Magnetic casein/CaCO₃/Fe₃O₄ microspheres stimulate osteogenic differentiation

Bone defects resulting from trauma and tumors profoundly impact patients’ quality of life.^1,2 Ideal bone repair materials necessitate biocompatibility, biodegradability, osteoinductivity, osteoconductivity, osteogenesis, and adequate hardness.^3–5 Current popular repair materials encompass metal alloys, bone cement, and polymer scaffolds.^6–8 Metal alloys, such as titanium, exhibit commendable mechanical strength, corrosion resistance, and biocompatibility, while their non-degradability and high-cost pose limitations.^9,10 Calcium phosphate, a significant bone constituent, can release calcium and phosphate and offer bone healing capabilities; however, its poor fracture toughness and osteoinductivity restrict its utility in bone cement.^11,12 This limitation often requires the material to be combined with osteoinductive agents, such as bone morphogenetic proteins (BMPs), to improve its efficacy in bone regeneration. Polymer scaffolds play a pivotal role in tissue guidance and growth factor delivery, yet many lack robust self-osteoconductivity, self-osteogenesis, and mechanical strength compared to metal or ceramic-based materials.^13–15 Consequently, there is a pressing demand for a composite structure combining biodegradability, osteoinductivity, osteoconductivity, and osteogenesis properties, featuring inherent voids for bone defect remediation.

Casein (CS), abundant in proline amino acids and devoid of disulfide bridges, tends to self-assemble, forming micelles in aqueous solutions supersaturated with calcium phosphate.¹⁶ These micelles, stabilized by calcium ions and hydrophobic interactions, exhibit a highly hydrated interior, creating a microporous scaffold structure. Various structural models of casein micelles have been proposed,^17,18 with a prevalent model suggesting that α_s- and β-caseins, phosphorylated and bound to calcium phosphate nanoclusters, constitute the micelle core. At the same time, κ-caseins predominantly occupy the surface, impeding micelle aggregation. Thus, casein micelles,¹⁹ negatively charged in neutral and alkaline environments, offer opportunities for composite fabrication via self-assembly and stabilization of CaCO₃ microsphere crystallization.²⁰ Moreover, casein has shown biodegradability and enhances the osteoinductive differentiation of bone marrow mesenchymal stem cells, contributing to bone regeneration.^16,20 Nonetheless, its mechanical strength is insufficient for standalone bone defect repair.

Calcium carbonate (CaCO₃), widely utilized in bone cement composites, boasts biocompatibility,²¹ osteoconductivity,²² and favorable mechanical properties.²³ Spherical CaCO₃ microspheres, with their porous structure and ample surface area, are conducive to drug loading. Various molecules have been employed to regulate CaCO₃ formation, such as type I-collagen,²⁴ sodium myristate,²⁵ lipopeptide biosurfactants,²⁶ and linear polymers.²⁷ Yet, few synthesized polymorphs exhibit osteoinductivity, a crucial trait for bone repair materials.

The superparamagnetic material, magnetic Fe₃O₄ nanoparticle, characterized by its notable biocompatibility and ability to respond to magnetic fields, has witnessed a surge in its utilization across a broad spectrum of biological applications.^28,29 These include magnetic biosensing, magnetic imaging, magnetic separation, and the delivery of drugs and genes, as well as hyperthermia therapy.^30–32 Surface-modified Fe₃O₄@SiO₂ nanoparticles exhibit heightened stability³³ and enhanced biocompatibility³⁴ compared to their uncoated counterparts. In addition, research suggests that magnetic Fe₃O₄ nanoparticles, when combined with appropriate magnetic stimulation, possess the capacity to stimulate osteogenesis and facilitate the formation of bone tissue.³⁵ 

Herein, we designed a novel kind of magnetic casein/CaCO₃/Fe₃O₄ microspheres (CCFM), in which casein micelles were used as an osteoinductive scaffold to form osteoconductive calcium carbonate microspheres and to load osteogenic magnetic Fe₃O₄ nanoparticles. CaCO₃ microspheres without casein and casein/CaCO₃ microspheres without Fe₃O₄ nanoparticles were synthesized as contrasts. The products were characterized by several analytical techniques, including scanning electron microscopy (SEM), transmission electron microscopy (TEM), zeta potential, x-ray diffraction (XRD), Fourier transform infrared (FTIR) spectroscopy, and thermogravimetric analysis (TGA). The in vitro bioactivity of CCFM confirmed osteogenic differentiation behaviors using MC3T3-E1 cells. This new kind of osteoinductive, osteoconductive, and osteogenic magnetic platform would be a promising tool for bone regeneration in clinical practice.

