Titanium and its alloys are widely used as orthopedics and dental implants due to their excellent biocompatibility, robust mechanical properties and corrosion resistance. However, in bone reconstruction therapy, peri-implantitis [1,2] and insufficient or poor prior bone healing at the interface of bone/implant in the early post-implantation stage [3] remain serious challenges for clinical application. The traditional treatment contains systemic antibiotic administration, wound drainage, surgical debridement and implant exclusion [4], etc. However, these approaches are quite inefficient and may cause further surgical intervention for the patients [5]. In order to solve these problems and develop a multifunctional implant material surface with osteogenic ability, antibacterial activity and long-term stability, implant surface modification is being required.

Surface modifications of implant materials have attracted the attention of researchers for osseointegration and antibacterial properties in the past decade [[6], [7], [8], [9]]. As a result, many investigations have been published to enhance the surface modifications of a bone implant to obtain preferred natural biocompatibility. One triumphant solution of the surface modification is sandblasting and large-grits etching (SLA), which can improve the bonding strength between the implants and tissue and also promote the new bone formation. The combined positive effects of grit-blasting and etching to produce a rough surface (macroscopically) with micro pits [10,11]. Rough surfaces have been shown to remarkably improve the bonding force between the implants and bone tissue and also enhanced the rate of bone regeneration [[12], [13], [14], [15]].

On the other hand, bactericidal coatings [[16], [17], [18], [19]] have been extensively studied against implant-associated infections, while there are some concerns regarding rapid drug delivery and weak binding with matrix [20,21]. If one can endow implants with the native osteogenic ability and inherent antibacterial activity, this multifunctional implant will be beneficial to overcome the existing problems of current implant materials.

Copper (Cu) is not only a commonly alloying element, but also utilized desirably in the biomedical field because it is an essential cofactor of several enzymes as well as demonstrates osteogenic and proangiogenic funcations [22] and antibacterial activity [23]. Previous studies suggested that the copper-bearing stainless steels have shown excellent anti-infective function [24] and outstanding endothelialization ability [25]. Moreover, a novel titanium-copper alloy (TiCu) [26] also possessed antibacterial [27] and osteogenic ability in vivo and in vitro [[28], [29], [30]] as well as good mechanical properties and excellent corrosion resistance [31], which consisted of α-phase and Ti₂Cu.

From all these reasons, it is believed that a surface roughness treatment of TiCu alloy can further enhance the antibacterial activity and osteogenic ability. The rough surfaces of both titanium-copper alloy and pure titanium (SLA-TiCu and SLA-Ti) were obtained by SLA treatment. Compared with mechanically ground surfaces of titanium-copper alloy and pure titanium (M-TiCu and M-Ti), the effects of various surfaces on the proliferation, adhesion, apoptosis, differentiation and bone-related gene expressions of MC3T3-E1 cells were systematically studied, and the antibacterial properties of different surfaces against Staphylococcus aureus (S.aureus) were also investigated, which would make contribution to the development of a novel titanium based multifunctional implant material with persistent osteogenic ability and antibacterial activity.

Titanium-copper alloy (TiCu) was produced by arc melting CP-Ti (99.86%) and pure Cu (99.99%) in a 30 kg vacuum consumable furnace, whose chemical composition was analyzed as (wt.%): Cu 5.02, C 0.011, N 0.002, H 0.001, O 0.045, and Ti in balance. The cast ingot was heat treated at 850 °C for 2 h followed by air cooling. Commercial-grade 2 pure titanium (Ti) was used as the reference material. These materials were cut into discs with diameter of 10 mm and thickness of 2 mm. M-TiCu and M-Ti surfaces were mechanically ground with various grades silicon carbide papers. Rough surfaces were achieved by sandblasting with large alumina grits of 190 μm under 0.2 MPa pressure and etching with HCl/H₂SO₄ for 3-5 min, named as SLA-Ti and SLA-TiCu. All the samples were ultrasonically cleaned in acetone, anhydrous ethanol and deionized water, and sterilized in the autoclave prior to experiments.

