Characteristics and properties of N-FN patches
The abbreviations for all samples in this work are as follows: flexible flat patch (FF patch), flexible nanotopographic patch (FN patch), N2 gas plasma-treated flexible flat patch (N-FF patch), N2 gas plasma-treated flexible nanotopographic patch (N-FN patch), O2 gas plasma-treated flexible flat patch (O-FF patch), O2 gas plasma-treated flexible multiscale nanotopographic patch (O-FMN patch).
Figure 1a shows a schematic of the plasma treatment process used in this study (described in detail in “Methods”). In our previous works, we reported the structures and topography on microenvironment of natural tendon and bone ECMs17,27. Briefly, native tendons consist of well-organized and highly aligned collagen fibers in ECMs with ~86% type I collagen of type I and small amounts of type III collagen. These collagen fibers were cross-linked with proteoglycans, revealing closely packed parallel structures28. In the tendon ECM, a collagen fiber is composed of a large number of fibrils. Collagen fibers that come together to form collagen fibers vary in diameter from 500 nm to 1 μm. Collagen fibers are assembled to form bundles (or fascia) that are ~10 mm long and 1–20 μm in diameter. These fiber bundles finally assemble to form tendon units 20–500 μm in diameter29. Cells are mainly located between the collagen fibers and are affected by the aligned collagen parallel array structure and pores which contributed to the exchange of oxygen and provided nutrition. Furthermore, the aligned nanotopographies of collagen fibers and nanopores are similarly observed in bone extracellular matrix17. SEM images of the surface morphology of N-FN patches revealed a highly aligned topography with grooves and ridges (~800 nm size), similar to the well-organized topography of the native tendon ECM, without deformation and etching of surfaces with increasing plasma treatment times (Fig. 1b).
a Fabrication of a polycaprolactone (PCL)-based FN patch. Inspired by the high aligned and well-organized nanotopography of the native ECM, the PCL-based FN patch was fabricated using capillary force lithography (CFL). Plasma surface modification on the surface of the FN patch. The surface of the patches was modified using N2 plasma reaction gas. b SEM images of the surface of FN patches treated by N2 plasma reaction gas under various times. Scale bars = 2 µm. c FT-IR analysis of the N-FF and N-FN patches treated plasma at various times. d XPS survey scans and e high-resolution N1s XPS spectra of FN patches and 30-min N-FN patches.
To verify whether the N2 reaction gas and various plasma treatment time conditions used in the plasma treatment process affected PCL properties, the polymer’s chemical characteristics were analyzed. The functional groups of plasma-treated patches were investigated by FT-IR spectroscopy (Fig. 1c). The characteristic absorption bands related to PCL (i.e., CH2 asymmetric stretching at 2944 cm−1, symmetric stretching at 2866 cm−1, C = O stretching vibration of carbonyl groups at 1721 cm–1, and deformation of C–O at 1161 cm–1) were detected in all patches. The chemical changes of the plasma-treated PCL patches were not detected compared to the PCL patches. The surface chemical composition of the N-FN patches was analyzed by XPS. Comparison of the survey scan spectra of FN and N-FN patches showed three separated peaks in all XPS spectra, which correspond to C1s (285 eV), N1s (400 eV), and O1s (532 eV) (Fig. 1d). A distinct N1s peak at 400 eV in the N-FN patch spectrum indicated that N2 plasma was successfully applied onto the FN patch. The surface atomic compositions of the FN patch were calculated to be 73.91%, 0.29%, and 25.81% for C1s, N1s, and O1s, respectively. The surface atomic compositions of the N-FN patch were calculated to be 61.83%, 3.34%, and 34.82% for C1s, N1s, and O1s, respectively. The high-resolution XPS N1s spectra of the N-FN patches showed that the N1s peaks of the FN and N-FN patches can be decomposed into a component: one main N–C = O (399.9 eV) (Fig. 1e). The atomic configuration