Iron chloride hexahydrate (FeCl₃·6H₂O), trisodium citrate dihydrate (TC), sodium acetate anhydrous (NaAc), ammonia solution (NH₃·H₂O, 28%), tetraethyl orthosilicate (TEOS), isopropyl alcohol, and (3-aminopropyl) triethoxysilane (APTES) were purchased from Aladdin Chemicals (China). Ethanol and ethylene glycol (EG) were supplied by Sinopharm Chemical Reagent Co., Ltd. (Shanghai, China). Sodium carbonate (Na₂CO₃) and calcium chloride (CaCl₂) were purchased from Tianjin Zhiyuan Reagent Co., Ltd. (Tianjin, China). Poly(acrylic acid) (PAA, Mw∼1800) was acquired from Sigma-Aldrich (St. Louis, MO, USA). Casein (CA) was purchased from Macklin (China). The ultrathin carbon support film for TEM analysis was obtained from Guangzhou Zhongjingkeyi Technology Co., Ltd. All the reagents are of analytical grade and were used without further purification. Deionized water (DI water) was used in all experiments.

At first, magnetite (Fe₃O₄) nanoparticles were fabricated by a high-pressure hydrothermal strategy modified from our previous publications.^36,37 Trisodium citrate dehydrate (0.318 g) and FeCl₃·6H₂O (0.819 g) were dissolved in 15 ml EG and stirred for 30 minutes to obtain solution A and solution B, respectively. NaAc (1.5 g), DI water (1 ml), and solution B were added to solution A under continuous stirring for 15 minutes to obtain solution C. Then solution C was transferred to a Teflon-line autoclave (50 ml) and treated at 200 °C for 10 h. The obtained products were isolated with a magnet, subsequently washed with ethanol and DI water several times, and then dried in a vacuum oven for 2 h for further processing. Successively, Fe₃O₄@SiO₂ core–shell colloids were prepared through a modified Stöber process. Prepared Fe₃O₄ nanoparticles (50 mg) were suspended in ethanol (40 ml) and DI water (6 ml) by ultrasonication for 5 min. The solution was added to an ammonia solution (4 ml) and then transferred into a three-neck flask in a water bath under vigorous mechanical stirring. TEOS (120 µl) was added to the flask every 20 min until the total amount of TEOS reached 240 µl. After mechanical stirring for 1 h, the Fe₃O₄@SiO₂ products were collected through magnetic separation, washed with ethanol and DI water three times, and finally dried at 40 °C in a vacuum. Then, the Fe₃O₄@SiO₂ particles (50 mg) were dissolved in a mixture of isopropyl alcohol (40 ml) and (3-aminopropyl) triethoxysilane (APTES) (100 µl) and heated up to 80 °C for 2 h to functionalize the silica surface with amino groups. Fe₃O₄@SiO₂–NH₂ nanoparticles were washed with isopropyl alcohol and dispersed in DI water.

Solutions of CaCl₂ and Na₂CO₃ at a concentration of 0.1M were prepared first. Fe₃O₄@SiO₂–NH₂ solution (5 mg/ml) was obtained by ultrasonic dispersion. Fe₃O₄@SiO₂–NH₂ solution (600 µl) was added to the CaCl₂ solution (20 ml) under stirring for 20 min. Casein (160 mg) was dissolved in a Na₂CO₃ solution (20 ml). Na₂CO₃/casein solution was dripped into the CaCl₂ solution under continuous stirring at 600 rpm. After addition, the mixture was agitated for 30 min. The prepared magnetic casein/CaCO₃/Fe₃O₄ microspheres (CCFM) were collected by centrifugation, followed by washing several times with DI water, freeze-dried, and stored at room temperature as control groups; casein/CaCO₃ microspheres (CCM), without adding Fe₃O₄@SiO₂–NH₂, and PAA/CaCO₃ microspheres (PCM), where PAA was used as nucleation micelle, were also prepared.