Surface mapping microscope (Micro XAM, KLA-Tencor, USA) is a non-contact measuring instrument using white-light interferometry in a microscopic range, which is also used to measure 3D profiles and amplitude roughness parameters. It was used to measure different surface roughness characteristics in a square field with 146 × 146 μm. Furthermore, the surface morphology and cross-sectional microstructure were observed by scanning electron microscopy (SEM, SHIMADZU SSX-550). For cross-section views, samples were etched in Kroll solutions for 10 s and the chemical compositions were detected by energy dispersive spectrum under BSE model.

The copper ions released from SLA-TiCu and M-TiCu surfaces were measured according to international standard ISO 10993-12. Each sample was immersed in 0.9% NaCl solution with an area to volume ratio (1.25 cm²/mL) under 37 °C for 1,3,7,14 and 21 days. The ions concentration was calculated by inductively coupled plasma-atomic emission spectroscopy ICP-MS, Thermo, America.

Contact angles were examined through the sessile drop method with deionized water and calculated by the mean average of five independent surfaces per group via contact angle-measuring instrument (Theta Lite, Biolin Scientific, Sweden). The angles image of the deionized water drop was taken after 10 s of release, which was the expected required time to get equilibrium. The wettability is strongly influenced by surface roughness [32,33]. Therefore, the Wenzel model [34] Eq. (1) was chosen to calculate the roughness effects on contact angle.

Where, R_w shows the surface area ratio, also referred to as Wenzel factor, θ_m is the apparent contact angle associated to the system equilibrium state and θ_Y is the Young contact angle.

2.5.1. Cell culture

MC3T3-E1 cells (Cyagen, China), a clonal osteoblast-like mouse calvarial cell line, were cultured in the Eagle’s alpha minimum essential medium (α-MEM) supplemented with 10% fetal bovine serum (FBS, Gibco, USA) and 1% penicillin/streptomycin (P/S, Hyclone, USA) at 37 °C in a humidified atmosphere of 5% CO₂. The culture media was refreshed at two-day intervals. For the osteogenic differentiation experiment, MC3T3-E1 cells were seeded on the material surfaces at a density of 1 × 10⁴ cells/cm² in 48 well plates. Three days later, cells got confluence and subsequently cultured for 4-21 days in differentiation medium containing 50 μg/mL concentration of the phosphate ester of ascorbic acid (Sigma, USA), 100 nM dexamethasone (Sigma, USA), 10 mM β-glycerophosphate (Sigma, USA), 10% FBS as well as α-MEM.

2.5.2. Cell proliferation assay

To estimate the effects of various surfaces on the proliferation of MC3T3-E1 cells, CCK-8 assay (Beyotime, China) was performed in triplicate for different days. Briefly, cells were cultured on all samples and placed in 48 well plates at an initial concentration of 1 × 10⁴ cells/cm². After incubating for 1, 4 and 7 days, respectively, the culture media was removed and these samples were rinsed with phosphate buffered saline (PBS, Hyclone, USA) to remove the unattached cells. 300 μL of medium with 10% CCK-8 assay was added and stored in a dark incubator for 2 h at 37 °C. Afterward, 100 μL of solution was moved to a new plate (96-well), and the optical density (OD) was obtained at 450 nm on a microplate reader (MuLtiskan GO, Thermo Scientific). All the results were calculated as the optical density values minus the absorbance of blank wells, and the cell viability was demonstrated by a relative growth rate (RGR) [26].