of the FN patch was calculated to be 0.51% and those of N-FN was 2.96% for N–C = O. Assessment of wettability of patches through measurement of the water contact angle showed that FN patches have a lower contact angle (82.56 ± 1.8°) when compared with FF patches (88.32 ± 1.7°) (Fig. 2a). The water contact angle of both FF and FN patches gradually decreased with increasing N2 plasma treatment time, and the 30-min N-FN patches had a lower contact angle (17 ± 1.4°) than that of 30-min N-FF patches (21 ± 2.1°). The wettability and hydrophilicity of polymer surface treated with plasma are transient. The hydrophobic properties of synthetic polymers are restored several hours after plasma treatment because uncured hydrophobic polymer chains migrate to the surface30. These phenomena, in addition to the oxidation process, include charge leakage from the surface, as well as migration and diffusion and redirection of polar groups, and therefore depend on the surface treatment conditions, material properties, and storage conditions31. Accordingly, we conducted the hydrophobic recovery analysis to confirm the maintenance of hydrophilicity and the storage period of scaffold surfaces in air and room temperature conditions immediately after the plasma treatment. The plasma-treated scaffolds were used immediately after the plasma treatment for in vitro and in vivo experiments within an hour. To confirm the hydrophobicity variation on the surface after N2 plasma treatment on the FN patch, static contact angle measurements were made at various time points (30 min, 1 h, 2 h, 4 h, 1 day, 2 days, and 6 days) after plasma treatment (Fig. 2b). N-FN patches exhibited slight hydrophobic recovery at 30 min to 4 h and showed substantial hydrophobic recovery from 1 to 6 days after plasma treatment. However, the hydrophilicity of N-FN patches was maintained when compared with the static water contact angle of FN patches.
a Water contact angle measurement of the N-FN treated under various times (n = 10 for each group). b Wettability recovery on the surfaces of 30-min N-FN patches (n = 5). c Attachment and proliferation of cells on N-FN patches. Quantitative analysis of cell attachment and proliferation on N-FF patches and N-FN patches showed a gradual increase with plasma treatment time (n = 6 for each group). d Effect of N-FN patches on the osteogenic mineralization of tenocytes. Alizarin Red staining and quantification of the degree of osteogenesis showed that the 30-min N-FN patch promoted higher calcium expression levels of tenocytes when compared with other groups (n = 6 for each group). Scale bars = 200 µm. Error bars = mean ± standard deviation (*P < 0.05).
In vitro analysis of cell behavior on N-FN patches
To investigate whether the N2 plasma treatment on the surfaces of FF and FN patches influenced cell proliferation and attachment, we cultured human tenocytes on the patches for 6 h (cell attachment assay), 3 days (cell proliferation assay), and 5 days (cell proliferation assay), respectively (Fig. 2c). After 6 h of cell culture, unattached cells were removed by washing with PBS, and cells attached to the patches were quantified by the WST-1 assay. Tenocytes were well attached on all patches, irrespective of topographic properties, and cells on the 30-min N-FF and N-FN patches showed higher attachment than did those under other plasma treatment times. After 3 and 5 days of cell culture, cell proliferation was higher on the 30-min plasma treatment patches when compared with those of the other groups (Fig. 2c). The osteogenic mineralization of tenocytes on the N-FN patches was examined by culturing cells on the two scaffolds in osteogenic induction medium for 14 days. Alizarin Red staining (Fig. 2d) revealed slightly higher calcium expression levels on the 30-min N-FN patches than on other samples and the tissue culture polystyrene substrate (TCPS). Although the N-FN patches treated for 30 min showed the highest degree of quantification on the osteogenic mineralization, there was no significant difference from the other control groups.
