Calcium and phosphorus co-doped carbon dots enhance osteogenic differentiation for calvarial defect repair in situ

Calvarial bone defect remains a clinical challenge due to the lack of efficient osteo-inductive agent. Herein, a novel calcium and phosphorus codoped carbon dot (Ca/P-CD) for bone regeneration was synthesized using phosphoethanolamine and calcium gluconate as precursors. The resultant Ca/P-CDs exhibited ultra-small size, stable excitation dependent emission spectra and favorable dispersibility in water. Moreover, Ca/P-CDs with good biocompatibility rapidly entered the cytoplasm through endocytosis and increased the expression of bone differentiation genes. After mixing with temperature-sensitive hydrogel, Ca/P-CDs were injected in situ into calvarial defect and promoted the repair of bone injury. These Ca/P-CDs provide a new treatment method for the bone repair and should be expended the application in the biomedical fields.


Introduction
Calvarial bone defect represents a frequent pathological issue of orthopedic which can be caused by trauma, congenital malformations, infections, and surgery [1][2][3]. In clinic, the therapeutic effect of calvarial bone defect is unsatisfactory owing to the lack of efficient osteo-inductive agent. In addition, traditional surgical transplantation has serious side effects and limited application including graft selection and immune rejection [4,5]. Therefore, there is an urgent need to develop effective treatment strategy for promoting bone regeneration effectively and restoring bone function.
It is well-known that calcium and phosphorus are important elements in the process of maintaining normal bone differentiation and development. They cooperate with parathyroid hormone and calcitonin to promote the expression of genes related to bone differentiation, and achieve the regulation of bone cell function. In addition, both of elements have an impact on bone metabolism and bone tissue morphology [6]. For instance, hydroxyapatite with a general formula of Ca 10 (OH) 2 (PO 4 ) 6 has extensively applied in bone defect repair due to its chemical similarities with the natural bone [7,8]. However, the uncontrollability of morphology and metabolism in vivo of hyaluronic acid (HA) weakens its therapeutic effect. As a new type of nano-biomaterial, carbon dots (CDs) have an ultra-small size (2-8 nm), excellent optical properties, good biological safety and stability [9,10]. It has attracted widespread attention in the fields of bioimaging, immunolabeling, molecular tracing, tumor targeting, and drug delivery [11]. By modifying the abundant active groups on the surface, CDs can be endowed with different specific functions to meet specific biomedical needs [12]. Our previous works has confirmed that functional CDs possessed a huge potential in diagnosis and treatment of tumor [13][14][15].
In view of this facts, we hypothesized that CDs could be used as a promising nanocarrier to integrate the calcium and phosphorus elements for bone regeneration and repair of bone injury.
The physical and chemical properties of Ca/P-CDs was characterized via a series of methods. Afterwards, the function of osteogenesis induced by Ca/P-CDs was explored using mouse osteoblastic cell line (MC3T3-E1) in vitro. Meanwhile, Ca/P-CDs mixed with thermo-sensitive hydrogel (Pluronic F127) were injected into calvarial defect in rats to investigate the bone regeneration in situ. Finally, we verified the long-term biocompatibility of Ca/P-CDs through histologic section.

Materials
Phosphoethanolamine, calcium gluconate, diethylenetriamine pentaacetic acid (DTPA), and glycine were obtained from Aladdin Reagent Company (Shanghai, China). Dulbecco's minimum essential medium (DMEM) and Fetal bovine serum were purchased from Hyclone (Logan, UT, USA). Lyso tracker and Mito tracker were purchased from Beyotime Biotechnology (Beijing, China). qPCR kits were purchased from Sigma (NY, USA). ICR rats were purchased from Center for Experimental Animals of Jiangsu University. Recombinant Human Bone Morphogenetic Protein-2 (rhBMP-2) was obtained from R&D Systems Europe (United Kingdom). All of these chemical agents were of analytical grade and were utilized without further purification.