2.5.3. Cell morphology

The attachment and morphology of cells on different surfaces were examined via fluorescence microscopy. Cells were seeded at a density of 1 × 10⁴ cells/cm² and cultured for 1 and 4 days, respectively, then fixed with 4% (w/v) paraformaldehyde for 10 min, followed by rinsing with PBS and permeabilizing with 0.1% (v/v) Triton X-100 (Beyotime, China) for 5 min. Afterward, 50 μL rhodamine-phalloidin (Molecular Probes® Thermo Fisher Scientific) was utilized to stain MC3T3-E1 cells for 40 min to visualize the filamentous actions of the cell cytoskeletons. Moreover, 5 mg/mL 40,6-diamidino-2-phenylindole (DAPI) solution was added to counterstain the cell nuclei. The density of adherent cells (cell numbers per unit area, 10⁴ cells/cm²) was calculated by the number of nuclei on the projected area (magnification 100×). Stained images were obtained from five different areas per sample.

2.5.4. Cell apoptosis assay

Cell apoptosis was determined through an Annexin V-fluorescein isothiocyanate (FITC) apoptosis detection kit (BD, USA). MC3T3-E1 cells were cultured on various surfaces at a concentration of 1 × 10⁴ cells/cm² for 4 and 7 days and then digested by non-pancreatic enzyme for 2 min and washed with fresh culture medium and cold PBS buffer solution twice. Cells were stained with the apoptosis detection kit and then samples were evaluated by a FACS Calibur (BD, USA) flow cytometry.

2.5.5. Alkaline phosphatase activity

The early stage differentiation of MC3T3-E1 cells on different surfaces was examined by the expression of alkaline phosphates (ALP). After 4 and 7 days, the ALP activity was quantified by the alkaline phosphatase assay kit (Beyotime, China) according to the manufacturer’s instructions. For normalization, the total protein concentration was measured by a bicinchoninic acid protein assay kit (Beyotime, China). ALP activity was normalized and expressed as ALP/total protein (OD/μg).

2.5.6. Extracellular matrix (ECM) mineralization

After culturing for 7 and 14 days the ECM mineralization of osteoblasts was stained by 20 mg/mL alizarin red (ARS). For quantitative comparison, the samples were treated with 10% (w/v) paraformaldehyde for 15 min. Then 500 μL ARS was adjusted PH to 4.2 and put in each well at 37 °C, after incubating for 30 min, then cleaned by PBS. The newly-formed bone-like nodules on the sample surfaces were observed in red color. The same volume of 10% hexadecylpyridinium chloride (Aladdin, China) was added to dissolve the crystals. The solution was moved to a new plate (96-well), and the ECM mineralization was quantified by detecting the optical density of dissolution with a microplate reader at 620 nm.

2.5.7. Real-time quantitative PCR (RT-qPCR) analysis

The effect of various surfaces on the MC3T3-E1 cells differentiation was studied by RT-qPCR to analyze the mRNA expression of bone-related genes. The genes of alkaline phosphates (ALP), collagen type I (COL I), Runt-related transcription factor 2 (RUN × 2), and osteopontin (OPN) were evaluated, and all the mRNA levels of above genes were measured and normalized to the housekeeping gene of GAPDH (Table 1).

2.6.1. Quantitive evaluation

The plate counting method was used for investigating the antibacterial effect of various surfaces. 50 μL of S.aureus (ATCC6538) suspension at a concentration of 2 × 10⁷ CFU/mL in PBS was put on each surface in a 24 well plate and stored at 37 °C for 1day. Afterward, the incubated strain was collected in a sterilized tube with 5 mL PBS and vortexed for 1 min to get a homogeneously mixed solution. Then 50 μL of the solution was spread onto the L.B plates and cultured at 37 °C for 24 h. The antibacterial rate [26] was calculated as the equal-volume bacterial suspension without samples as control group.

2.6.2. Analysis of bacterial vitality

In order to further explore the effect of different sample surfaces on the vitality of adhered bacteria, samples were stained with LIVE/DEAD® BacLightTM Bacterial Viability Kit solution (Leiden, Netherlands) at room temperature. After incubation with an initial concentration of 10⁷ CFU/mL bacterial suspension with 10% L.B for 1 day, the non-invasive confocal imaging of bacteria on different surfaces was carried out via a confocal laser scanning microscope (CLSM), and the image was reconstructed in 3D model by the NIS Viewer software.