Characteristics and properties of O-FMN patches
Figure 3a shows a schematic of the plasma treatment process used in this study (described in detail in “Methods”). SEM images of the surface morphology of the O-FMN patches revealed a highly aligned topography with grooves and ridges, similar to the well-organized topography of the tendon ECM (Fig. 3b). Nanosized pores generated by O2 plasma treatment for 30 min were observed due to etching and volatilization, which were not generated in the N-FN patches. The O-FMN patches showed the uniform distribution of various pore sizes and more pores, depending on the O2 plasma treatment time compared to the O-FF patches (Fig. 3b and Supplementary Fig. 1). The average pore sizes of 30 min O-FMN and 30 min O-FF patches were analyzed as 192.84 and 145.16 nm, respectively. These results are due to when oxygen plasma is treated on the flat surfaces of polymeric materials, the collision radius or sidewall collision of the reactive oxygen species occurs significantly less. In contrast, on the surface with nanotopography, the sidewall collision and the collision radius increase due to the diffusion of reactive oxygen species in the nanostructures, so that the bombardment and oxidizing effect occurs actively, forming numerous large pores32.
a Schematic of plasma surface modification using O2 plasma reaction gas on the surface of the FN patch. b SEM images of the O-FMN patch under various treatment times. The surface topography of the O-FMN patch showed the generation of nanopore structures without damage to the aligned nanotopography. Scale bars = 2 µm. c FT-IR analysis of the 30-min O-FMN. d XPS survey scans and e high-resolution O1s XPS spectra of FN patches and 30-min O-FMN patches. f Water contact angle measurements (n = 10 for each group) and wettability recovery (n = 5) on surfaces of the O-FMN patch. g Attachment and proliferation of cells on O-FMN patches (n = 6 for each group). After 5 days of cell culture, cell proliferation was higher on the 30-min O-FMN than on the 30-min N-FN. h Effect of O-FMN patches on the osteogenic mineralization of tenocytes. Alizarin Red staining revealed higher calcium expression levels on the 30-min O-FMN patches than on other samples and the tissue culture polystyrene substrate (TCPS). Quantification of osteogenic mineralization demonstrated the highest degree of osteogenesis by the cells on the 30-min O-FMN patches (n = 6 for each group). Scale bars = 200 µm. Error bars = mean ± standard deviation (*P < 0.05).
However, these pores did not result in notable deformation to the highly aligned nanotopography. This result indicated that the generation of nanosized pores onto the aligned nanotopography could form multiscale nanostructures similar to the complex microenvironment of the ECM. Thus, we hypothesized that the combination of multiscale nanostructures comprising nanopores and a well-defined nanotopography and functional groups applied by O2 plasma treatment would provide a physiochemically synergetic effect to improve cell affinity along with cell function and tissue regeneration.
To confirm whether the O2 reaction gas used in the plasma treatment process changed PCL properties, the chemical characteristics of the polymer were analyzed. The functional groups of the O-FMN patch were investigated by FT-IR spectroscopy (Fig. 3c). The characteristic absorption bands related to PCL (i.e., CH2 asymmetric stretching at 2944 cm−1, symmetric stretching at 2866 cm−1, C = O stretching vibration of carbonyl groups at 1721 cm–1, and deformation of C–O at 1161 cm–1) were detected in all samples, showing that functional groups were well maintained and were not affected by plasma treatments. The surface chemical composition of the O-FMN and FN patches was analyzed by XPS. As shown in Fig. 3d, all XPS spectra had three separated peaks corresponding to C1s (285 eV), N1s (400 eV), and O1s (532 eV). A distinct O1s peak at 532 eV in the O-FMN patch spectrum indicated that the O2 plasma had been successfully applied onto the FN patch. The surface atomic compositions of the FN patch were calculated to be 73.91%, 0.29%, and 25.81%, and those of the O-FMN patches were calculated to be 63.7%, 0.23%, and 36.07% for C1s, N1s, and O1s, respectively. The high-resolution XPS O1s spectra showed that the O1s peak of the FN and O-FMN patches can be decomposed into three components: C–O component (531.69 eV) and C = O (532.39 eV), and O = C–O (533.3 eV) components (Fig. 3e). The atomic configurations of the FN patch were calculated to be 11.7%, 6.83%, and 7.49% for C–O, C = O, and O = C–O, respectively. The atomic configurations of the O-FMN patch were