Synthesis of bare CDs, P-CDs, Ca-CDs and Ca/P-CDs
Three types of CDs were synthesized using different precursors by a one-pot hydrothermal method, as described in the literature previously with a little modification [15]. Bare CDs (without P or Ca doping): 0.4 g of glycine and 0.40 g of DTPA were dissolved in 20 ml of double distilled water and stirred continuously for 10 min to form a transparent solution. The solution was transferred to a Teflon-lined stainless steel autoclave and heated in a muffle furnace at 200 • C for 4 h. After cooled to room temperature, the reaction solution was poured into a 50 ml centrifugal tube and the black large particles were discarded by centrifuging at 2000 rpm for 15 min. The obtained supernate was then dialyzed against water for 5 d in a cut-off dialysis bag (MWCO = 3 kD, Solarbio Company, Beijing, China). Afterward, the supernate was collected and freeze-dried with vacuum freeze dryer. Thus, bare CDs powder was obtained and stored for further characterization. The P-CDs (with phosphorus doping), Ca-CDs (with calcium doping) and Ca/P-CDs (with phosphorus and calcium doping) were prepared as above method using the phosphoethanolamine and DTPA, calcium gluconate and DTPA, and phosphoethanolamine and calcium gluconate as precursors, respectively. In addition, the molar rate of P/Ca in Ca/P-CDs is 1:1.

Morphological and chemical characterization of Ca/P-CDs
The morphologies of the Ca/P-CDs were observed by high-resolution transmission electron microscopy (HRTEM) on a JEM-2100 microscope (JEOL, Japan) under an accelerating voltage of 200 kV. The surface chemical components of Ca/P-CDs were examined using a Fourier transform infrared (FTIR) spectrometer (Nicolet Nexus 470, USA) ranged from 4000 to 400 cm −1 . Elemental composition of the Ca/P-CDs was characterized by x-ray photoelectron spectroscopy (XPS) on Escalab 250Xi (Thermo Scientific, America). The crystal structure of Ca/P-CDs was analyzed by x-ray diffraction (XRD) on a Rigaku-D/MAX2500 diffractometer (Japan) with a scanning speed of 4 • min −1 in the range from 5 • to 80 • . The optical properties of the Ca/P-CDs were obtained with a UV-2450 UV/Vis spectrophotometer (Shimadzu, Japan), and the photoluminescence emission spectra was recorded using a Cary Eclipse Fluorometer (Varian, USA).

Cell culture and biocompatibility of Ca/P-CDs
The cytotoxicity of the Ca/P-CDs in vitro was measured using the CCK-8 cell viability kit assay (Solarbio, Beijing, China). Briefly, MC3T3-E1 (mouse embryo osteoblast precursor) cells (1 × 10 4 cells per well) were seeded into a 96-well plate with four replicates in each group. After incubation for 24 h, the growth medium were replaced with the different concentrations of Ca/P-CDs (10-800 µg ml −1 ) in fresh DMEM and continued to incubate for another 24 h. Then, the wells were washed with phosphate-buffered saline (PBS), and 10 µl of CCK-8 and 90 µl of DMEM solution were added to keep incubation for 4 h at 37 • C and 5% CO 2 . At the end, the absorbance of each well was measured at 450 nm using a Synergy HT Multi-Mode Microplate Reader (Bio Tek, USA). The relative cell viability (mean ± standard deviation (SD), n = 3) was expressed as (Abs sample-Abs zero sitting)/(Abscontrol-Abszero sitting) × 100%.