All the results were expressed as mean ± standard deviations and each experiments were repeated three times. For the statistical analysis, the student’s t-test was used to analyze the significant differences among the groups via the SPSS software, and P-values<0.05 were accepted statistically.

Surface topography of a medical implant in association with other surface characteristics, such as surface wettability, surface energy, charge and functional groups, can influence the biological cascade of events at the biomaterial/host interface, including protein adsorption, hard- and soft-tissue interactions with the implant and bacterial film formation [35]. Therefore, surface morphology and wettability of all the materials were characterized in this study. The 3D surface profiles of M-Ti, M-TiCu, SLA-Ti and SLA-TiCu surfaces were shown in Fig. 1. Mechanically ground surfaces demonstrated grooves in the same direction, while rough surfaces were more uniform and non-directional. The surface topography parameters of these surfaces were given in Table 2, the average surface roughness (S_a) values of M-Ti and M-TiCu surfaces were about 0.25 μm, however, SLA-Ti and SLA-TiCu surfaces roughness significantly increased up to 1.89 μm. Under the same surface treatment, surfaces of TiCu alloy were much rougher than pure Ti surfaces, but their S_a values had no significant difference.

Fig. 2 depicts the surface and cross-sectional morphologies of SLA-Ti and SLA-TiCu surfaces. In Fig. 2(a), SLA-Ti surface showed classical SLA topographies, honeycomb-like holes with diameters ranging from 1 to 2 μm. In Fig. 2(b) a large number of streaks appeared on the SLA-TiCu surface marked by a red circle, with numerous round cavities with diameters varying from 3 to 5 μm, indicated by yellow stars, in addition to the honeycomb-like structure. The previous study confirmed that Widmannstatten structure of TiCu alloy consisted of α-Ti and ellipsoidal Ti₂Cu phase with long axis smaller than 1 μm [26]. The cross-sectional photography (Fig. 2b) revealed that the distributions of Ti₂Cu were multidirectional, parallel distribution, bulge pattern and inclined distribution shown in box 1,2 and 3 respectively, which were closely related to the cross-section morphology of the round cavities. Fig. 2(b₁ and b₂) were high magnified images of box 2 and box 3. Fig. 2(b₂) shows that the round cavities developed along the distribution direction of Ti₂Cu and at the bottom of these cavities there were visibly partial gaps between the Ti₂Cu and the matrix (marked by yellow arrow), which indicated that Ti₂Cu was about to separate from the matrix and the small Ti₂Cu particles with fuzzy boundary spread along both sides of these cavities. The EDS results (Table 3) also showed that the Cu content of these tiny particles (Point 3) was significantly lower than that of Ti₂Cu in the matrix (Point 2), which confirmed that the Ti₂Cu dissolved gradually after separating from the matrix, diffused outside along the circular cavities formed by acid etching, and finally developed the serrate inclined morphology.

Fig. 3(a) shows the contact angles on different surfaces, the contact angles of machined surfaces were less than 90°, while those of rough surfaces were higher than 90°, in other words, rough surfaces were hydrophobic. To examine the biosafety of the experimental materials, Cu²⁺ ions released from M-TiCu and SLA-TiCu surfaces were shown in Fig. 3(b). The Cu²⁺ ions from SLA-TiCu surface was higher and after 21 days reached 83.5 μg L^-1, while the Cu²⁺ ions from M-TiCu surface was very slow and reached 9.5 μg L^-1 after 21 days, which was far lower than that from SLA-TiCu surface.