Thermal field emission environmental scanning electron microscopy (TFE-SEM, Quanta 400F, Oxford, UK) was applied to observe the shape and surface morphology of microspheres. The Gemini SEM 500 (Gemini 500, Zeiss, Germany) was used to study the more detailed structure of the surface of the microspheres. Transmission electron microscopy (TEM) images were obtained by using JEM-2100F equipment (JEOL, Japan) at an accelerating voltage of 200 kV. The zeta potential of Fe₃O₄@SiO₂–NH₂ nanoparticles and magnetic casein/CaCO₃/Fe₃O₄ microspheres was measured with the Zetasizer Nano-ZS90 (Malvern, UK). The mineral deposition was performed by x-ray diffraction (XRD, Empyrean, PANalytical B.V., the Netherlands) with a Cu K source to identify the crystalline structure. The presence of casein within the synthesized microspheres was confirmed by Fourier transform infrared (FTIR) spectroscopy (NICOLET 6700, Thermo Fisher Scientific, USA) and thermogravimetric analysis curves (TGA, TG209F1 Libra, Netzsch, Germany). TGA analysis was performed within a temperature range of 30–900 °C and with a rate of increasing temperature of 10 °C/min under nitrogen protection. To evaluate in vitro mineralization, the microspheres were soaked in 1× simulated body fluid (SBF) and incubated at 37 °C for 14 days. The morphologies of mineralized agglomerates were recorded by SEM.

MC3T3-E1 cells were seeded in α-Minimal Essential Medium (α-MEM, Gibco, USA) supplemented with 10% fetal bovine serum (FBS), 100 mg/ml streptomycin, and 100 U/ml penicillin (all from Gibco, USA) in an atmosphere of 100% humidity and 5% CO₂ at 37 °C, with the medium changing every 3 days. The cells were digested by TrypLE Express (Gibco, USA) when reaching 80% confluence. Lyophilized microspheres were exposed to ultraviolet light for 5 h at the same time. Successively, the microspheres were washed three times with PBS and incubated with a complete culture medium overnight.

Cell proliferation was evaluated with the Cell Counting Kit-8 (CCK-8, Dojindo, Japan) according to the manufacturer’s instructions. Cells were seeded at a density of 3 × 10³ cells per well in 96-well plates, with each sample prepared in triplicate and treated with different microspheres (CaCO₃ microspheres, CCM, and CCFM, while the control group was treated with the same amount of PBS) at a final concentration of 0.25 mg/ml. At 24, 48, and 72 h after microspheres were treated, cell proliferation was determined using CCK-8 reagent by incubating cells at 37 °C for 30 min. The optical density was measured with a multi-mode microplate reader (SynergyHTX, Bio-Tek, USA) at a wavelength of 450 nm (650 nm for reference). All results were confirmed by repeating the experiment three times.

Alkaline phosphatase (ALP) staining and alizarin red staining (ARS) were performed to evaluate the early and late osteogenic differentiation of MC3T3-E1 cells, respectively. Cells were treated with microspheres (CaCO₃ microspheres, CCM, and CCFM, while the control group was treated with the same amount of PBS) after being cultured in 12-well plates for 24 h. Another 24 h later, the medium was replaced with an osteogenic medium composed of α-MEM supplemented with 10% FBS, antibiotics (100 mg/ml streptomycin and 100 U/ml penicillin), 100 nM dexamethasone, 10 mM β-glycerophosphate, and 50 µM L-ascorbic acid (all from Sigma-Aldrich, USA). The medium was changed every 3 days. ALP staining and ARS staining were carried out using kits (ALP staining, Beyotime, China; ARS staining, Cyagen, China) according to the manufacturer’s instructions on the seventh and 14th days, respectively.

Quantitative data were presented as mean (standard deviation). Statistical analysis was performed by one-way analysis of variance (ANOVA) with Dunnett’s post hoc tests using the SPSS 20.0 software. A P value of <0.05 was considered statistically significant.

In this study, magnetic casein/CaCO₃/Fe₃O₄ microspheres (CCFM) are prepared by electrostatic self-assembly between positively charged Fe₃O₄@SiO₂–NH₂ nanoparticles and casein/CaCO₃ microspheres. Magnetic nanoparticles were coated with SiO₂, and then, the silica surface was functionalized with amino groups.^38, Figure 1(a) shows the TEM characterization of Fe₃O₄@SiO₂–NH₂ nanoparticles, which are multilayer spheres with a Fe₃O₄ core and silica shell with a narrow size distribution of 100–150 nm. As shown in Fig. 2(a), the –NH₂ terminated Fe₃O₄@SiO₂ colloids display a positive potential of 41.1 mV, indicating the existence of amino groups on the surface of the nanoparticles. Casein/CaCO₃ microspheres (CCM) were fabricated using a fast precipitation method with casein micelles as a template, which is a promising bone repair material with favorable biocompatibility, biodegradability,²⁰ osteoinductivity, and osteoconductivity.³⁹ The positive-charged Fe₃O₄@SiO₂–NH₂ nanoparticles were assembled on the surface of CCM through electrostatic interaction to form magnetic CCFM, which was observed by TEM characterization [Fig. 1(b)].