Hemolysis assay
All animal procedures were conducted in accordance with the Management Rules of the Ministry of Health of the People's Republic of China and approved by the Institutional Animal Care and Use Committee of Jiangsu University (permit number: SYXK2018-053). The hemocompatibility of Ca/P-CDs was carried out according to the protocol reported in the literature with slight modification [16]. In brief, fresh mouse blood was stabilized with heparin sodium and centrifuged (1200 rpm, 15 min) to remove the supernatant. The sediment was washed with PBS five times

Name
Primer Sequence Product to obtain the mouse red blood cells. Next, the mouse RBCs were resuspended using 0.9 ml of Ca/P-CDs in PBS from 50 to 200 mg ml −1 (water as positive control and PBS as negative control). These samples were incubated for 2 h at 37 • C after gentle shaking, thus centrifuged to collect the supernatant at 12 000 rpm for 1 min. The absorbance of samples at 541 nm was recorded by a UV-Vis spectrophotometer (UV-2450). The hemolysis percentages of Ca/P-CDs was calculated using the absorbance compared with the control.

Cytophagocytosis assay
The internalization of Ca/P-CDs in the cells was observed using laser confocal microscope. MC3T3-E1 cells at a density of 1 × 10 5 /well were seeded into the 24-well plates that pre-filled with 10 mm coverslips. When the cell density reached 60%, the MC3T3-E1 cells were administered with the DMEM medium containing 200 µg ml −1 of Ca/P-CDs and further incubated for 4 h. Then, these MC3T3-E1 cells on the coverslips were washed with PBS twice and fixed with 4% paraformaldehyde (PFA). In order to determine the intracellular distribution, these MC3T3-E1 cells stained with Lyso tracker and Mito tracker were observed under a confocal laser scanning fluorescence microscope (Zeiss LSM-710, Carl Zeiss Meditec AG, Jena, Germany)

Alizarin red staining
Alizarin red staining was choosed to verify the osteogenic differentiation of MC3T3-E1 cells after incubation with Ca/P-CDs in DMEM medium. The experiments was divided into five groups, including control, bare CDs P-CDs, Ca-CDs and Ca/P-CDs. In each group, the dosages of CDs were added according to the proportion of calcium and phosphorus equal mole. MC3T3-E1 cells were seeded into a 96-well plate at a density of 1 × 10 4 cells per well with four replicates. After incubation with 50 µg ml −1 Ca/P-CDs for 21 d, the MC3T3-E1 cells were washed with PBS and fixed with PBS PFA (3.6%) for 1 h at 4 • C. Then, the MC3T3-E1 cells were stained with 1% alizarin red (sigma Aldrich) solution for 30 min. Finally, these MC3T3-E1 cells were washed by PBS to remove free alizarin red, and the Ca/P-CDs-induced mineralization was measured by microscope and Image-J software.

Real-time PCR
The osteogenic induction of MC3T3-E1 cells using Ca/P-CDs was same as the above. After incubation with the Ca/P-CDs, these MC3T3-E1 cells were collected by enzymatic digestion and their total RNA was extracted by Trizol reagent. RT-PCR was conducted on the ViiA 7 RT-PCR System (Thermo Fisher Scientific) using the QuantiTect SYBR Green Kit (Qiagen, Quanta, France). The primer sequences of the related genes were listed in the table 1. The expression of each related gene, including alkaline phosphatase (ALP), osteocalcin (OCN), and Runt-related transcription factor 2 (RUNX2), was normalized to the housekeeping gene (β-actin), and fold differences were calculated using the comparative C t method.
The osteogenic markers ALP, OCN and RUNX2, were analyzed.

In vivo animal study
All animal procedures were conducted in accordance with the Management Rules of the Ministry of Health of the People's Republic of China and approved by the Institutional Animal Care and Use Committee of Jiangsu University (permit number: SYXK2018-053). Eight-week-old Wistar SD rats were used (body weight ∼20 g) for the experiments in vivo. Ten rats were randomly divided into two groups with five rats in each group. The brief surgical procedure was as follow: Firstly, the rats were anaesthetized with 10% chloral hydrate in normal saline by intraperitoneal injection at 3.5 ml kg −1 body weight. Their fur on the skull surface was shaved off using an electric razor and this area was sterilized with iodine. Then, a 1.5 cm long longitudinal skin incision was made on the mouse scalp from the back of the eyes to the skull area using a sterile scalpel. The fascia was stripped laterally to expose the skull. Next, a 5 mm diameter defect was made using an electric drived low-speed (about 1500 rpm) trephine bur on one side of the parietal bone and cooled with normal saline during the whole process. Finally, 20 µl of Ca/P-CDs loaded F127 hydrogel (100 µg ml −1 ) was injected into the calvarial defect. Pure F127 hydrogel (20%) was used as control. After the surgery, the wound was sutured with simple interrupted suture and then the rats were placed separately and fed normally. After 8 weeks, the major organs (heart, liver, spleen, kidneys, and lungs) were collected for conventional histocompatibility analysis.