calculated to be 10.29%, 13.48%, and 11.97% for C–O, C = O, and O = C–O, respectively. The wettability of the O-FF and O-FMN patches was evaluated by water contact angle measurement, showing that FN patches had a lower contact angle (82.56 ± 1.8°) than FF patches (88.32 ± 1.7°) (Fig. 3f). Similar to the N-FF and N-FN patch, the water contact angle of O-FF and O-FMN patches after plasma treatment were decreased, and 30-min O-FMN patches had a lower contact angle (18.4 ± 1.1°) than did 30-min O-FF patches (22.01 ± 2.2°). To confirm the hydrophobicity variation on the surface of O-FMN patches, the static contact angle measurements were conducted over O2 plasma treatment time (30 min, 1 h, 2 h, 4 h, 1 day, 2 days, and 6 days). As shown in Fig. 3f, O-FMN patches exhibited small hydrophobic recovery at 30 min to 4 h and substantial hydrophobic recovery at 1 day to 6 days after plasma treatment. However, the hydrophilicity of the O-FMN patch was well maintained compared with the high static water contact angle of FN patches. To confirm whether the generation of the pores on the scaffold surfaces affects the change in mechanical strength, the tensile strengths of the FF, FN, 30 min O-FF, and 30 min O-FMN patches were measured using a tensile tester and assessed (Supplementary Fig. 2). When a load was applied along the direction of the aligned topography, the FN patches with the aligned nanotopography exhibited slightly larger tensile stress (~10.38 MPa) than that (~9.62 MPa) of FF patches with the flat topography. However, the 30 min O-FF patches and the 30 min O-FMN patches with nanopores generated oxygen plasma treatment showed no significant difference compared to the FF patch and the FN patch, respectively (Supplementary Fig. 2). This trend was measured in breakpoint strain analysis of FF, FN, 30 min O-FF, and 30 min O-FMN patches.
In vitro cell behaviors on O-FMN patches
Cell attachment and proliferation on N-FN and O-FMN patches were both higher than those on FF patches. In addition, after 5 days of cell culture, cell proliferation was considerably higher on O-FMN patches treated with O2 plasma for 30 min than that observed on 30-min N-FN patches (Fig. 3g). This suggests that the surface modified by O2 plasma may provide various cell-friendly functional groups and hierarchically topographical environments that are similar to the complex microenvironment of the ECM, thus promoting the proliferation and attachment of tenocytes when compared with the surface modified by N2 plasma. In addition, we examined the osteogenic mineralization of tenocytes on the N-FN and O-FMN patches by culturing cells in osteogenic induction medium for 14 days. Alizarin Red staining (Fig. 3h) revealed higher calcium expression levels on the O-FMN patches for 30 min than on N-FF and N-FN patches for 30 min and the TCPS. Quantification of osteogenic mineralization further demonstrated the highest degree of osteogenesis by cells cultured on the 30-min O-FMN patches. These results suggest that plasma treatment of FN patches may provide cell-friendly functional groups and topographical environments to enhance the attachment, proliferation, and differentiation of tenocytes. Therefore, the highly aligned nanotopography, generation of nanosized pores, and various functional groups induced by plasma treatment using O2 gas synergistically contributed to the proliferation and differentiation of tenocytes.
In vivo animal study for soft- and hard-tissue regeneration
RC tendon tears are one of the most common causes of shoulder pain33. Surgical repairs of RC tendon tears have high re-tear rates, and thus many devices have been developed to augment the repair efficacy34. However, repairing defected mineralized fibrocartilage of the tendon–bone interface that causes the high re-tear rate of RC remains a challenge. Here, we propose an approach for the regeneration of mineralized fibrocartilage tissue of the tendon–bone interface. According to the results of in vitro analyses with tenocytes, we determined the optimum plasma treatment time and plasma reaction gas (O2 vs. N2) based on their effects on cell proliferation and differentiation. Since the O-FMN patch mimicking the multiscale ECM nanostructure with various functional groups could provide an efficient environment for cell proliferation and osteogenic mineralization, we hypothesized that the O-FMN patch with a highly aligned nanotopography, nanosized pore, and various functional groups would effectively guide RC tendon and mineralized fibrocartilage tissue regeneration. To confirm the tissue regeneration efficacy of the O-FMN patch, an acute RC tendon tear rat model was established.
