Nano-topographic titanium modulates macrophage response in vitro and in an implant-associated rat infection model

Abstract

The macrophage-implant response plays a pivotal role in the interactions between implants and tissue involving inflammation and tissue remodeling. In this study, we investigated the proliferation, migration, and inflammatory cytokine secretion of macrophages adhering to titania nanotubes (TNT) and polished titanium (pTi) surfaces in a lipopolysaccharide (LPS)-simulated infection environment. An infection model in rats was used to analyze cell and bacterium responses to TNT and pTi implants in vivo. The in vitro results indicated that TNT surfaces restricted macrophage proliferation and migration and reduced pro-inflammatory cytokine secretion. Notably, LPS loading onto the TNT surface resulted in macrophage elongation with increased levels of secreted pro-inflammatory cytokines within 24 h followed by a decrease to the lowest level of all tested samples at 72 h. Analogously, increased amounts of inflammatory cytokines were observed for the TNT implants in vivo at 24 h with fewer detected at 72 h compared with pTi in the presence of Staphylococcus epidermidis (Se). Additionally, TNT implants exhibited lower total amounts of viable bacteria at 72 h than pTi implants. Furthermore, the tissues exhibited preferential spreading on TNT-Se implants at 72 h. These results suggested that the TNT surface in an infection environment could regulate the inflammatory response and promote tissue remodeling more effectively within the initial implantation compared with pTi. This study indicated the ability to construct functional nano-topographic surfaces by loading proteins or cytokines on implant surfaces that then could effectively modulate macrophages to support a healing process in lieu of chronic inflammation.

---

1. Introduction

The host-implant response determines the success rate of implant surgery.^1–3 However, whereas the majority of studies have focused on osteoblast or soft tissue cell responses to implant surfaces, few have elucidated the immune cell responses.^4,5 After implantation, immune cells are recruited to the implant site in a timely manner to phagocytose biomaterial fragments, bacteria, and dead cells to effect an immune defense function.^6,7 Macrophages have long been considered to be important immune effector cells that mediate inflammation by inducing a common wound healing response from the body toward the implant. Macrophages exhibit remarkable plasticity and either display predominantly inflammatory or anti-inflammatory activity, which in turn regulates tissue remodeling. Therefore, the regulation of macrophage function can be an effective strategy for reducing the risk of implant infection and improving long-term implant stability.

The functionality of macrophages can be altered by the physical and chemical properties of the implant surface, such as its surface topography and wettability.^4,8–10 For example, it has been shown that topography-induced changes can affect cell adhesion, cell morphology, and cytokine secretion in vitro.⁴ Furthermore, a hydrophilic implant surface was found to exhibit immunomodulating properties and dampened inflammatory macrophage activities by impeding the signaling pathways.¹¹ In particular, previous studies have suggested that the surfaces of nano-structured titanium, which has shown the potential for implantation owing to its excellent biocompatibility, were associated with a decrease in the inflammatory response of macrophages.^12–15 In addition, Lee et al. identified that macrophage movement was restricted on nanostructured compared with flat titanium surfaces. Moreover, the nanostructured titanium elicited the secretion of fewer pro-inflammatory enzyme molecules and cytokines as well as reduced nitric oxide production.⁵ Similarly, it was shown that the number of adherent macrophages on the nanotubular Ti surface was obviously reduced, up to 67% compared with conventional flat Ti.¹⁵

The initial interaction between implant surfaces and macrophages is noteworthy as the implanted materials inevitably cause an immune response of the host. Once the implant-associated infection occurs, bacteria at the infection sites then may affect the implant–macrophage interaction. In the present study, we created a pro-inflammatory environment by exposing the cells to the bacterial cell wall component lipopolysaccharide (LPS), which can be recognized by the innate immune system as a sign of infection.¹⁶ The pro-inflammatory environment was simulated by loading LPS onto the titania nanotube surfaces or adding LPS into the culture medium. The in vitro experiment examined the behavior of macrophages on nano- and flat titanium in standard and pro-inflammatory conditions. To further analyze the host-implant response in the early implantation period (over approximately 72 h), we then focused on the interactions between the host cells, bacteria, and nano-structured or polished titanium implants in soft tissues by evaluating the inflammation-associated cytokine changes and the surrounding tissue response to implants using an in vivo infection model in rats.

2. Experimental

2.1. Titanium sample preparation

Titanium discs (10 mm diameter, 0.5 mm thickness) were used for in vitro analysis and cylindrical titanium implants (2 mm diameter, 10 mm length) were utilized in the in vivo experiment. Titania nanotubes were prepared by anodization on titanium surfaces (TNT) and polished titanium, prepared by sanding and pickling, was used as the control (pTi). The method of sample preparation and the characterization of the samples were all as described in our previous study.¹⁷ The electrolyte was a miscible liquid of H₃PO₄ (2 M) and HF (0.15 M). The voltage was set at 20 V and the anodization was carried out for 1 h.

2.2. LPS loading and release

One group of TNT was immersed in LPS (from Escherichia coli, Sigma-Aldrich) solution (10 mL, 10 μg mL⁻¹) for 6 h at 37 °C to allow physical adsorption. After the prescribed time period, each sample was gently rinsed with endotoxin free water to remove unabsorbed LPS. Samples were then transferred to 24-well plates for release studies. The resultant sample of LPS loaded TNT was termed TNT-LPS.