CT scanning
CT images of the calvarial defect were acquired on a clinical 64-slice multidetector CT scanner (SOMA-TOM Emotion, Siemens, Bavaria, Munich, Germany) every two weeks. Then, trabecular bone volume/tissue volume (BT/TV) were calculated using the builtin program.

Statistical analysis
All experimental data were presented as mean ± SD and analyzed by one-way analysis of variance or Tukey's multiple comparison tests between different treatments. P < 0.05 was considered statistically significant.

Characterization of Ca/P-CDs
The successful preparation of Ca/P-CDs was performed by using calcium gluconate as calcium source and phosphoethanolamine as surface passivation agent based on scheme 1. The morphology of synthesized Ca/P-CDs was observed using HRTEM. As shown in figures 1(A)-(D), the Ca/P-CDs possessed discrete and quasi-spherical shape without apparent aggregation, which similar with Ca-CDs, P-CDs, bare CDs. The average diameter of Ca/P-CDs was approximately 2.4 nm that could be conducive to cross the physiological barrier and discharge from the body by the kidney. Figure 1(B) showed that there was no discernible lattice fringes, indicating the amorphous nature of Ca/P-CDs. XRD pattern was used to study the phase structure of Ca/P-CDs. As shown in figure 1(C), there was an strong and broad diffraction peak at 24.98 • , which was similar to the characteristic peaks of graphite ({002} planes, 2θ = 26.5 • ). Subsequently, we used FT-IR spectrum to study the chemical composition of Ca/P-CDs ( figure 1(D)). FTIR spectrum of Ca/P-CDs showed the stretching vibration of O-H at 3450 cm −1 , C=O at around 1650 cm −1 , C-N/C-C at around 1400 cm −1 , C-O at around 1100 cm − 1 , suggesting the abundant hydrophilic groups on the surfaces of Ca/P-CDs.

Chemical characterization of Ca/P-CDs
The surface chemical structure of Ca/P-CDs was further studied using x-ray XPS. As shown in figure 2(A), there were five obvious peaks at 350, 133, 284.0, 400.0, and 530.6 eV, which indicated that the Ca/P-CDs Scheme 1. Schematic diagram of Ca/P-CDs preparation and application in osteogenic differentiation.
were mainly composed of calcium, phosphorus, carbon, nitrogen, and oxygen atoms. The content of Ca and P in the Ca/P-CDs was 3.5% and 2.3%, which was higher than that of Ca in the Ca/-CDs (2.2) and P in the P-CDs (1.6%) in the figures 2(B) and (C). The Ca/P ratio in the Ca/P-CDs was 3:2 according the XPS analysis. It is worth noting that Ca/P ratio before (1:1) and after (3:2) the reaction was inconsistent, probably because phosphorus atom is more similar to carbon atom than calcium. As shown in figures 2(D) and (E), the N 1s spectrum displayed remarkable N-H peaks at 399.9 eV and N-C peak at 399.3 eV, where the incorporation of N atoms could be used as auxochrome group to enhance the optical performance of Ca/P-CDs [17,18]. The XPS spectrum showed the P 2p peak at 133.60 and 132.85 eV, which was contributed to P-O and P-C bonds, respectively (figures 2(F) and (G)). In the figure 2(H), Ca 2p contained two groups of peaks were attributed to Ca 2p 1/2 (350.70 and 350.16 eV) and Ca 2p 3/2 (347.30, 346.55 and 346.85 eV), which were assigned to the presence of CaO and CaCO 3 . It was verified that the Calcium and phosphorus elements were successfully doped into Ca/P-CDs, where both of the elements were essential for bone growth and differentiation [19].