To characterize the morphology of the microspheres and verify the template function of casein during the crystal growth of CaCO₃ microspheres, different microspheres were observed under SEM, including PAA/CaCO₃ microspheres (PCM), CCM, and CCFM. The PCM were in ellipsoidal shapes [Fig. 1(c)] with smooth surfaces [Fig. 1(f)]. CCM [Fig. 1(d)] and magnetic CCFM [Fig. 1(e)] both had an average diameter of ∼12 µm and exhibited a rough surface composed of CaCO₃ nanocrystals interspersed with irregular nanopores [Figs. 1(g) and 1(h)], which confirmed the speculation of the formation mechanism of CCM in a previous study.²⁰ In brief, casein micelles were negatively charged outside with –COO^- and positively charged in the interior with –NH₃⁺. Hence, CO₃²⁻ was located much more in the interior than on the exterior due to the electrostatic attraction between –NH₃⁺ and CO₃²⁻ ions after casein was dissolved in the NaCO₃ solution. Ca²⁺ and CO₃²⁻ were bonded in situ to form CaCO₃ nanocrystals, which were continuously deposited to fabricate micrometer-scale CCM. In this paper, casein/NaCO₃ solution was added to Fe₃O₄/CaCl₂ solution drop by drop, which led to the casein/CO₃²⁻ micelles being surrounded by a high concentration of Ca²⁺ ions and Fe₃O₄@SiO₂–NH₂ nanoparticles at an earlier stage. Therefore, the average size of CCM fabricated in prophase was larger than in anaphase, which required centrifugal collection to obtain microspheres of uniform size. During the formation of CCM, Fe₃O₄@SiO₂–NH₂ nanoparticles with positive charge were electrostatically adsorbed on the periphery of the casein micelles by self-assembly [Figs. 1(e) and 1(h)].

X-ray diffraction (XRD) was used to investigate the crystal phase of various microspheres. As shown in Fig. 2(b), only calcite (PDF No. 47–1743) peaks were present in CaCO₃ with PAA, while both calcite and vaterite (PDF No. 33–0268) diffraction peaks were detected in CCFM and CCM, indicating that casein is related to the formation of vaterite. Casein micelles are essential in controlling nucleation and affecting the deposition of vaterite crystals, as well as maintaining the spherical shape of CCM.²⁰ There was little calcite present in CCM in other reported work, in which casein/NaCO₃ solution was quickly added to the CaCl₂ solution.²⁰ The following is the explanation for the difference: Due to the high concentration of calcium ions in the reaction system at the initial stage, some free carbonate ions combine with calcium ions to form amorphous CaCO₃ nanoparticles when casein/NaCO₃ solution drops into Fe₃O₄/CaCl₂ solution.⁴⁰ Some of the amorphous nanoparticles were immediately transformed into calcite and deposited on the surface of the formed microspheres to increase their size. In addition, calcite crystals have been formed in the presence of PAA.⁴¹ Therefore, both calcite and vaterite were observed in CCFM and CCM.

The function groups in the samples were analyzed by Fourier transform infrared (FTIR) spectrum analysis [Fig. 2(c)]. The absorption bands of CO₃²⁻ at 870 cm⁻¹ and between 1400 and 1500 cm⁻¹ were observed. While the absorption peak of –CONH– at 1650 cm⁻¹ exists in CCM and CCFM except for PCM, the characteristic absorption bands for calcite at 1800 and 706 cm⁻¹ were all observed in the spectra of three microsphere samples.⁴² The results confirm that casein exists in CCM and CCFM.

The thermogravimetry (TGA) results are displayed in Fig. 2(d). The decomposition of CCM and CCFM was presented in three stages. The first stage showed an initial weight loss of 2.5% due to the dissipation of the coordinated water molecules, which was also observed in the curve of pure casein with a 5% initial weight loss. The second stage, occurring at temperatures ranging from 176 to 620 °C, corresponded to the decomposition of casein, which was reported to be thermally stable at temperatures 0–176 °C.⁴³ Notably, the casein decomposition stage was absent in the curve of CaCO₃ with PAA (PCM). The decomposition of calcium carbonate, starting at 620 °C, occurred in the third stage. The results of TGA further support the presence of casein in CCM and CCFM. In addition, the residue percentages of CCM and CCFM were 42.5% and 48.9%, respectively, indicating the existence of Fe elements in magnetic casein/CaCO₃/Fe₃O₄ microspheres.

