mTORC2 regulates hierarchical micro/nano topography‐induced osteogenic differentiation via promoting cell adhesion and cytoskeletal polymerization

Abstract Surface topography acts as an irreplaceable role in the long‐term success of intraosseous implants. In this study, we prepared the hierarchical micro/nano topography using selective laser melting combined with alkali heat treatment (SLM‐AHT) and explored the underlying mechanism of SLM‐AHT surface‐elicited osteogenesis. Our results show that cells cultured on SLM‐AHT surface possess the largest number of mature FAs and exhibit a cytoskeleton reorganization compared with control groups. SLM‐AHT surface could also significantly upregulate the expression of the cell adhesion‐related molecule p‐FAK, the osteogenic differentiation‐related molecules RUNX2 and OCN as well as the mTORC2 signalling pathway key molecule Rictor. Notably, after the knocked‐down of Rictor, there were no longer significant differences in the gene expression levels of the cell adhesion‐related molecules and osteogenic differentiation‐related molecules among the three titanium surfaces, and the cells on SLM‐AHT surface failed to trigger cytoskeleton reorganization. In conclusion, the results suggest that mTORC2 can regulate the hierarchical micro/nano topography‐mediated osteogenesis via cell adhesion and cytoskeletal reorganization.


| INTRODUC TI ON
Implant surface topography can regulate cell behaviour and ultimately be involved in cell fate decisions. [1][2][3] For dental implants, the clinical widely used grit-blasted and acid-etched (SLA) titanium surface with single micro-scale topography which has been proved much better in osteointegration than smooth (S) titanium surface, partly due to the facilitating role of the micro-scale structure in bone locking and implant initial stability. [4][5][6] However, it was reported that the single micro-scale feature might inhibit cell attachment and proliferation. 6,7 To achieve long-term success of the intraosseous implants, it is necessary to alleviate the inhibitory effects of the single micro-scale topography. Recently, the hierarchical micro/nano topography has attracted extensive attention since the nano-scale feature can increase the adsorption of proteins and subsequently enhance cell attachment. 8,9 Notably, mimicking the natural bone structure which consists of micro-scale collagen fibers and nanoscale hydroxyapatite, the hierarchical micro/nano topography titanium surface provides a better microenvironment than that with signal micro-scale topography for cell-surface interaction. [10][11][12][13][14][15] In our previous study, we have revealed that the hierarchical micro/nano topography was superior to the SLA titanium surface in improving the osteogenesis. 16,17 However, the elaborate regulation process of the cell-surface interaction and the underlying mechanism have remained to be elucidated.
Cells perceive the implant surface through various of mechanosensors. It is widely known that cell adhesion and actin cytoskeleton have a central role in sensing and transmitting extracellular stimuli based on the connection between cell membranes and nuclears mechanically. Biochemically, cell adhesion is mediated by the integrin (at the nano-scale). 18 The interaction and gathering of integrin result in the assembly of several intracellular ankyrins (talin, vinculin, etc) to induce the formation of mature focal adhesion (FA) (at micro-scale), which connects the material surface and the cytoskeleton to propagate the biochemical signalling. 19 Among the diverse adhesion-related signalling pathways, focal adhesion kinase (FAK) was considered as an important one regulating the actin cytoskeleton organization-related signallings, including Cdc42, Rac and ROCK, as well as participating in cell fate decisions. 20 Concomitantly, linked to the adhesion-related molecules, actin cytoskeleton also acts as a critical modulator generating intracellular tension which contributes to the regulation of cell phenotype. Overall, compelling evidence supports a critical role of the cell adhesion and actin cytoskeleton in the hierarchal micro/nano topography-elicited osteogenesis. 12,21,22 Our former works have indeed shown that the micro/ nano topography could direct cell fate via promoting cell adhesion, 16 polymerization of cytoskeleton, and the regulation of chromatin modifications. 17 However, relatively little has been known about the molecular mechanism how topography regulates cell adhesion and cytoskeleton up to now. Herein, we sought to observe the underlying mechanism of the hierarchical microgroove/nanopore topography we fabricated in regulating the cell adhesion, actin cytoskeleton and finally osteogenic differentiation.
Mammalian target of rapamycin (mTOR) is an evolutionarily conserved serine/threonine protein kinase which could interact with different proteins and form two functionally distinct complexes termed mTORC1 and mTORC2. mTORC1 involves Raptor as a unique adaptor protein rather than Rictor in mTORC2. mTOCR1 can sense various environmental conditions, like insulin and serum, coordinating multiple cell processes from catabolism and anabolism of protein and lipid to autophagy. [23][24][25] By contrast, the main function of mTORC2 is to phosphorylate AGC subfamily of kinases, such as AKT and PKCα, which regulate cell proliferation, survival and actin cytoskeleton, 25,26 while the function and underlying mechanism of mTORC2 are still in the exploratory stage. 27 In recent years, growing evidence has implicated that mTORC2 plays a critical role in bone homeostasis. [28][29][30] Rictor knock-down in mature osteoblast and BMSCs resulted in impaired osteogenic differentiation in vitro and compromised bone formation in vivo. 31,32 In the process of osteoblast differentiation, mTORC2 was activated as the downstream of canonical Wnt signalling pathway. 33,34 In turn, osteogenic gene RUNX2 could directly bind to the promoter of mTOR and activate the mTORC2/AKT signalling pathway. 35,36 The phosphorylation of Akt-Ser473, the best characterized substrate of mTORC2, has been demonstrated to be necessary for osteogenesis. 37 Given that Rictor has been found to regulate cytoskeleton through PKCα initially, 38,39 and overall proteome analyses have shown that the function of mTORC2 was highly associated with cell adhesion in cancer cells, 40 we supposed that mTORC2 might be responsible for the topographical cues-induced osteogenic differentiation, through regulating cell adhesion and cytoskeletal polymerization.
In the presented study, we fabricated the titanium surface with hierarchal microgroove/nanopore topography by using the selective laser melting (SLM) technique combined with alkali heat treatment (AHT). We hypothesized that, during cell reading hierarchal micro/ nano topography, mTORC2 was activated to enhance cell adhesion and cytoskeletal polymerization, which in turn promoted osteogenesis. To verify this, we first proved that cell adhesion and mTORC2 signalling pathway could be activated by the hierarchal micro/ nano topography. Moreover, Rictor stable knock-down MC3T3-E1 cells were used to confirm the role and underlying mechanism of mTORC2 in topographical cues-induced osteogenic differentiation. Our results demonstrated that mTORC2 was essential in this process. In the absence of mTORC2 signalling, topographical cuesinduced signalling transduction based on the cell adhesion and the actin cytoskeletal polymerization will be blocked and consequently impair the osteogenesis. Then, the specimens were treated by 5 mol/L NaOH at 100℃ for 2 hours. Next, the specimens were heated in muffle furnace from 0 ℃ to 600 ℃ (5℃/min). The resultant titanium specimens were cleaned ultrasonically in acetone, absolute ethanol and doubledistilled water (ddH2O) sequentially for 15 min and sterilized at 120℃/2 h. The specimens were sterilized with ultraviolet light for at least 30 minutes before use.