Fig. 4 demonstrates the cell proliferation of MC3T3-El cells on all surfaces after 1, 4 and 7 days. The cell proliferation increased on all the surfaces during the experiment. After 4 days, the OD value of SLA-TiCu surface was lower than other surfaces, but the RGR was still higher than cytotoxicity limit 75%. While after 7 days, all surfaces favored the proliferation of MC3T3-El cells and there was no considerable difference.

The morphologies of MC3T3-El cells attached on different surfaces for 1 and 4 days was shown in Fig. 5 (a). For 1 day, the adhered cells on all surfaces presented a spindle-like morphology, but the cytoskeleton spreading area of MC3T3-El cells on rough surfaces was smaller than ground surfaces. Moreover, the cellular connections on SLA-TiCu and M-TiCu surfaces were closer than SLA-Ti and M-Ti surfaces, and the numbers of pseudopodia were considerably higher. The statistical results of adherent cell density on different surfaces also showed that the cell density on rough surfaces was less than that on ground surfaces (Fig. 5 (b)), but the cell density on SLA-TiCu surface was significantly higher than SLA-Ti surface. After 4 days, all surfaces were covered with cells and the adhered cell density showed no remarkable difference. The cells on M-Ti and M-TiCu surfaces spread along the grooved surface topographies, on the other hand, the cells on SLA-Ti and SLA-TiCu surfaces adapted to the rough surface topographies, where cells got together to facilitate the exchange of information which contributed to the signal transmission during osteogenic differentiation periods.

In addition to evaluating proliferation and attachment, apoptosis of MC3T3-E1 cells was further analyzed using flow cytometry (Fig. 6). Apoptosis is a normal physiological process, which is a selective death mechanism for cells to adapt to environments. As shown in Fig. 8, cells in the Q3 were alive, cells in the Q2 were in the end stage of apoptosis, undergoing necrosis or already dead, and cells in the Q4 were undergoing apoptosis. Hence, the apoptosis rate was used to reflect the cell adaptability to different material surfaces, the apoptosis rate on the rough surfaces was slightly higher than that of the ground surfaces, and under the same surface treatment, the TiCu alloy surfaces with slightly lower apoptosis were superior to Ti surfaces.

The osteogenic differentiation of cells at the implant/host interface determines the success of bone regeneration. ALP is an early osteogenic marker, which takes part in the regulation of the extracellular matrix calcification process and influences the osteoblast mineralization ability. Fig. 7(a) shows the ALP activity of MC3T3-E1 cells on different samples. After 4 days, the ALP activity of rough surfaces was remarkably higher than ground surfaces, after 7days the ALP activity further increased. Moreover, the ALP activity on SLA-TiCu surface was significantly higher than SLA-Ti surface. This result verified that rough surfaces promoted the osteogenic differentiation of MC3T3-E1 cells at early stage compared to ground surfaces. Among all surfaces, SLA-TiCu surface had the highest ALP activity. ECM mineralizations on various material surfaces were assessed for 14 and 21 days, respectively, which implied the late stage of osteogenic differentiation. As shown in Fig. 7(b), it was noticed that the ECM mineralization level on SLA-TiCu surface was far higher than other surfaces, in other words, SLA-TiCu also had the most substantial osteogenic effect during the late period. RT-qPCR analysis results were shown in Fig. 8, which clarified that SLA-Ti surface enhanced bone-related expression of ALP and COL I, compared to M-Ti surface, but there was no significant difference between two surfaces on the Run×2 and OPN gene expression. Excitingly, SLA-TiCu surface raised the expressions of Run×2 and OPN compared to SLA-Ti surface. In all samples, SLA-TiCu surface showed excellent bone-related gene expression of ALP, COL I, OPN and Run×2 of the MC3T3-E1 cells compared to the other surfaces.