Figure 4a shows the procedure of implantation of patches at the RC repair site (i.e., all samples were implanted after tenorrhaphy immediately following artificial tendon rupture). All rats survived to the designated sacrifice date, and no adverse events were observed. Hematoxylin and eosin (H&E) staining was performed to evaluate the histologic quality on the connection of collagen fibers in the tendon–bone interface, orientation and density of the collagen fibers, maturity of the tendon–bone interface, and cell confluency of the regenerated RC tendon and mineralized fibrocartilage tissue 4 weeks after repair with non-treated patches, 30-min N2 plasma-treated patches, and 30-min O2 plasma-treated patches (Fig. 4b and Supplementary Fig. 3). No infection, contracture, mobility disability, or inflammatory reaction were observed in any of the rats throughout the postoperative period. The tissues from rats treated with the N2 (N-FF and N-FN patches) and O2 (O-FF and O-FMN) plasma-treated patches showed native rat tendon tissue-like histological healing patterns (Supplementary Fig. 4), with high ratios of tendon tissue regeneration observed in the sections under the patches. The RC tendons repaired using the FN patches showed well-organized collagen fibers of high density, whereas those of the FF control groups showed sparse fibrous tissue and low cell confluence of the bone and fibrocartilage at the wound site (Fig. 4c). In addition, the tendon tissues treated with the N2 (N-FF and N-FN patches) and O2 (O-FF and O-FMN) plasma-treated patches showed organization of the collagen fibers similar to that of native tendon tissues, with increased amounts of cellular at the tendon–bone interface compared to those of non-treated groups. The fibrocartilage tissues at the tendon–bone interfaces treated with the O-FMN patch showed a more well-organized vertical arrangement compared to those of the non-treatment groups and N-FN patch groups. Importantly, the fibrocartilage tissues at the tendon–bone interfaces from rats treated with the O-FMN patch showed a vertical arrangement, alignment, and increased amounts of cellular similar to the native tendon tissue (Fig. 4c). The O-FMN patch had an obvious positive influence on the collagen organization, connection, and tendon tissue regeneration, resulting in a structure similar to that of native RC tendon tissues and fibrocartilage tissues at the tendon–bone interface. These results indicate the importance of a precisely aligned nanotopography and nanosized pore, which generated the multiscale structure of the native ECM microenvironment to effectively guide tendon tissue and fibrocartilage regeneration.
a Surgical procedure for RC tendon repair. All patches were grafted onto the defected tendon tissue after tenorrhaphy of the torn RC tendon (n = 3 for each group). b Representative histologic images of H&E staining and c Masson trichrome staining of the insertion site of FF, FN, 30-min N-FF, 30-min N-FN, 30-min O-FF, and 30-min O-FMN patches onto the supraspinatus tendon 4 weeks after repair. Scale bars = 200 µm. d Semiquantitative analysis of the histological evaluation scores on repaired tendon tissues of RC tear animal models. e Surgical procedure for rat calvarial bone repair (n = 5 for each group). f Representative histologic images of H&E staining of the insertion site of FN, N-FN, and O-FMN patches onto rat calvarial bone 6 weeks after repair. Scale bars = 1 mm. g Representative micro-CT image after 3 and 6 weeks of repair. Scale bars = 5 mm (3 weeks). Scale bars = 1 mm (6 weeks). h Quantitative analysis of bone volume and area of the bone regeneration site after 3 and 6 weeks of repair. Error bars = mean ± standard deviation (*P < 0.05).