The TNT-LPS sample was immersed in 10 mL endotoxin free water at 37 °C in an incubator shaker. After 4 h, 24 h, and 72 h, 1 mL release solution was removed for testing the cumulative release amount, and another 1 mL fresh endotoxin free water was added into the original solution to simulate the diluting effect of body fluid flow. This process was carried out for a total of 72 h to determine the LPS release dynamics. The solutions containing released LPS were periodically tested using the LAL Chromogenic Endotoxin Quantification kit (Thermo Fisher Scientific) per manufacturer instruction. LPS contamination was always below 0.1 EU per mL.

2.3. Macrophage culture

The murine macrophage-like RAW264.7 cell line was cultured in Dulbecco's modified Eagle medium supplemented with 10% fetal bovine serum (FBS; Hyclone) and 100 U mL⁻¹ penicillin and 100 μg mL⁻¹ streptomycin at 37 °C in a humidified atmosphere of 5% CO₂. The medium was refreshed every two days during cell culture. The cells were seeded at a density of 5 × 10⁴ cells per cm² onto TNT-LPS, TNT, and pTi surfaces in 24-well plates. After allowing the cell to adhere for 1 h, 200 ng mL⁻¹ LPS was added to one set of TNT and pTi groups to activate the macrophages. The following four groups of samples were used: TNT sample with original medium (T/D), TNT-LPS sample with original medium (TL/D), TNT sample with LPS-supplemented medium (T/DL), and pTi sample with LPS-supplemented medium (P/DL). The culture medium of all samples was refreshed after 4 and 24 h.

2.3.1. Cell proliferation and morphology. Cell proliferation was evaluated using the alamarBlue assay. After culture for 24 and 72 h, the culture medium in the 24-well plate was replaced by 300 μL medium 199 (GIBCO) supplemented with 10% FBS and 10% alamarBlue. The cells were incubated at 37 °C in a humidified atmosphere of 5% CO₂ for 4 h. The OD of the medium was read at 570 nm in a microplate reader (MQX200).

Cell morphology was observed under a fluorescence microscope (FM, Leica DM5500) and a scanning electron microscope (SEM, FEI Quanta 200) after 4, 24, and 72 h of culture. For observation, the medium was removed and the cells were fixed in a solution containing 2.5% glutaraldehyde for 4 h. For FM observation, the cells were stained with FITC–phalloidin (1 : 40 in 5% FBS/0.1 Triton X-100 in phosphate buffered saline (PBS), Sigma) for 30 min to visualize F-actin filaments as well as with 4,6-diamidino-2-phenylindole (DAPI, 1 : 30 in PBS, Sigma) for 5 min to visualize the cell nucleus. For SEM, the cells were dehydrated sequentially in a series of ethanol solutions, then critically point dried and gold-sputtered prior to observation.

2.3.2. NO production and pro-inflammatory cytokine expression. After 24 and 72 h, the cell culture supernatant was collected to evaluate the nitric oxide (NO) and cytokine release generated by the activated macrophages adhering on the samples. The production of NO was measured with a Nitric Oxide Assay Kit (Jiancheng Biotech) according to manufacturer instruction. The release of cytokines (TNF-α, IL-1β, and IL-6) was investigated using ELISA kits (R&D Systems).

2.3.3. Cytokine/chemokine gene expression. Total RNA was extracted (TRIzol, Invitrogen) after 4, 24, and 72 h and reverse transcription was carried out using a PrimeScript RT reagent Kit (TaKaRa Bio). Quantitative analysis of mRNA expression was performed with a PIKORed 96 real-time fluorescence quantitative polymerase chain reaction (RT-PCR, Thermo Fisher) system. The design of primers for inflammatory cytokines and chemokines was performed using Primer Premier web-based software. PCR was carried out with the following primers: monocyte chemoattractant protein-1 (Mcp1), (forward (fwd): 5′-CCA CTC ACC TGC TGC TAC TCA TTC A-3′, reverse (rev): 5′-GTT CAC TGT CAC ACT GGT CAC TCC T-3′), tumor necrosis factor-α (Tnfa, fwd: 5′-CGG TGC CTA TGT CTC AGC CTC TTC TC-3′, rev: 5′-TGG TGG TTT GTG AGT GTG AGG GTC TG-3′), interleukin-1β (Il1b, fwd: 5′-TGC ATA CAG GCT CCG AGA TGA ACA-3′, rev: 5′-TGC TCT GCT TGT GAG GTG CTG ATG T-3′), Il6 (fwd: 5′-TGG AGT CAC AGA AGG AGT GGC TAA GG-3′, rev: 5′-GCA TAA CGC ACT AGG TTT GCC GAG TA-3′), Il10 (fwd: 5′-TCC TAG AGC TGC GGA CTG CCT TCA-3′, rev: 5′-ACT CTT CAC CTG CTC CAC TGC TTG-3′), and transforming growth factor-β1 (Tgfb1, fwd: 5′-GGT GGA CCG CAA CAA CGC CAT CTA-3′, rev: 5′-TGG TTC AGC CAC TGC CGT ACA ACT-3′). β-Actin (fwd: 5′-GAA GAT CAA GAT CAT TGC TCC T-3′, rev: 5′-TAC TCC TGC TTG CTG ATC CA-3′) was used to verify the equal amounts of RNA used for amplification. The relative mRNA expression level was calculated by threshold cycle method using the value of 2^−ΔΔCT. The CT value for each test sample was analyzed by Sequence Detection software version 1.2.3 (Applied Biosystems).

