Amine modification of calcium phosphates by low-pressure CH 4 /N 2 /He plasma for bone regeneration

Calcium phosphates are promising materials for artificial bone but lack of satisfied osteogenic ability on their surfaces. In the present study, we applied a low-pressure plasma technology to chemically (amine) modify the surface of calcium phosphates (hydroxyapatite or beta-tricalcium phosphate) using a CH 4 /N 2 /He plasm gas mixture to improve their osteogenic ability. The CH 4 /N 2 /He plasma treatment produced a thin, stable amine-rich carbon polymer on the surface of the calcium phosphates, and enhanced hydrophilicity, deep infiltration of cells into porous calcium phosphates, cell adhesion and osteogenic differentiation on the surface of calcium phosphates. In a rat calvarial defect model, the CH 4 /N 2 /He plasma treatment afforded calcium phosphates a significant higher bone regeneration capacity. Together, these results suggest that surface modification of calcium phosphates with CH 4 /N 2 /He plasma might improve osteogenic ability of calcium phosphates in vitro and in vivo. these


Introduction
Autogenous iliac bone grafting remains the gold standard for repairing large bone defects caused by trauma or tumors and spinal fusion surgeries. However, the amount of autograft that can be harvested is limited and the harvesting procedure can cause donor site morbidity 1 . To overcome these limitations, the use of artificial bones, in combination with autograft is prevalent. However, the widespread use of artificial bone is hampered by its lack of satisfied osteogenic ability, despite their superiority in terms of bone conduction and availability 2 .
Calcium phosphates are representative materials for artificial bone. To enhance the osteogenic ability of calcium phosphates, surface modification mediated by plasma technology has gained considerable attention. Functional groups created by plasma polymerization (i.e., polymer formation via plasma discharges) can provide selected surface properties such as hydrophilicity/hydrophobicity, cytocompatibility, and bacterial resistance to meet different clinical needs 3 . We previously reported that plasma treatment of hydroxyapatite (HA, a typical calcium phosphate used for artificial bone) with O2/He gas could improve surface hydrophilicity and promote the osteogenic differentiation of rat bone marrow stromal cells (BMSCs). However, the O2/He plasma-treated HA exerted only minimal effects in vivo, possibly owing to the instability of the generated hydroxyl (-OH) groups [4][5] . In the present study, we focused on the addition of amino groups or amines (-NHx with x = 0-2) to the surface of HA and beta-tricalcium phosphate (β-TCP, another typical calcium phosphate used for artificial bone), which has been suggested to promote cell attachment [6][7] . Using a mixture gas of CH4/N2/He as the precursor, we 5 successfully generated a thin, stable amine-rich carbon polymer on and inside the surfaces of the HA and β-TCP through a low-pressure plasma treatment, and enhanced cell infiltration, migration, cell adhesion, osteogenic differentiation in vitro, and bone regeneration with a rat calvarial defect model in vivo.

Materials and reagents
Both dense disks and porous disks of HA [Ca10(PO4)6(OH)2] and β-TCP  Plasma polymerization of HA and β-TCP disks Plasma polymerization using a bipolar pulsed-plasma deposition system (also known as an inverter plasma system) was as shown in Fig. 1A. The details of the system are described elsewhere [9][10]   For cross section analysis, the single side treated porous disk were cut in half at the center and the cross sections were scanned using high-resolution XPS.
The measurements were performed on 21 spots with a diameter of 110 μm that were separated by an equal distance of 100 μm along the center axis of the crosssection (Fig. 1B). During the analysis, the pressure of the XPS chamber was maintained at < 10 −7 Pa. The peaks of XPS signals used for the analyses in this study were C 1s, Ca 2p, N 1s, O 1s, and P 2p.
Detection of primary amines (-NH2) on CH4/N2/He plasma-treated disk surfaces was performed using the standard derivatization with 4-trifluoromethyl- where A is the cell area and P is the perimeter.

=
where ConvexA is the area of the smallest convex hull that contains the cell.

Cell adhesion assay
Cell suspensions (5 x 10 3 cells/35 μl GM) were gently dropped on dense β-TCP disks (φ 5 mm x h 2 mm) to form centroclinal water drops on the disks. After incubating for 30 min, a centrifugation cell adhesion assay was performed 16 .
Briefly, 1 ml of PBS was gently added into the wells and fluorescent macro-photos were obtained to quantify the initial adherent cells. β-TCP disks on which cells were attached were then embedded into a 48-well culture plate containing 100 μl of Vaselin. After filling each well with PBS, the plate was sealed with sealing tape.
Then, the culture plate was set upside-down on a centrifuge and centrifuged at 10 g for 5 min to detach weakly adherent cells. PBS and detached cells were slowly aspirated, then the wells were carefully filled with 200 μl PBS. Macro fluorescence photos were obtained using the same conditions as prior to the centrifugation. Automatic cell counting was performed using ImageJ software 17 .