Based on the qualitative histological analysis of the effect of the O-FMN patch presented as a repair strategy for RC tendon tissue rupture, a semiquantitative histology analysis was performed using the Bonar scoring system (Fig. 4d and Supplementary Table 1). Cell morphology, ground substance, collagen arrangement, and vascularity changes were observed in all groups and were graded depending on Bonar score (Supplementary Table 2). The cell morphology and collagen arrangement scores of the O-FMN patch group were significantly lower than those of other groups (Fig. 4d). The ground substance and vascularity scores of the O-FMN patch group were similar to the score of the O-FF patch group but were significantly lower than those of other patch groups. Overall, the total histological score of the O-FMN patch group was significantly lower than those of other patch groups (Fig. 4d). These results indicate the importance of the oxygen plasma treatment, a precisely aligned nanotopography, and the nanoporous structures in synthetic ECMs for guiding tendon tissue and fibrocartilage regeneration.
In addition, we confirmed the effects of N2 and O2 plasma treatment and nanotopograhpy throughout bone regeneration in vivo (Fig. 4e). All mice used in the in vivo studies survived to the date of sacrifice, no adverse events were observed. FN patch, N-FN and O-FMN patch were engrafted onto the calvarial bone defect with 5 mm diameter (Fig. 4e). No infection or inflammatory response was observed in any mice during the postoperative period. The implanted patches were maintained for 6 weeks without deformation. To confirm the bone regeneration efficacy from N2 and O2 plasma treatment and nanotopograhpy, we conducted the H&E staining at 6 weeks after implantation (Fig. 4f). As the result, the bone of the defect groups was empty in the defect area. More bone formation and dense cytoplasm occurred in O-FMN patches compared to other groups were confirmed.
To quantitatively evaluate the effects of nanotopography and plasma treatment on bone formation, we performed micro-CT and 3D-image conversion using the MIMICS 14.0 software on new bone defects in vivo. As shown in Fig. 4g, formation of bone by the nanotopographic and plasma-treated patches occurred along the periphery of the bone defect and grew along the patches. After 3, 6 weeks, the compact bone formation was not observed in the defect group, FN patch groups, and N-FN groups whereas bone regeneration was significantly enhanced in the O-FMN patch groups after 3, 6 weeks of implantation. The new bone formation was observed from edge to center depending on nanotopograhpy direction. At the 6 weeks, the bone volume was 1.91 mm3 in the defect group, 2.43 mm3 in the FN patch groups, and 3.38 mm3 in the N-FN patch groups, and 4.25 mm3 in the O-FMN patch groups (Fig. 4h). The bone area was 19.81 mm2 in the defect groups, 26.09 mm2 FN patch groups, 37.92 in the N-FN patch groups, and 47.92 mm2 in the O-FMN patch groups (Fig. 4h). The results of bone regeneration and formation provide insight into the importance of nanotopography and O2 plasma treatment cues for inducing hard-tissue regeneration.
Quantitative investigation of relative contributions
Based on these in vitro and in vivo results, we quantified the capability of tissue regeneration and cell function enhanced by the beneficial effects of the O-FMN patch on proliferation and differentiation (Supplementary Fig. 5). Quantitative investigation of relative contributions was derived by setting the raw average values of the proliferation and osteogenic mineralization absorbance of the FF patch as 1, and calculating relatively the absorbance of the FN, 30-min N-FF, 30-min N-FN, 30-min O-FF, and 30-min O-FMN patches. The aligned nanostructure of the FN patch, functional groups applied by N2 plasma treatment, and nanosized pores and functional groups generated by the O2 plasma treatment could only promote proliferation by a factor of 1.01, 1.12, and 1.1, respectively, and osteogenic differentiation was promoted by a factor of 1.37, 1.32, and 1.44, respectively. The combination of a nanotopography and N2 plasma treatment increased proliferation and osteogenic differentiation by a factor of 1.36 and 1.4, respectively. In addition, the combination of nanotopography and O2 plasma treatment significantly increased the proliferation and osteogenic differentiation by a factor of 1.74 and 1.78, respectively, further confirming the enhanced effects of highly aligned nanotopography, nanosized pores, and functional groups. This synergetic effect may be due to improvement in cell topography or cell–cell interactions from the highly aligned nanotopography and nanosized pores generated by O2 plasma treatment.