2.4. Percutaneous implantation model

We utilized 12 Sprague-Dawley rats (180–220 g) fed on a standard pellet diet and water in the study. The Institutional Animal Ethics Committee of the Southwest Jiaotong University approved all animal experimental protocols and the experiments were conducted in accordance with the guidelines of Committee on the Use of Live Animals in Teaching & Research, Government of China, for animal care and experimentation. The rats were anesthetized by an intravenous injection of 3% pentobarbital sodium with a dose of 1 mL kg⁻¹ body weight and then the back was shaved and cleaned with ethanol and iodine prior to surgery. For implantation, 6 separate incisions were made in the dorsal surface and pockets were created by blunt dissection in the soft tissue under the skin. Of these, 4 pockets received implants (TNT and pTi) leaving approximately 1–2 mm of implant exposed; the other two pockets were sham sites (no implant). Next, 0.9% saline containing S. epidermidis (10⁶ CFU per mL) was injected in each pocket prior to closure on the left side. The right side was only injected with 0.9% saline, which was kept as the control. The wounds were then carefully closed. After 4, 24, and 72 h, the animals were sacrificed with an overdose of pentobarbital sodium. The back of each rat was cleaned with ethanol/iodine and all the sutures were removed. The implants or implant with attached soft tissue were retrieved and the exudates were obtained from the pockets, diluted with PBS, and stored at 4 °C.

2.4.1. Bacteria colony counting. The implants were placed in tubes with 1 mL PBS, vortexed at maximum speed for 1 min, and subjected to ultrasonic treatment for 30 s to dislodge adherent bacteria and disrupt aggregates. The exudate samples were vortexed at maximum speed for 1 min. The suspensions were then serially diluted in saline and 100 μL from each dilution was spread onto a nutrient agar plate, incubated at 37 °C for 24 h, and the number of resultant bacterial colonies counted.

2.4.2. Cytokines expression. The exudates were centrifuged for 5 min and the supernatant was diluted in saline. The diluents were investigated using ELISA kits (R&D Systems) for 3 cytokines: TNF-α, IL-1β, and IL-6. Briefly, the diluents were incubated with antibody-coupled wells for 2 h, followed by incubation of cytokine-bound wells with secondary antibody for 2 h. Then the substrate solution was added to each well; after 20 minutes, addition of the stop solution resulted in a change in color in the wells from blue to yellow. Cytokine concentrations (pg mL⁻¹) in triplicates were determined from the mean fluorescence intensity readings using standard curves.

2.4.3. Histology staining and SEM observation of the implant surface. The retrieved implants after 4, 24, and 72 h of implantation with or without S. epidermidis were fixed in 2.5% glutaraldehyde for 4 h, then dehydrated in a graded series of ethanol and dried. The morphologies of the cells and matrix on the implant surfaces were visualized by SEM after sputtering with palladium. The implants were removed and the surrounding tissue was formalin-fixed and paraffin-embedded. Sections with a thickness of 5 μm were stained with routine hematoxylin and eosin (H&E) and May-Grünewald Giemsa. All the sections were observed using a light microscope.

2.5. Statistical analysis

All the results are presented as the means ± standard error of the mean (SEM). The data were statistically analyzed using the software program SPSS (IBM). Statistical analysis was performed using one-way ANOVA. Significance was indicated at p values < 0.05, 0.01, and 0.001. Data from all experiments are representative from one of at least three repeated experiments.

3. Results

3.1. LPS release

All samples were characterized in terms of morphology, surface roughness, and wettability. Briefly, the TNT used in this work was 100 nm in diameter. The TNT surface exhibited a higher surface roughness and wettability than the pTi surface.¹⁷ Furthermore, the LPS-modified TNT surface also presented good hydrophilicity (Fig. S1†).

Fig. 1 shows the cumulative and single point release of LPS within 72 h. The cumulative release profile of LPS exhibited a sustained release without burst release fitting the zero-order kinetics (Fig. 1a). The single point release profile revealed that the LPS released from TL/D surface was 210 ng at 4 h within 0 to 4 h, 180 ng at 24 h within 4 to 24 h, and 220 ng at 72 h within 24 to 72 h, respectively. The amount of the released LPS at each time point approximated that added in the culture medium of T/DL.

Fig. 1 Cumulative and single point release of LPS from TNT.

3.2. Macrophage proliferation and morphology

As high release amounts of LPS would be expected to notably affect the behaviors of macrophages, we tested the proliferation of RAW 264.7 cells at 24 and 72 h post-seeding using the alarm Blue assay (Fig. 2a). T/D and P/DL exhibited the highest cell densities at 24 and 72 h, respectively. The cell density on TL/D was lower than that on T/DL at each time point although no significant difference was determined. TL/D, T/DL, and P/DL showed an increase in the number of viable cells with incubation time from 24 to 72 h, whereas the number on TNT without LPS (T/D) increased only marginally. To assess cellular morphology, FM visualization (Fig. 2b) of the actin cytoskeleton and the nucleus and SEM observation (Fig. 2c) were performed. The FM results showed that macrophages adhering to the T/D surface remained round-shaped whereas an elongated shape was observed on the other three surfaces after 24 h of culture (Fig. 2b). Magnified SEM images provided evidence of macrophage activation on the TL/D and T/DL surfaces (Fig. 2c), wherein cells displayed a higher degree of spreading to a spindle-shaped morphology after 72 h. This finding was consistent with an activated and migratory pro-inflammatory phenotype and was most noticeable on the T/DL surface at 72 h post-seeding.

