the cell-substrate interaction modeled according to our experimental geometry that confirmed the influence of the substrate flexibility on the cell–surface interaction. We increased the nuclei deformation by LIPUS, and through this approach we succeeded to amplify the cells osteogenic response.
The flexible topological surfaces consisted of arrays of vertical microtubes in a hexagonal configuration, fabricated by LDW via TPP of IP-L780 photopolymer. The microtubes have inner diameters of 5 μm and heights of 5, 10 and 20 μm (Figs. 2, 3). The center-to-center spacing between the microtubes was 10 μm. This spacing was small enough to have a minimal effect on normal cell adhesion and function . The high accuracy and reproducibility of the microtubes dimensions provided by LDW via TPP fabrication method allowed us to control the substrate flexibility. The geometrical arrangement and dimensions of the microtubes restricted direct cell–cell interactions and allowed us to observe individual cells, as shown also by .
It would be useful to monitor the osteogenic effects on microtubes arrays with equal differences between heights, e.g., 5, 10 and 15 μm and even to check the cellular response on as many microtubes heights as possible. To argument our choice concerning the microtubes’ heights of 5, 10 and 25 μm, we mention the following. First, the role of the pillars’ height on the cellular behavior was already proven and discussed in terms of the influence of the pillars height on their flexibility, for heights between 0.97 and 12.9 μm . Second, the criterion of equal distance between the microtubes’ height is not critical for our study, since we do not intend to establish a linear relationship between the microtubes’ heights and the cells behavior, given that latter is far more complex than a simple linear dependence. Third, the osteogenic effect of each of the two approaches involved in our study (LIPUS and topological surfaces, respectively) was already demonstrated in previous studies. Our aim was to validate the possibility to merge these two approaches in a synergistic manner for further amplification of the osteogenic effect. Therefore, for reaching the lower and upper limits of any effects of the topological surfaces on the cellular behavior, we focused on scanning a broad range of microtube heights. To this end, we monitored the cellular behavior between the extreme values of the microtubes heights that were achievable in our experimental conditions. The lowest height (5 μm) was determined by the voxel size at the laser focus and by the “spiral-like” fashion of microtubes fabrication that involved a certain degree of overlapping of these voxels (as described in the "Materials and methods" section in the manuscript). The largest microtubes height (20 μm) was limited by the microtubes’ stability. For heights between 20 and 30 μm, a slight bending of the microtubes’ tip occurred (Figure S9 (a) in Supporting information). For microtubes heights above 30 μm, some of the microtubes were detached from the glass substrate, while others formed pairs of attached microtubes and clusters of more than two microtubes (Figure S9 (b) in Supporting information). These findings are in good agreement with the general characteristic of delicate polymer structures fabricated by LDW via TPP. Microstructures with high aspect ratio often tend to shrink and collapse during the fabrication process. This might be related to the strength of the structures against capillary forces, which mainly arise during the evaporation of the developer and rinsing liquids. In the case of micropillar arrays, when the spacing between pillars is smaller than their height, the pillars undergo a lateral collapse, they cluster or form agglomerates . By changing the developer and using liquids with low surface tension and/or post UV-curing, the pillars may stay upright and have straight tips . Nevertheless, these procedures are time consuming and beyond the purpose of our study. The intermediate value of 10 μm was selected in order to establish a trend of the cellular behavior as a function of the height/flexibility of the microtubes. Considering the first and second arguments presented above, this intermediate value for the height of microtubes should not be necessary in the middle of the 5–20 μm interval.
SEM analysis revealed that the cells from all topological surfaces were lifted away from the flat glass substrate, the cells anchoring on the top of the microtubes (Fig. 4). This observation is consistent with the study of Zeinab Jahed et al. reporting that glioma cells were close from the flat substrate only for short micropillars, i.e., heights below < 2 μm, when the cell protrusions extended between the micropillars . Another interesting observation was that the microtubes situated at the cells’ periphery bended toward the center of the cells (red arrows in Fig. 4), because of the cells traction forces . Cell traction forces in adherent cells like the osteoblasts are generated through the actomyosin interactions and act on the underlying substrate through focal adhesion proteins such as integrins. Previous studies using micropipette aspiration and pyramidal AFM  report that the traction forces trigger biochemical signaling network that connects the extracellular matrix, cytoskeleton and nuclei. In our experiments, the fact that the higher microtubes bended more than the shorter ones accounts for the strongest nuclei deformation (Figs. 6, 7, 8).
We further performed a SolidWorks® (Finite Element Model) simulation of the interaction between the cells and the microtubes underneath, in response to applied horizontal traction force of 8.5 μN (Fig. 5). The results confirmed the experimental observations according to which the bending of the microtubes (displayed in Fig. 4 as “microtubes displacement”) increased with their height. A similar conclusion was reported by Fu et al., which used the finite element method for analyzing deflections of hexagonally spaced poly-(dimethylsiloxane) microposts with a post diameter of 2 μm and heights between 1 and 12 μm, under different applied horizontal traction forces exerted by human mesenchymal stem cells . The simulation results are very similar with the images obtained by electron microscopy (Fig. 4 vs. 5).