11
The adhesion rate was calculated as follows:

Cell proliferation assay
To balance the initial cell count of adherent cells on plasma-treated and

Results
Plasma polymerization on calcium phosphate porous disks   Both cell counts and adhesion rate were all significantly higher on CH4/N2/He plasma-treated than the untreated β-TCP disks (Fig. 3A, 3B). The morphology of attached cells in the CH4/N2/He plasma-treated group following 24h incubation, presented a significant wider spreading cytoplasm (larger cell area) and showed a decreased circularity and solidity, which emphasized the presence of membrane protrusions (Fig. 3C, 3D) Quantitative analysis of attached cells before and after centrifugation, and adhesion rate.
(C) Fluorescence photos of attached cells on untreated or plasma-treated dense β-TCP disks. Scale bars = 500 μm. (D) Quantitative analysis of area, which represents the size, and solidity which represents the shape of attached cells. Data are expressed as mean ± SD with n = 1860 for "untreated" and n = 1440 for "plasma". ****p < 0.0001. T-test.

Figure 4. In vitro cell proliferation and osteogenic differentiation of GFP rat BMSCs.
Cell proliferation curve (A). Macro photos of ALP-stained BMSCs from GFP rats subcultured on untreated (B) and CH4/N2/He plasma-treated (C) dense β-TCP disks. Scale bars = 1 mm. ALP activity assay (D). ALP activity was normalized by total protein content.

Discussion
In this study, we generated a thin, stable amine-rich carbon polymer on the surface of calcium phosphates (HA and β-TCP) by low-pressure plasma technology using a mixed gas containing CH4/N2/He as precursors.
Low-pressure CH4/N2/He plasma allows chemically reactive gaseous species Amine constitutes a hydrophilic functional group. The present study demonstrated that amine modification of calcium phosphates could enhance its hydrophilicity, and fasten the infiltration of cell suspension dropped on the calcium phosphate surface. Notably, the CH4/N2/He plasma-treated β-TCP resorbed tissue fluid and rapidly became wet, in contrast to non-treated β-TCP, which remained almost completely dry during the implantation (Supplementary Fig. S3).
Moreover, histological sections at early phase (postoperative 3 weeks) revealed denser tissue formation in the CH4/N2/He plasma-treated disks. Taken together, 26 these findings suggest that the amine-rich carbon polymers can efficiently promote early tissue integration into the deep center of the disks, which makes preliminary preparations for later bone regeneration.
β-TCP coated with amine-rich carbon polymers enhanced in vitro osteoblastic differentiation and in vivo new bone formation. Several effects provided by amine modification must likely contributed to this enhanced bone formation.
Firstly, amine modification strengthens cell adhesion by enhancing integrin binding, which has been shown to be required for osteoblastic differentiation [18][19][20] . In detail, cell adhesion is mainly mediated by the binding of cellular integrins and adhesive proteins such as fibronectin in the extra-cellular matrix. Hydrophilic and positively charged amines can increase the density of fibronectin and change its conformation 21 . These changes in fibronectin strengthen the cell adhesion via integrin [22][23] and trigger rapid phosphorylation of focal adhesion-associated tyrosine kinase (FAK) 24 , subsequently triggering ERK/MAPK signaling to upregulate Runt-related transcription factor 2 (Runx2), which is a master regulator of osteoblastic differentiation [25][26][27][28][29][30][31] .
In addition, amine modification was found to modulate the cell morphology (e.g., larger cell area and decreased circularity and solidity), which has been shown to be relevant to the osteogenic differentiation of rat mesenchymal stem cells 32 . Moreover, human mesenchymal stem cells exhibiting a spreading rather than a round shape appear inclined toward an osteogenic lineage as well 33 . The underlying mechanism is considered to involve upregulation of Ras homolog family member A (RHOA), a transcription factor that regulates the actin cytoskeleton, along with increased osteogenesis [32][33][34] .
Lastly, positively charged amines can also improve osteogenesis by affecting pH 35 . In an aqueous environment, amine is protonated and becomes positively charged (-NH3 + ), which can increase the interfacial pH. A high pH environment around implant materials has been reported to enhance osteoblastic differentiation [35][36] .
In summary, our results support that amine modification of calcium phosphates with low-pressure CH4/N2/He plasma influences cell adhesion, cell spreading, and possibly the interfacial environment of calcium phosphates (here β-TCP). We consider that these factors in combination underlie the high ability of the modified β-TCP to improve bone regeneration.
There are limitations to this study. Firstly, the stability of the amine modification was observed for only two months, which should be elongated in future study for confirmation of longer off-the-shelf use. Secondly, we investigated the bone regeneration capacity only by the calvarial defect model. The calvaria has a rich blood supply and therefore is easy for bone regeneration. Further investigations of the bone regeneration effects in harsh environment such as malunions of fractures is needed. Thirdly, as the volume of newly formed bone inside the pores of the β-TCP disks was tiny, we could only evaluate the bone volume even with even a high-resolution micro-CT, but could not evaluate other bone parameters such as trabecular numbers and thicknesses. Finally, lacking biomechanical tests, we do not exactly know the real strength of the regenerated bone.
In treatment of large bone defects with artificial bone, the successful introduction of cells and blood vessels at a substantial distance from the host bone remains challenging [37][38] . To treat "critical size" bone defects, pre-loading of bone or vascular-forming cells or vascular transplantation inside artificial bones has been attempted [39][40][41][42] , which usually require long and costly pre-treatments.
In the present study, we have reported a "chemical modification" approach as a next-generation surface treatment and processing strategy to form a stable amine-rich carbon polymer on the surfaces of calcium phosphates

Data availability
The datasets generated and/or analysed during the current study are available from the corresponding authors on reasonable request.