Fig. 2 (a) RAW 264.7 macrophage adhesion on T/D, TL/D, T/DL, and P/DL after 24 and 72 h. Data represent the means ± SEM, N = 4, *p < 0.05; (b) fluorescent images of macrophages on T/D, TL/D, T/DL, and P/DL after 24 h; cells were stained with FITC–phalloidin (green) and DAPI (blue). (c) SEM images of macrophages on T/D, TL/D, T/DL, and P/DL after 72 h.

3.3. NO production and inflammatory cytokine expression

NO is a prominent indicator of pro-inflammatory signal transduction during inflammatory response and antimicrobial defense.¹⁸ In the current study, NO was examined by measuring the level of nitrite accumulation in the cell culture media (Fig. 3a). In the absence of LPS (Fig. 3a T/D), NO production from macrophages on the TNT surface was decreased from 24 to 72 h and the lowest value among the samples was identified at 72 h. When LPS was loaded on the TNT surface (Fig. 3a TL/D), NO production was the lowest at 24 h and increased from 24 to 72 h, although the production was significantly lower as compared to that from T/DL at 72 h. LPS added to the culture medium resulted in more NO production after 72 h on both TNT and pTi surfaces (Fig. 3a T/DL and P/DL), with higher production from TNT.

Fig. 3 (a) NO production at 24 and 72 h and (b) TNF-α, IL-1β, and IL-6 secreted by macrophages at 72 h in response to incubation with T/D,TL/D, T/DL, and P/DL. Data represent the means ± SEM, N = 4, *p < 0.05.

Furthermore, the concentrations of three pro-inflammatory cytokines were measured after 72 h of culture (Fig. 3b). The levels of TNF-α and IL-1β secreted by macrophages on the TNT surface without LPS (Fig. 3b T/D) were significantly lower compared to the TNT and pTi surfaces in the presence of LPS (Fig. 3b TL/D, T/DL, and P/DL). T/DL yielded more cytokine release from macrophages than T/LD. For IL-6, T/D exhibited the lowest cytokine release, whereas T/DL released more IL-6 than TL/D and P/DL demonstrated the highest cytokine release.

The gene expression of MCP-1, TNF-α, IL-1β, IL-6, IL-10, and TGF-β1 was assessed by PCR after 4, 24, and 72 h (Fig. 4). The pro-inflammatory cytokine TNF-α and IL-1β and chemokine MCP-1 genes exhibited similar expression profiles. For the three cytokines, the mRNA levels of T/D was maximal at 4 h and progressively decreased from 4 to 72 h. TL/D showed an approximately linear increase in mRNA levels between 4 and 24 h, followed by a decrease phase from 24 to 72 h. In contrast, the opposite tendency was observed for T/DL. For P/DL, the mRNA levels of TNF-α and IL-1β increased from 4 to 72 h whereas those of MCP-1 increased between 4 and 24 h and then decreased until 72 h. The mRNA level of the pro-inflammatory cytokine IL-6 for TL/D, T/DL, and P/DL decreased between 4 and 24 h and then increased, whereas the Il6 mRNA level of T/D was stable.

Fig. 4 Relative mRNA expression of TNF-α, IL-1β, MCP-1, IL-6, IL-10, and TGF-β1 secreted by macrophages at 4, 24, and 72 h of culture on T/D, TL/D, T/DL, and P/DL.

Results from the prominent anti-inflammatory cytokines IL-10 and TGF-β1 indicated that the mRNA level of IL-10 for T/D and TL/D first increased and then decreased whereas the mRNA level of T/DL remained approximately constant between 4 and 24 h and then increased from 24 to 72 h. P/DL exhibited a continuous increase of Il10 mRNA level from 4 to 72 h. The T/D mRNA level for TGF-β1 was almost stable, whereas the TGF-β1 expression profile shared the same trend as IL-10 expression for the other three samples.

3.4. Inflammatory cytokine secretion in vivo

The cytokine secretion of inflammatory cells in the exudate surrounding the implant was measured at 4, 24, and 72 h (Fig. 5). At all time points, the secretion of TNF-α, IL-1β, and IL-6 was higher in infected implant sites than in non-infected sites. The secretion of TNF-α was observably decreased in the infected TNT implant (TNT-Se) at 72 h, whereas a large amount of TNF-α was consistently secreted in pTi-Se. Infected TNT sites secreted lower IL-1β compared with pTi-Se at 4 h, exhibited 3-fold enhanced secretion at 24 h, then significantly decreased again at 72 h. From 4 to 72 h, the level of IL-1β was reduced in infected pTi sites. The IL-6 secretion profile shared the same trend as observed for IL-1β secretion.

Fig. 5 Cytokine secretion of TNF-α, IL-1β, and IL-6 in the exudate around TNT, pTi, and sham sites in the presence or absence of inoculated Staphylococcus epidermidis (Se) at 4, 24, and 72 h after surgery. Data represent the means ± SEM, N = 4, *p < 0.05.

3.5. Colonization of S. epidermidis in vivo

Next, we examined the bacteria adhering to the implants and existing in the exudates following implantation in the presence or absence of inoculated S. epidermidis to explore whether bacteria preferentially located on the implant surfaces in vivo (Fig. 6). The number of viable bacteria on the infected implant surfaces was low at 4 h, with a subsequent significant increase until 24 h and a decrease by 72 h whereupon more bacteria were retrieved from pTi-Se than TNT-Se (Fig. 6a). The numbers of implant-adherent bacteria were fewer in non-infected sites. The number of viable bacteria in the exudates showed a similar trend (Fig. 6b). Fig. 6c shows the total viable counts on the implants and in the exudates, indicating lesser numbers on and around TNT-Se compared with pTi-Se after 24 h.