We also examined the cytoskeletal arrangements on the topological surfaces by fluorescence microscopy (Fig. 6). On the flat surfaces, the actin filaments were homogenously spread. On all topological surfaces, irrespective of their height, the cells exhibited localized, elongated, stress fibers aligned along free spaces between the microtubes. A similar cellular response was reported for glioma cells grown on of silicon micropillar arrays  and confirms the role of surface topography in modulating the cellular behavior.
We further looked into more details related to shape changes of the cells nuclei (Fig. 7). When the cells attached on the topological surfaces, the nuclei changed their shape and wrapped around the microtubes. Nuclei deformation triggered by topological surfaces with micropillar arrays has been reported in several studies. Micropillar arrays of poly(lactide-co-glycolide) were used for nuclei deformation of mesenchymal stem cells and osteoblasts . Cellular functions were also modulated via the mechanical stiffness of the cellular environment . Patterned micropillars with various heights, where increasing height corresponds to decreased stiffness, have been used to manipulate the mechanical regulation of human MSC differentiation . Microtube cell culture structures were employed to determine the spatial limit for the “self-imposed” nuclei deformation .
In our study, we quantified of nuclei deformation through the shape index (SI) and eccentricity (ECC) parameters. We recall that SI values reflect the deformation of the nucleus’ edges, while the eccentricity provides information about the nucleus deformation as a whole. According to Fig. 8, SI values below 0.6 and ECC higher than 0.6 indicate a significantly deformed nucleus. The nuclei deformation increased with increasing microtubes’ height thus with substrate flexibility. Moreover, the nuclei deformation increased after LIPUS treatment in comparison with the cells cultured in static conditions. Overall, these results provide evidence that the interplay between the cellular traction forces, substrate flexibility and LIPUS treatment results in different degrees of nuclei deformation.
The “inside-out” signals from the nuclei are conducted to the cytoskeleton network; the action filaments exert a pulling force upon the deformed nuclei that counts for the osteogenic response . Marino et al. demonstrated that the enhancement of the osteogenesis can be triggered by changes in the nucleus shape induced by bioinspired 3D structures . Therefore, we further evaluated if the changes of the nuclei shape observed in our experimental conditions, related to the topological surfaces and LIPUS treatment, influence the cells osteogenic differentiation.
LIPUS is transmitted to the living cells as an acoustic pressure wave that triggers biochemical events at cellular level and has been successfully used for promoting bone formation in animal models and clinical treatments [37, 38, 39, 40, 41]. However, previously reported studies on LIPUS treatment of bone-forming cells on flat surfaces indicate lower levels of osteoblastic differentiation than those obtained through our experimental approach. LIPUS treatment of human alveolar bone-derived mesenchymal stem cells on flat culture dishes reported an increase in ALP activity of about 25% , while LIPUS exposure of human periodontal ligament cells on flat surfaces induced only a 20% increase in ALP activity and about 25% increase in osteocalcin secretion, as compared to the samples in static regime [16, 23].
In our experiments, ALP activity related to surface topology increased up to 150% (at day 7) and was slightly improved by microtubes height by 20–40%. ALP activity related to LIPUS treatment increased up to 50% and was less sensitive to microtubes height (Fig. 9b). Osteocalcin secretion had about 50% increase related to topological surfaces, in comparison with the flat surfaces. This increase was not sensitive to the microtubes’ height. On the contrary, the osteocalcin secretion increase related to LIPUS treatment was strongly dependent on microtubes’ height as measured on the same surfaces with or without ultrasound stimulation. The osteogenic effect increased with increasing the microtubes height. We observed over 100% increase in osteocalcin levels for the 20-µm-long microtubes (Fig. 9c). The osteogenic markers levels are following closely the nucleus deformation data. Especially, the eccentricity of cell nuclei and both osteogenic markers have similar trends (Fig. 8b vs. 9c for example).
To conclude, we succeeded to amplify the nuclei deformation on the topological surfaces by LIPUS treatment. In fact, in our experimental conditions, the LIPUS treatment and topological surfaces have both an osteogenic effect that increased with the microtubes’ flexibility. The level of influence on osteogenesis of the topological surfaces and LIPUS is varying between 50 and 150% for each of them. However, the weighting trends on osteogenesis of the two stimulations are dissimilar. Overall, both topological surfaces and LIPUS treatment are increasing the level of osteogenic markers up to 200%.
The positive effect on osteogenesis of topological surfaces and LIPUS treatment depends strongly on the geometry, but nevertheless they are additive. We are aware that are two concurrent processes. The more flexibility of the substrate, the more increase in osteogenesis, but a decrease in structural strength. A further optimization should be done in order to address this issue, involving different structures for mechanical strength increase.