S.aureus is one of the most common bacteria causing infection in orthopedics. Hence, it was used to evaluate the antibacterial activity of different surfaces. As shown in Fig. 9(a), compared to the control group, the antibacterial rates of M-TiCu and SLA-TiCu surfaces could reach up to 99.9% and 79.9%, which showed that SLA-TiCu surface had better antibacterial activity. However, the antibacterial rates of M-Ti and SLA-Ti surfaces were 36.6% and 23.9%, respectively. In order to further evaluate the viability of bacteria on different surfaces, LIVE/DEAD® BacLight™Bacterial Viability Kit was used, where alive bacteria could be dyed green with syto-9 and dead bacteria were stained red with PI, as shown in Fig. 9(b). The bacteria attached on the M-Ti and SLA-Ti surfaces were mainly alive and the red fluorescence on these surfaces was minute, suggesting that most of these bacteria attached to these two surfaces were viable. Conversely, most of the bacteria attached to M-TiCu and SLA-TiCu surfaces were dead, and the red fluorescence on SLA-TiCu demonstrated that most of the bacteria found (single or colonized) was dead. From the above results it is clear that SLA-TiCu surface had higher antibacterial properties.

In the present study, the rough surfaces of Ti and TiCu alloy were prepared by SLA treatment, which is a simple and effective surface modification technique for dental implants [36,37]. The results represented that S_a values of SLA-Ti and SLA-TiCu surfaces were 1.73 μm and 2.06 μm, which were consistent with the most active bone response rough range of implants, S_a: 1-2 μm [38]. The surface and cross-sectional morphologies were shown in Fig. 2. On the SLA-Ti surface, etched holes even existed in the side of the pits formed by sandblasting, which revealed typical SLA morphology. However, the surface topography of SLA-TiCu was lamellar with many circular cavities close to the size of Ti₂Cu, and the formation direction of these cavities is related to the distribution of Ti₂Cu. So what is the reason for such difference in the surface topographies?

Schematic diagrams demonstrated the formation of these two material morphologies during the acid etching process in Fig. 10. During the corrosion of pits formed by sandblasting on pure titanium surface (Fig. 10 (a₁)). It was hard for the matter diffusion in the holes, where H⁺ ions concentration decreased, Cl^- ions and other ions increased so that the difference of corrosion potential between inside and outside the holes could lead to crevice corrosion [39,40], finally the honeycomb-like morphology was formed in Fig. 10 (a). Besides, the etching process of TiCu alloy was described in Fig. 10 (b₁) and (b₂). In addition to crevice corrosion similar to pure titanium, the galvanic corrosion between Ti₂Cu and α-Ti phase also leads to the corrosion developing along with the distribution of Ti₂Cu phase. Since the Ti₂Cu intermetallics are nobler than both the α-Ti matrix [41], the α-Ti matrix phase dissolved preferentially around Ti₂Cu phase so that Ti₂Cu phase could separate from the matrix. Ti₂Cu phase also dissolved with smaller size and blurred the grain boundary as entering the solution, and the serrated morphology of SLA-TiCu was obtained at last, as shown in Fig. 10(b).

These surfaces possessed good cytocompatibility without cytotoxic effect on MC3T3-E1 cells viability (Fig. 4). At the early stage, one of the most interesting results is that the cytoskeleton spreading area and density of MC3T3-El cells on rough surfaces were less than ground surfaces (Fig. 5), as well as higher apoptosis rate (Fig. 6). These results suggested rough surfaces slightly decreased cell adhesion at the early stage, compared to the ground surfaces, which was likely due to the hydrophobic properties of rough surfaces (Fig. 3(a)). The similar results were also found in the study of early osteogenic cells response on different modified implant surfaces [42]. On the other hand, SLA-TiCu and M-TiCu surfaces had more pseudopodia and lower apoptosis rates than SLA-Ti and M-Ti surfaces respectively, which were mainly because TiCu alloy not only promoted the early cytocompatibility but also improved the response of rough surfaces on the initial cell adhesion [43,44]. Moreover, the differences between various surfaces gradually disappeared with culture time.