Fig. 6 Counts of bacteria adherent to the implants (a), present in the exudates (b), and the total viable counts (c) at 4, 24, and 72 h after implantation of TNT and pTi. Data represent the means ± SEM, N = 4, *p < 0.05.

3.6. SEM observations of the implant surfaces

The proliferative phase of wound healing involves cellular proliferation, angiogenesis, new extracellular matrix (ECM) deposition, and the formation of granulation tissue. The SEM images (Fig. 7) show that the implants have already been covered by tissue cells and ECM at the point of observation. In the absence of S. epidermidis (Fig. 7a), considerably more tissue cells attachment and ECM deposition on the TNT implants was observed than on pTi at 4 and 24 h; conversely, a thin layer of tissue cells and ECM were found on the TNT at 72 h. At the infected sites (Fig. 7b), a large quantity of tissue cells had spread on the TNT implant surface, whereas some bacteria-like granules existed along with the tissue cells at 24 h although these had almost disappeared at 72 h; these findings were consistent with the bacteria count results (Fig. 6a). For the pTi implant, tissue cells were seen to have spread and gathered to form a membrane on the surface at 4 and 24 h, with bacteria appearing on the surface at 72 h.

Fig. 7 SEM images of the implant surfaces in the absence (a) or presence (b) of inoculated Staphylococcus epidermidis (Se) at 4, 24, and 72 h after implantation.

3.7. Histology

The soft tissues around the percutaneous implants were analyzed by H&E (Fig. 8a and b) and May-Grünewald Giemsa (Fig. 8c) staining. In absence of S. epidermidis (Fig. 8a), blood vessels were observed across the tissue for TNT and pTi at 4 h. The number of inflammatory cells increased considerably at 24 h and more inflammatory cells were observed for the TNT implants compared with pTi at 72 h. The presence of S. epidermidis induced a larger quantity of inflammatory cells in the tissue at 24 h for TNT-Se as well as fewer blood vessels compared with non-infected sites (Fig. 8b). The proportion of inflammatory cells for TNT-Se was higher at 24 h compared with pTi-Se whereas the inflammatory cells were markedly reduced for TNT-Se at 72 h. No obvious differences were detected for pTi-Se at 4, 24, and 72 h after implantation. The presence of bacteria is illustrated in Fig. 8c, showing the presence of some clusters of bacteria distributed in the tissue for TNT-Se and pTi-Se at 24 h.

Fig. 8 H&E stained images of tissue morphology in the absence (a) or presence (b) of inoculated Staphylococcus epidermidis (Se) at 4, 24, and 72 h after implantation. (: inflammatory cells, : blood vessel); May-Grünewald Giemsa stain images (c) at 24 h (: bacteria, blue cocci of approximately 1 mm).

4. Discussion

The present in vitro experiments illustrated the different inflammatory responses of macrophages on nano- or flat titanium surfaces in the absence or presence of LPS. Previously, it was revealed that a hydrophilic surface inhibited the adhesion of macrophages; this might be due to a lack of the integrin binding sites necessary for macrophage adherence and spreading.¹⁹ Consistent with these results, our study demonstrated that titania nanotubes presented good wettability and that the TNT surface restricted the movement and proliferation of macrophages (Fig. 2 T/D). Furthermore, the TNT surface inhibited the inflammatory cytokine/chemokine gene expression/protein secretion of macrophages (Fig. 3 and 4 T/D), indicating that the nano-structured titanium surface restrained the inflammatory activity of macrophages. It is conceivable that the nano-topography might affect some stages of the signal pathways during macrophage activation and migration such as PI3K/Akt, MAPK, and FAK.^5,20 For example, in a study on FAK activation in macrophages, FAK phosphorylation (indicative of FAK activation) declined on the nano-structured titanium surfaces, which reflected restricted macrophage migration on the nano-structured surfaces as compared to the flat ones.⁵

It is known that LPS induces macrophage activation by recognizing the Toll-like receptor of macrophages and activating NF-κB, which stimulates macrophages to release inflammatory cytokines.²¹ The results of our in vitro analyses indicated that LPS also stimulated the elongation and proliferation of macrophages on the TNT surface (Fig. 2 TL/D and T/DL). However, macrophages on the pTi surface showed less spreading in the inflammatory environment at 72 h, although they exhibited the highest degree of cell adhesion (Fig. 2 P/DL). Notably, it has been established that cell proliferation and motility can be mediated by cytoskeletal rearrangement.²² Additionally, MCP-1 binding to its receptor triggers cytoskeletal rearrangement and the activation of MAP kinase and FAK,^23,24 further inducing macrophages activation and migration. Therefore, the higher degree of elongate morphology observed at 72 h likely resulted from the higher Mcp1 gene expression exhibited by the macrophages on the T/DL surface (Fig. 2 and 4).