We proved that the osteogenic differentiation of osteoblast-like cells can be dramatically enhanced through cells nuclei deformation induced by flexible topological surfaces in conjunction with ultrasound stimulation. The topological surfaces were fabricated as arrays of vertical microtubes with precise positioning and dimensions that allowed us to explore single-cell behavior. These surfaces induced dramatic changes of the cells nuclei that wrapped around the microtubes. The nuclei deformation increased with increasing microtubes’ height. Experimental results and numerical simulations showed that, under the action of the cells traction forces, the higher microtubes bended more significantly than the shorter ones, accounting for stronger deformation of the cells nuclei. Low-frequency LIPUS caused additional changes of the nuclei shape. This further enhanced the cells osteogenic differentiation in comparison with the corresponding samples in static regime. A causal relationship between the nuclei deformation and the cells osteogenic response was demonstrated. A proof-of-concept on the synergy between the topological surfaces and LIPUS treatment that allowed us to increase the level of osteogenic markers up to 200% was proved. The results strongly support the validity of the proposed approach that offers great potential for bone tissue engineering and can be easily adapted for other cell types.
This work was supported by a grant of the Romanian National Authority for Scientific Research and Innovation, CNCS/CCCDI—UEFISCDI, project number PN-III-P2-2.1-PED-2016-1787. A part of this work was supported by the National Program LAPLAS VI (16N/08.02.2019).
Compliance with ethical standards
Conflict of interest
The authors declare that they have no conflict of interest.
- 1.Amini AR, Laurencin CT, Nukavarapu SP (2012) Bone tissue engineering: recent advances and challenges. Crit Rev Biomed Eng 40:363–408. https://doi.org/10.1615/CritRevBiomedEng.v40.i5.10 CrossRefGoogle Scholar
- 5.Fu X, Xu M, Liu J, Qi Y, Li S, Wang H (2014) Regulation of migratory activity of human keratinocytes by topography of multiscale collagen-containing nanofibrous matrices. Biomaterials 35:1496–1506. https://doi.org/10.1016/j.biomaterials.2013.11.013 CrossRefGoogle Scholar
- 12.Dimitriou R, Babis GC (2007) Biomaterial osseointegration enhancement with biophysical stimulation. J Musculoskelet Neuronal Interact 7:253–265Google Scholar
- 14.Paun IA, Zamfirescu M, Luculescu CR, Acasandrei AM, Mustaciosu CC, Mihailescu M, Dinescu M (2017) Electrically responsive microreservoires for controllable delivery of dexamethasone in bone tissue engineering. Appl Surf Sci 392:321–331. https://doi.org/10.1016/j.apsusc.2016.09.027 CrossRefGoogle Scholar
- 15.Paun IA, Popescu RC, Mustaciosu CC, Zamfirescu M, Calin BS, Mihailescu M, Dinescu M, Popescu A, Chioibasu D, Soproniy M, Luculescu CR (2018) Laser-direct writing by two-photon polymerization of 3D honeycomb-like structures for bone regeneration. Biofabrication 10:025009. https://doi.org/10.1088/1758-5090/aaa718 CrossRefGoogle Scholar
- 17.Romano CL, Romano D, Logoluso N (2009) Low-intensity pulsed ultrasound for the treatment of bone delayed union or nonunion: a review. Ultrasound Med Biol 35:529–536. https://doi.org/10.1016/j.ultrasmedbio.2008.09.029 CrossRefGoogle Scholar
- 18.Escoffre JM, Bouakaz A (eds) (2016) Therapeutic ultrasound. Springer, Belrin. https://doi.org/10.1007/978-3-319-22536-4_1
- 21.FA Duck (1990) Physical properties of tissue: a comprehensive reference book. Academic Press, London. https://doi.org/10.1016/B978-0-12-222800-1.50009-7
- 25.Bandow K, Nishikawa Y, Ohnishi T, Kakimoto K, Soejima K, Iwabuchi S, Kuroe K, Matsuguchi T (2007) Low-intensity pulsed ultrasound (LIPUS) induces RANKL, MCP-1, and MIP-1beta expression in osteoblasts through the angiotensin II type 1 receptor. J Cell Physiol 211:392–398. https://doi.org/10.1002/jcp.20944 CrossRefGoogle Scholar
- 28.Nikkhah M, Edalat F, Manoucheri S, Khademhosseini A (2012) Engineering microscale topographies to control the cell–substrate interface. Biomaterials 33:5230–5246. https://doi.org/10.1016/j.biomaterials.2012.03.079 CrossRefGoogle Scholar
- 41.Zheng H, Lu L, Song JL, Gao X, Deng F, Wang ZB (2011) Low intensity pulsed ultrasound combined with guided tissue regeneration for promoting the repair of defect at canines periodontal fenestration in beagle dogs. Zhonghua Kouqiang Yixue Zazhi 46:431–436Google Scholar
- 42.Lim K, Kim J, Seonwoo H, Park SH, Choung PH, Chung JH (2013) In vitro effects of low-intensity pulsed ultrasound stimulation on the osteogenic differentiation of human alveolar bone-derived mesenchymal stem cells for tooth tissue engineering. Biomed Res Int 2013:269724. https://doi.org/10.1155/2013/269724 Google Scholar