Optical characterization of Ca/P-CDs
The optical properties of Ca/P-CDs were characterized by fluorescent spectrum and UV/Vis absorption spectrum. Figure 3(A) showed there was no obvious absorption peak in an aqueous solution of Ca/P-CDs. Meanwhile, the aqueous solution was pale brown and transparent in daylight, but exhibited bright blue fluorescence under UV irradiation (inset, figure 3(A)). The Tyndall effect occurred when a bean of red laser gone through the aqueous solution, indicating that Ca/P-CDs had good dispersibility and colloid stability in water. The quantum yield of Ca/P-CDs was calculated at 11.3% using quinine bisulfate as a standard. As shown in figure 3(A), the Ca/P-CDs displayed a broad range of emission wavelengths, resulting in preeminent multicolor fluorescent emission. In addition, the fluorescent spectrum of Ca/P-CDs showed excitation-dependent emission manner, which was well similar to the conventional CDs [20,21]. The maximum emission of Ca/P-CDs at 480 nm was observed under  the excitation wavelength at 396 nm, which further confirmed the blue fluorescence performance ( figure 3(D)).

Biocompatibility of Ca/P-CDs
Biocompatibility of nanomaterials is of great importance for biomedical applications. Herein, the cytocompatibility of Ca/P-CDs was assessed by CCK-8 assay using incubation with VSMC and MC3T3-E1 cells. The cell viability of VSMC and MC3T3-E1 cells were detected after 24 h incubation with Ca/P-CDs. The figure 4(A) shown that their cell viability did not show significant changes after treatment with Ca/P-CDs at 0, 10, 50, 100, 200, 400, 800 µg ml −1 . Even at 800 µg ml −1 , both of cell viability were still as high as 88% and 91%, respectively. Furthermore,  the hemocompatibility of Ca/P-CDs was assessed by hemolytic assay. As shown in figure 4(B), there was strong and sharp absorption peak at 541 nm in the positive control (DI Water group), indicating the presence of the hemolysis of red blood cells. Compared to the negative control (PBA group), the percentages of hemolysis were all less than 3% after treatment with Ca/P-CDs from 10 to 1000 µg ml −1 , indicating that no significant hemolysis occurred. Taken together, these findings clearly suggested that the prepared Ca/P-CDs had negligible cytotoxicity and good biocompatibility.

Intracellular distribution of Ca/P-CDs after endocytosis
Numerous studies have reported that the conventional CDs readily entered into cytoplasm in living cells due to their ultra-small size [22,23]. However, the intracellular localization of Ca/P-CDs after endocytosis was still unknown. In this study, three fluorescent probes (MitoTracker, ER-Tracker, or LysoTracker) were chosen to investigate the distribution of Ca/P-CDs at the organelle level. As shown in figure 5(A), the presence of strong green fluoresce signal in three groups verified that Ca/P-CDs could be successfully internalized into cytoplasm after 6 h of co-incubation. In addition, MC3T3-E1 cells had no obvious morphological changes. Furthermore, the red fluorescence from MitoTracker and ER-Tracker highly overlapped the green fluoresce from Ca/P-CDs, respectively. However, this fluorescence signal of colocalization for LysoTracker was relatively weak. Quantitative analyses using Image-J software demonstrated that Ca/P-CDs were mainly accumulated in mitochondria and endoplasmic reticulum in figure 5(B). According to these findings, we proposed that the Ca/P-CDs entering the cytoplasm were mainly distributed in the mitochondria and endoplasmic reticulum.