Macrophages grown on the different surfaces displayed different tendencies for cytokine expression in the inflammatory environment simulated with LPS. Higher levels of pro-inflammatory cytokines, i.e., NO, TNF-α, IL-1β, and IL-6, were produced by adding LPS in the culture medium than by loading LPS on the TNT surface (Fig. 3 TL/D and T/DL). The TNT surface modified by LPS facilitated the inflammatory cytokine/chemokine expression of macrophages within 24 h but suppressed the expression from 24 to 72 h. However, the macrophages on the TNT surface with LPS-supplemented medium exhibited a completely opposite expression tendency. Notably, whereas a higher expression of inflammatory cytokines favors the elimination of inflammation in the early response period (about 24 h), continued excessive expression has been shown to lead to tissue damage.^25,26 Conversely, the anti-inflammatory cytokines can regulate macrophages switching from a pro-inflammatory phenotype to a wounding-healing phenotype, thereby reducing the secretion of pro-inflammatory cytokines.²⁷ In our study, the mRNA level of IL-6 mediated by pTi (Fig. 4, P/DL) was much higher than that for the three TNT samples at 72 h. As IL-6, a crucial cytokine in the acute phase response, dictates the change from acute to chronic inflammation by altering the nature of leukocyte infiltration,²⁸ this finding indicated that the flat titanium surface might induce chronic inflammation under conditions of LPS stimulation.

For TL/D, macrophages were activated by LPS loaded on the TNT surface, initially increasing the expression of inflammatory cytokines (except IL-6). As LPS was released from the surface, the nano-topography began to affect macrophage activation, which led to a subsequent decrease of inflammatory cytokine (except IL-6) expression. For T/DL, nano-topography initially restricted macrophage activation, as LPS was added to the medium, macrophages were activated and released more pro-inflammatory cytokines. Furthermore, proteins and LPS in the medium were also commonly or competitively absorbed on the TNT surface. In contrast to this situation, LPS loaded on the TNT surface occupied most of the binding sites of the material surface, which might have obstructed subsequent protein adsorption to a certain extent. The difference in the behaviors of protein absorption might therefore affect the inflammatory activity of macrophages on the TL/D and T/DL surfaces.

An ideal implant would be expected to regulate the inflammatory function of immune cells and rapidly eliminate the inflammation. The nano-structured titanium surface examined in this study inhibited the migration and inflammatory factor secretion of macrophages. However, LPS on the TNT surface (TL/D) stimulated the adhesion and migration of macrophages and significantly increased the secretion of inflammatory cytokines within 24 h, owing to the anti-inflammatory response. Subsequently, the pro-inflammatory cytokines of TL/D were decreased from 24 to 72 h, which suggested that the macrophages had switched to a wounding-healing phenotype by 72 h. Generally, the wounding-healing polarization state is characterized by little to no secretion of pro-inflammatory cytokines, thus promoting tissue remodeling and repair.²⁹ These results indicated that both LPS and the nano-structure on the surface were able to regulate an effective switch from a pro-inflammatory to a wounding-healing phenotype.

The nano-topography of titanium surfaces influences the response of tissue cells as well as that of macrophages in vitro and in vivo. Accordingly, in this study a rat percutaneous implantation model was applied to investigate the immune and tissue cell response to the TNT and pTi implant surfaces in vivo. In the absence of S. epidermidis, the inflammatory response to the TNT implant continued until 72 h, while the response reached its termination for pTi at this time point (Fig. 8a). The continuing inflammatory response resulted in a smaller number of tissue cells on the TNT implant surface than the pTi surface at 72 h. The TNT implant also exhibited higher pro-inflammatory cytokine secretion than pTi at 72 h (Fig. 5 TNT, pTi). After implantation, the acute inflammation phase, which is dominated by macrophages and neutrophils, inevitably occurred within 72–96 h.³⁰ The in vivo study revealed that the TNT implant triggered a weak but continuous inflammatory response within 72 h, resulting in poor tissue adhesion. This might have been due to the nano-structured titanium restricting the integration with macrophages, affecting the activation and immune modulatory property of the macrophages. However, a previous study noted that nanotubular titanium implants exhibited greater cellular inhabitation and tissue ingrowth over longer implantation times (3 weeks).³¹ Additionally, our in vivo study revealed that there were more tissue cells spread on the TNT implant surface at 7 d than on the pTi surface in the absence of S. epidermidis (Fig. 9). Together, these findings suggested that an inhibition of the macrophage activation was likely disadvantageous for the host-implant response within the initial 72 h. As the functionality of macrophage was restricted, the inflammatory response was not likely able to be weakened in a timely manner and the process of tissue remodeling was delayed. These results suggest that macrophage activation via an appropriate inflammation response is conducive for opportune tissue remodeling.

Fig. 9 SEM images of the implant surfaces in the absence of inoculated Staphylococcus epidermidis at 7 d after implantation.

In the presence of S. epidermidis, the interaction between bacteria and implant surfaces is a critical event in the pathogenesis of infections. Notably, it has been shown that only a low dose of inoculum is sufficient to result in the infection at an implant site. The inoculated bacteria attach onto the surface, synthesize extracellular polymeric substances (EPS), and aggregate into micro-colonies.³² In such situations, tissue cells are not able to displace these primary colonizers. In addition, EPS can protect bacteria from antimicrobial agents and the host immune system. Thus, the inhibition of bacterial adhesion is often regarded as a critical step to prevent implant-associated infection. However, surface topography and the properties of titanium implants are also considered important and determinant factors in bacteria attachment on implant surfaces. In the current study, TNT-Se exhibited a higher viable bacteria level than pTi-Se at 4 h followed by a significant decrease of viable bacteria from 24 to 72 h (Fig. 6 TNT-Se, pTi-Se). One possible explanation is that the TNT implant with a higher surface roughness and higher surface energy was able to recruit not only bacteria but also macrophages and other cytokines in the in vivo infection environment, and that these macrophages were able to prevent the subsequent reproduction and adhesion of bacteria by secreting pro-inflammatory cytokines, resulting in eventual bacteria elimination.