Another interesting result was that SLA-TiCu surface remarkably promoted the osteogenic differentiation, including bone-related gene expression (ALP, OPN, COL I and Run×2) and ECM mineralizations (Fig. 7, Fig. 8). The formation of bone extracellular matrix [45,46] has been classified into proliferation (I), matrix development and maturation (II) and mineralization (III), as shown in Fig. 11. Initially, during the period I, maximal levels of COL I mRNA are observed to support the formation of extracellular matrix mineralization, and the period of matrix maturation occurs. When ALP gene is maximally expressed and the expression of OPN gene gradually increases, OCN is closely related to the third period, which is an essential marker at the late differentiation stage. As a transcriptional activator of osteoblast differentiation, Run×2 [47] can activate the gene transcriptions and expressions of OCN, OPN, BSP and COL I. The present results suggested that SLA-Ti surface significantly promoted the expressions of ALP and COL I genes compared to M-Ti surface (P < 0.05), but there was no difference in the OPN gene expression and mineralization staining between these two surfaces, which demonstrated rough surfaces can only promote the osteogenesis differentiation at the early stage, but had no visible effect at the late stage, the osteogenic effect of rough surfaces (SLA) was represented by the orange arrow in Fig. 11. Marinucci had also confirmed the rough titanium surfaces increased secretion of TGFβ2 (a growth factor involved in differentiation and osteoblast proliferation). Moreover, rough surface also enhanced the expressions of COL I and BSP genes, but had no effect on OCN expression in the late osteogenesis stage, which also revealed that rough surface could promote the osteoblast differentiation in the initial stage [48]. Besides, Cu²⁺ ions have been reported to promote the osteogenic differentiation [49,50], as shown by red arrow in Fig. 11. Interestingly, in this study, it was found that SLA-TiCu surface not only enhanced the expressions of ALP and COL I in period I compared to SLA-Ti surface, but also significantly increased the content of ECM and the expression of OPN gene in the mineralization stage. In the cooperation of SLA surface and Cu²⁺ ions, SLA-TiCu surface possessed excellent osteogenic ability in all the stages of osteoblast differentiation.

Furthermore, the corporation between rough surface and Cu²⁺ ions also offered an excellent antibacterial activity for SLA-TiCu surface (Fig. 9). The quantitative and qualitative results showed that SLA-TiCu surface had more significant bactericidal effect on S.aureus compared to M-TiCu surface. The number of bacteria attached on SLA-Ti surface was more than M-Ti surface, which may be related to the high surface area [[51], [52], [53]]. Moreover, the release concentration of Cu²⁺ ions from SLA-TiCu surface was much higher than M-TiCu surface due to the rough treatment (Fig. 3(b)), which made SLA-TiCu surface better antibacterial performance than M-TiCu surface.

In the present work, rough surface of TiCu alloy with antibacterial and osteogenic ability was studied. SLA-TiCu surface showed no cytotoxicity to MC3T3-E1 cells and also inhibited cell apoptosis compared with SLA-Ti surface. Furthermore, the SLA-TiCu surface remarkably promoted the osteogenic differentiation compared to the other surfaces (M-Ti, M-TiCu and SLA-Ti) in the cell and gene level. Moreover, the SLA-TiCu surface also effectively inhibited S.aureus activity. Therefore, SLA-TiCu surface demonstrated multifunctional characteristics with excellent osteogenic ability and antibacterial activity, which makes it promising to be a novel implant material.

This work was financially supported by the National Key Research and Development Program of China (2018YFC1106601, 2016YFC1100601), LiaoNing Revitalization Talents Program (XLYC1807069), National Natural Science Foundation (No. 51631009, 31870954), and Key Projects for Foreign Cooperation of Bureau of International Cooperation Chinese Academy of Sciences(174321KYSB2018000). The authors are especially thankful for the technical support and assistance from Xun Qi and Zixuan Li (Key Laboratory of Diagnostic Imaging and Interventional Radiology of Liaoning Province, China Medical University, China).