Osteogenic induction of Ca/P-CDs in vitro
The function of osteogenic induction of Ca/P-CDs was explored by the RT-PCR and Alizarin red staining using MC3T3-E1 cells. Three different types of CDs was chosen as control treatments, including bare CDs (B-CDs), calcium-doped CDs (Ca-CDs), Phosphorous-doped CDs (P-CDs). The dosages were added according to the proportion of calcium and phosphorus equal mole. As shown in figures 6(A) and (B), the mineralization of MC3T3-E1 osteoblast cell in the Ca/P-CDs treatment group was significantly higher than that of the other group. The function of Ca/P-CDs was further enhanced by the addition of the G. In addition, the Ca-CDs group exhibited better osteogenic induction than the P-CDs group. Meanwhile, the expression of ALP, mature bone markers (OCN) and osteogenic transcription factor (RUNX2) were significantly up-regulated under the same concentration of Ca-CDs, Ca/P-CDs treatments, among which Ca/P-CDs group has the highest expression. Ca/P-CDs could promote the expression of bone differentiation-related RNA to be 2-3 times higher than that of the control group (figures 6(C) and (D)). It can be proved that the Ca/P-CDs could effectively promote the expression of genes related to bone differentiation and promote development and maturity of bone differentiation.

Osteogenic potential of Ca/P-CDs in vivo
Encouraged by the above results, we evaluated the ability of Ca/P-CDs to promote bone differentiation and development through in vivo experiments. In order to achieve administration in situ, F127 thermal sensitive hydrogel was chosen to load the Ca/P-CDs.  After in situ injection, F127 hydrogel loading Ca/P-CDs could form a gel quickly in the calvarial defect due to the body temperature rising. During this treatment, CT scans were performed to observe the healing of the calvarial defects every two weeks. As shown in figure 7(A), the area of calvarial defect in F127 hydrogel (control group) and Ca-CDs group had change barely after 8 weeks of treatment. Surprisingly, the area of calvarial defect in Ca/P-CDs group decreased obviously. Quantitative analyses from bone histomorphometry of 3D reconstruction showed that new bone area in control and Ca-CDs group was 2.1 and 2.4 mm 3 while new bone area in Ca/P-CDs group was 6 mm 3 ( figure 7(B)). Furthermore, BV/TV in control, Ca-CDs and Ca/P-CDs group was 10%, 12% and 31%, indicating that amount of new bone at defect area dramatically increased after CD Ca/P-CDs treatment compared with control and Ca-CDs treatment ( figure 7(C)). In addition, the corresponding hematoxylin eosin (HE) staining showed that Ca/P-CDs treatment promoted the regeneration of osteoblasts and infiltration of soft tissue ( figure 7(D)). The results verified that Ca/P-CDs could effectively promote the repair of bone injury in vivo. Besides, histological analysis of major organs were used to evaluated the potential long term biosafety of Ca/P-CDs. As shown in figure 8, it showed no visible pathological changes in the lung, liver, spleen, kidney and heart after 8 weeks treatment with PBS, P-CDs, Ca-CDs, Ca/P-CDs. There was no obvious inflammation or quantum dot deposition in various organs. Our previous study and other reports had confirmed that CDs can be excreted out of the body through the kidneys, which reflected the good biological safety of CDs [24,25].

Conclusion
In this study, we have successfully prepared the Ca/P-CDs with osteogenic potential via introduction of calcium and phosphorus element by onepot hydrothermal carbonization. The as-prepared Ca/P-CDs had uniform particle size distribution and good dispersibility in water. Furthermore, the Ca/P-CDs exhibited stable optical properties and favorable biocompatibility. After entering into the cytoplasm, the Ca/P-CDs could escape from the lysosome and mainly distributed in the mitochondria and endoplasmic reticulum (ER). More importantly, these internalized Ca/P-CDs upregulated the expression of bone differentiation-related genes, including ALP, RUNX2 and OCN. Finally, the Ca/P-CDs promoted the development of osteoblasts in vitro and the repair of calvarial defect in vivo. In summary, the study can provide a novel alternative treatment for clinical bone injury repair and expand the applications of CDs in the biomedical fields.

Data availability statement
All data that support the findings of this study are included within the article (and any supplementary files).