In our study, S. epidermidis induced strong pro-inflammatory cytokine secretion. The inflammatory cytokine secretion for TNT-Se was the highest at 24 h and decreased at 72 h compared with pTi-Se and the control site (Fig. 5), displaying a favorable inflammation-regulatory function of TNT in the infection environment. The result of H&E staining also indicated that the inflammatory response for TNT-Se was most intense at 24 h but had ended at 72 h (Fig. 8b), and that the tissue cells were preferentially spread on TNT-Se surfaces at 72 h (Fig. 7b). Correspondingly, the in vitro study showed that TL/D stimulated the migration of macrophages and that the production of pro-inflammatory cytokines was increased between 4 and 24 h and decreased from 24 to 72 h (Fig. 4. TL/D). This revealed that nano-structured titanium in the presence of a pathogen stimulus was able to immediately activate macrophages to destroy potential pathogens after an appropriate inflammatory reaction and then weaken the inflammatory response within 72 h, which was beneficial for wound healing. This is consistent with the result obtained in our previous study for a prolonged implantation time, which indicated that the TNT implant significantly inhibited the infection risk and enhanced tissue integration of the implant compared to pTi in an infection environment.¹⁷

5. Conclusions

The in vitro study without inflammatory stimuli demonstrated that the macrophages on nano-topographic titanium surfaces were restricted in proliferation, migration, and pro-inflammatory cytokine expression. LPS loaded on the TNT surface caused the macrophages to begin to migrate, with pro-inflammatory cytokine secretion increasing at 24 h and then decreasing at 72 h, suggesting that the LPS-modified TNT surface could effectively regulate macrophages function for both the inflammatory response and tissue remodeling within 72 h. In the absence of S. epidermidis, the nano-structured titanium implant surface presented a higher inflammatory response and poorer tissue growth compared with the polished implant surface at 72 h. The continuous inflammatory response and impaired tissue remodeling within 72 h might be ascribed to the repressed functionality of macrophages. Conversely, in the presence of S. epidermidis the pro-inflammatory cytokines of the TNT implant were increased, resulting in destruction of the bacteria colony within 24 h; subsequently, the total amount of viable bacteria and secretion of pro-inflammatory cytokines were reduced from 24 to 72 h, accompanied by marked spreading of the tissue cells on the TNT implant. These findings indicated that nano-topography titanium implants could modulate macrophage behaviors to support a healing process in lieu of chronic inflammation in an infection environment. In turn, this indicated that appropriate inflammation stimuli effected macrophage activation, which is conducive for the regulation of inflammatory response and tissue remodeling upon initial implantation. As the mechanism of LPS-modified TNT modulation of the macrophage response has not yet been fully elucidated, our ongoing work involves exploration of the signaling pathways of macrophage-modified TNT interaction.

Acknowledgements

This work was supported by the National Basic Research Program of China (973 Program, 2012CB933600), Natural Science Foundation of China (31570955), Applied Basic Research Programs of Sichuan Province, China (2015JY0036).