| Construction of Rictor knock-down cell lines
To explore the role of mTORC2 in the hierarchical micro/nano topography-induced osteogenesis, we designed two short hairpin RNAs (shRNAs) to knock down Rictor in MC3T3-E1 cells. Scramble shRNA was employed as a control group. Rictor-shRNA oligos (as shown in Table 1) were purchased from GENEWIZ (China), and scramble-shRNA plasmid was preserved in our laboratory. Briefly, shRNA oligos were annealed and ligated into digested pLKO.1 vector, and the correctly identified sequence was transfected into 293T cells using PAX8 (packaging) and VSVG (enveloping) plasmid. Virus supernatants were harvested 48 hours later to infect MC3T3-E1 cells at 70% confluence. After 48 hours infection, puromycin was added to selected positive cells. RT-qPCR and Western blot were used to examine the Rictor knock-down efficiency in the method detailed in the corresponding sector.

| MTS
To evaluate the cells' proliferation, scramble cells and Rictor knockdown cells were seeded at a density of 1 × 10 3 cells/well. After 24 hours, 2 and 4 days culturing, the culture medium was replaced by empty DMEM, MTS and PMS (100:20:1) and then cells were cultured at 37℃ for 2 hours. Finally, the absorbance at 490 nm wavelength was detected and OD value was calculated.

| Immunoprecipitation
To investigate the interaction of vinculin and Rictor, immunoprecipitation was performed. Wild-type cells cultured on the three surfaces were collected and lysed on ice for 30 minutes, sonicated and centrifugated, and 3% supernatant was collected as input. Vinculin antibody (CST) was added to the incubation at 4℃ overnight, followed by 3-hour incubation with Protein A/G beads (Smart-Lifesciences).
Immunoprecipitates were washed three times and resuspended when buffer is loaded for SDS-PAGE analysis.