Notes and references

1. B. N. Brown and S. F. Badylak, Expanded applications, shifting paradigms and an improved understanding of host-biomaterial interactions, Acta Biomater., 2013, 9(2), 4948–4955.
2. D. Campoccia, L. Montanaro and C. R. Arciola, A review of the clinical implications of anti-infective biomaterials and infection-resistant surfaces, Biomaterials, 2013, 34(33), 8018–8029.
3. S. Svensson, M. Trobos, M. Hoffman, B. Norlindh, S. Petronis, J. Lausmaa, F. Suska and P. Thomsen, A novel soft tissue model for biomaterial-associated infection and inflammation – bacteriological, morphological and molecular observations, Biomaterials, 2015, 41, 106–121.
4. S. Chen, J. A. Jones, Y. Xu, H. Y. Low, J. M. Anderson and K. W. Leong, Characterization of topographical effects on macrophage behavior in a foreign body response model, Biomaterials, 2010, 31(13), 3479–3491.
5. S. Lee, J. Choi, S. Shin, Y. M. Im, J. Song, S. S. Kang, T. H. Nam, T. J. Webster, S. H. Kim and D. Khang, Analysis on migration and activation of live macrophages on transparent flat and nanostructured titanium, Acta Biomater., 2011, 7(5), 2337–2344.
6. J. M. Anderson, Inflammatory response to implants, ASAIO Trans., 1988, 34(2), 101–107.
7. R. S. Labow, E. Meek and J. P. Santerre, Neutrophil-mediated biodegradation of medical implant materials, J. Cell. Physiol., 2001, 186(1), 95–103.
8. S. Hamlet, M. Alfarsi, R. George and S. Ivanovski, The effect of hydrophilic titanium surface modification on macrophage inflammatory cytokine gene expression, Clin. Oral Implants Res., 2012, 23(5), 584–590.
9. M. A. Alfarsi, S. M. Hamlet and S. Ivanovski, Titanium surface hydrophilicity modulates the human macrophage inflammatory cytokine response, J. Biomed. Mater. Res., Part A, 2014, 102, 60–67.
10. K. A. Barth, J. D. Waterfield and D. M. Brunette, The effect of surface roughness on RAW 264.7 macrophage phenotype, J. Biomed. Mater. Res., Part A, 2013, 101(9), 2679–2688.
11. S. Franz, F. Allenstein, J. Kajahn, I. Forstreuter, V. Hintze, S. Möller and J. C. Simon, Artificial extracellular matrices composed of collagen I and high-sulfated hyaluronan promote phenotypic and functional modulation of human pro-inflammatory M1 macrophages, Acta Biomater., 2013, 9(3), 5621–5629.
12. K. M. Ainslie, S. L. Tao, K. C. Popat, H. Daniels, V. Hardev and C. A. Grimes, et al., In vitro inflammatory response of nanostructured titania, silicon oxide, and polycaprolactone, J. Biomed. Mater. Res., Part A, 2009, 91, 647–655.
13. M. L. Chamberlain, S. K. Brammer, W. G. Johnston, S. Chien and S. Jin, Macrophage inflammatory response to TiO₂ nanotube surfaces, J. Biomater. Nanobiotechnol., 2011, 2, 293–300.
14. B. S. Smith, P. Capellato, S. Kelley, M. Gonzalez-Juarrero and K. C. Popat, Reduced in vitro immune response on titania nanotube arrays compared to titanium surface, Biomater. Sci., 2013, 1, 322–332.
15. A. Rajyalakshmi, B. Ercan, K. Balasubramanian and T. J. Webster, Reduced adhesion of macrophages on anodized titanium with select nanotube surface features, Int. J. Nanomed., 2011, 6, 1765–1771.
16. M. J. Fenton and D. T. Golenbock, LPS-binding proteins and receptors, J. Leukocyte Biol., 1998, 64, 25–32.
17. J. Tan, Y. T. Li, Z. Y. Liu, S. X. Qu, X. Lu, J. X. Wang, J. Weng and F. Bo, Anti-infection activity of nanostructured titanium percutaneous implants with a postoperative infection model, Appl. Surf. Sci., 2015, 344, 119–127.
18. R. Radi, Nitric oxide, oxidants, and protein tyrosine nitration, Proc. Natl. Acad. Sci. U. S. A., 2004, 101(12), 4003–4008.
19. B. G. Keselowsky, D. M. Collard and A. J. Garcia, Integrin binding specificity regulates biomaterial surface chemistry effects on cell differentiation, Proc. Natl. Acad. Sci. U. S. A., 2005, 102(17), 5953–5957.
20. S. Heydarkhan-Hagvall, C. H. Choi, J. Dunn, S. Heydarkhan, K. Schenke-Layland, W. R. MacLellan and R. E. Beygui, Influence of systematically varied nanoscale topography on cell morphology and adhesion, Cell Commun. Adhes., 2007, 14(5), 181–194.
21. G. Trinchieri and A. Sher, Cooperation of Toll-like receptor signals in innate immune defence, Nat. Rev. Immunol., 2007, 7, 179.
22. E. T. den Braber, J. E. de Ruijter, L. A. Ginsel, A. F. von Rectum and J. A. Jansen, Orientation of ECM protein deposition, fibroblast cytoskeleton, and attachment complex components on silicone microgrooved surfaces, J. Biomed. Mater. Res., 1998, 40, 291–300.
23. T. R. Kyriakides, M. J. Foster, G. E. Keeney, A. Tsai, C. M. Giachelli and I. Clark-Lewis, et al., The CC chemokine ligand, CCL2/MCP1, participates in macrophage fusion and foreign body giant cell formation, Am. J. Pathol., 2004, 165(6), 2157–2166.
24. A. Mantovani, A. Sica, S. Sozzani, P. Allavena, A. Vecchi and M. Locati, The chemokine system in diverse forms of macrophage activation and polarization, Trends Immunol., 2004, 25(12), 677–686.
25. R. D. Stout, C. Jiang, B. Matta, I. Tietzel, S. K. Watkins and J. Suttles, Macrophages sequentially change their functional phenotype in response to changes in microenvironmental influences, J. Immunol., 2005, 175(1), 342–349.
26. T. A. Wynn, A. Chawla and J. W. Pollard, Macrophage biology in development, homeostasis and disease, Nature, 2013, 496(7446), 445–455.
27. D. F. Fiorentino, A. Zlotnik, T. R. Mosmann, M. Howard and A. O'Garra, IL-10 inhibits cytokine production by activated macrophages, J. Immunol., 1991, 147(11), 3815–3822.
28. C. Gabay, Interleukin-6 and chronic inflammation, Arthritis Res. Ther., 2006, 8(2), S3.
29. M. Alfano, F. Graziano, L. Genovese and G. Poli, Macrophage polarization at the crossroad between HIV-1 infection and cancer development, Arterioscler., Thromb., Vasc. Biol., 2013, 33(6), 1145–1152.
30. B. N. Brown, B. D. Ratner, S. B. Goodman, S. Amar and S. F. Badylak, Macrophage polarization: an opportunity for improved outcomes in biomaterials and regenerative medicine, Biomaterials, 2012, 33(15), 3792–3802.
31. B. J. Farrell, B. I. Prilutsky, J. M. Ritter, S. Kelley, K. Popat and M. Pitkin, Effects of pore size, implantation time, and nano-surface properties on rat skin ingrowth into percutaneous porous titanium implants, J. Biomed. Mater. Res., Part A, 2014, 102(5), 1305–1315.
32. A. Trampuz, D. Osmon, A. Hanssen, J. M. Steckelberg and R. Patel, Molecular and antibiofilm approaches to prosthetic joint infection, Clin. Orthop. Relat. Res., 2003, 414, 69–88.

---

Footnote

† Electronic supplementary information (ESI) available. See DOI: 10.1039/c6ra22667a

---