| Immunofluorescence
To visualize FA formation, the state of the actin cytoskeleton and the subcellular localization of Rictor, wild-type cells and Rictor knock-down cells were seeded at a density of 1 × Finally, the moderate mounting media (changjia) was added on microscope slide (changjia), then the specimen was putted on the microscope slide carefully to prepared for subsequent confocal image observation. Briefly, Image J software was employed to calculate the area of vinculin stain. The area which was greater than 3.14 µm 2 was acted as mature FA.

| RT-qPCR
Wild-type cells were seeded at a density of 1 ×  Table 2.

| Western blot
Wild-type cells, scramble cells and Rictor knock-down cells were

| Surface topography
As the FE-SEM images shown in Figure 1A,   Figure 2A,B), demonstrating that the SLM-AHT surface could activate adhesion-related FAK signalling pathway. Furthermore, as shown in Figure 2C-F, cells on SLM-AHT surface exhibited fewer FAs in total but more mature FAs than they did in the other two groups, suggesting that SLM-AHT surface could promote mature FAs formation.

| The effect of hierarchical micro/nano topography on cell adhesion, actin cytoskeleton and eventually cell osteogenesis
As shown in Figure 2C, cells cultured on SLM-AHT surface showed a typical polygonal, elongated morphology, and the actin fibres were arranged in an orderly way with higher intensity. In contrast, cells on control groups were round in shape, and the cytoskeleton was in a disorderly state with lower intensity, indicating SLM-AHT surface could trigger the polymerization of the cytoskeleton.
After 3 days of culture, the gene expression level of RUNX2 was slightly enhanced, while the protein expression level of RUNX2 was significantly enhanced on SLM-AHT surface ( Figure 3A,B). After 7 days of culture, a considerable increased expression of RUNX2 were detected on SLM-AHT surface in comparison with control groups (Figure 3D,E).
Consistently, the strongest RUNX2 positive stain was observed in the cells cultured on SLM-AHT surface ( Figure 3C,F). Meanwhile, the gene expression level of late-stage osteogenic differentiation marker OCN was notably increased after 7 days instead of 3 days of culture on SLM-AHT surface ( Figure 3A,D). Collectively, it could be inferred that SLM-AHT surface has greater potential in promoting cell osteogenic differentiation than the single micro-scale surface.

| The role of mTORC2 in topographical cuesinduced cell osteogenic differentiation
As shown in Figure

| The relationship between mTORC2 activation and hierarchical micro/nano topographyinduced cell adhesion and cytoskeletal polymerization
To further unravel the molecular mechanism of mTORC2 in hierarchical micro/nano topography-mediated osteogenesis, immunofluorescence was employed to visualize the subcellular localization of Rictor. Rictor  increase protein adsorption, cell adhesion and ultimately osseointegration. 8,42 However, the underlying mechanism that surface topography manipulates cell fate still requires further investigation.

| D ISCUSS I ON
In the current study, we utilized SLM to fabricate microgroove titanium surface, on which AHT was employed to create nanopore features. The resultant specimens were used to explore the effect of the SLM-AHT surface on cell adhesion, actin cytoskeleton and eventually osteogenesis.
Adhesion of cells to implant surface was considered as the very beginning of the osseointegration. 18,43,44 In the present study, we Compelling evidence has indicated that mTORC2 plays a crucial role in regulating bone homeostasis including both bone formation and absorption. [49][50][51] Given the evidence that mTORC2 was sensitive to mechanical cues and was essential in osteogenesis, 32   in the process of cell-reading hierarchical micro/nano topography. mTORC2 could be activated by a variety of biochemical signallings, such as WNT/LRP5 34 and Hedgehog, 58 involved in the regulation of osteogenesis. 59 Therefore, mTORC2-involved biochemical signalling pathway in SLM-AHT surface-mediated osteogenesis still awaits further investigation.
In conclusion, mTORC2 activation in response to hierarchical microgroove/nanopore topography leads to enhancement of cell adhesion and polymerization of the cytoskeleton, which allows for an amplification of topographical cues orchestrating cell osteogenic differentiation. It is considered that there was an interaction between mechanical and biochemical signalling pathways, as well an interplay between the intrinsic and the extrinsic mechanical environment. Further experiments are needed to explore the more detailed molecular mechanisms both in vivo and in vitro.

CO N FLI C T O F I NTE R E S T
No potential conflicts of interest.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available on request from the authors.