Figure 3 shows the magnetic response of CCFM under an external magnetic field. The saturation magnetization strength of CCFM is 5.6 emu g⁻¹. The results indicate that CCFM are superparamagnetic due to the loading of Fe₃O₄@SiO₂–NH₂ nanoparticles. Therefore, we successfully prepared the magnetic casein/CaCO₃/Fe₃O₄ microspheres for further study of their osteoinductivity in vitro.

For analysis of in vitro osteoinductive bioactivity, the samples were soaked in 1× simulated body fluid (SBF) at 37 °C for 14 days and characterized under SEM. The sample of PCM displayed no obvious change in surface morphology, as shown in Figs. 4(a) and 4(d). In contrast, plenty of agglomerates interconnecting each other, such as vegetable sponges, were observed on the surfaces of CCM and CCFM [Figs. 4(b) and 4(c)]. Magnified images [Figs. 4(e) and 4(f)] showed that these agglomerates have the same morphology as the bone-like apatite reported in the literature.⁴⁴ These results indicate that the CaCO₃ microspheres with casein formed apatite on their surfaces in SBF. Thus, it can be speculated that these microspheres will bond to living bone through the apatite layer in the living body.⁴⁴ 

Figure 5(a) shows the structure of the Fe₃O₄@SiO₂–NH₂ nanoparticle. The silica-coated Fe₃O₄ nanoparticle is functionalized with amino acids.⁴⁵ The formation mechanism of ellipsoidal CaCO₃ in the presence of PAA is displayed in Fig. 5(b).⁴¹ Donnet et al. proved that there are four stages to the growth mechanism of the seeded calcite precipitation at the high concentration of PAA, including the formation of an amorphous hydrous CaCO₃ gel, hydrous amorphous CaCO₃ transforming into calcite through growth on the calcite seeds stabilized by PAA, the agglomeration of primary particles, and PAA inhibiting the ripening of calcite crystallite.⁴¹ The calcite crystallite is coated with CaCO₃ nanocrystallites with no alignment among the crystals. Magnetic casein/CaCO₃/Fe₃O₄ microspheres are comprised of vaterite and calcite, as shown in Fig. 5(c). CO₃²⁻ are located in the interior of casein micelles. Ca²⁺ is absorbed into the PAA chain. When CaCl₂ solution is added to Na₂CO₃ solution, Ca²⁺ is bonded with CO₃²⁻ and nucleates in situ to form a vaterite crystal with casein micelle as a template. Positively charged Fe₃O₄@SiO₂–NH₂ nanoparticles are loading on the exterior of casein micelles through electrostatic attraction. The vaterite and calcite crystals are growing with the agglomeration of more and more CaCO₃ nanocrystallites. Finally, magnetic casein/CaCO₃/Fe₃O₄ microspheres are formed with the deposition of vaterite and calcite crystals.

In order to evaluate the effect of CCFM on cell viability and proliferation, cytotoxicity assays were performed using the CCK-8 method. MC3T3-E1 cells were cultured with CaCO₃ microspheres, CCM, and CCFM for 72 h using cells grown on tissue culture polystyrenes (TCPS) as a control. The measured data were obtained at 24, 48, and 72 h after the microspheres were seeded. As shown in Fig. 6, there was no significant difference between groups, although the cell viability of the group with CCFM displayed a slightly descending trend compared to the control group. The results indicated that co-culture with microspheres of proper concentration did not promote cell proliferation; it also did not lead to adverse effects on cell viability.

The effects of various CaCO₃ microspheres on osteogenic differentiation of MC3T3-E1 cells were evaluated by in vitro osteoinductive studies in comparison with TCPS as a control. ALP staining and Alizarin Red S staining were adopted as golden standards to assess the osteoinductive activity in different stages of the biological processes of osteogenic differentiation after seven and 14 days of cell culture, respectively. The biological processes of osteogenic differentiation are defined as three periods: cell proliferation, extracellular matrix (ECM) maturation, and mineralization.⁴⁶ Each period is characterized by the expression of distinctive osteogenic markers such as ALP activity and calcium deposition, which are early markers of immature osteoblast activity and later markers of mature osteoblasts, respectively.⁴⁷ As shown in Figs. 7(a)–7(d), ALP staining intensity was highest for CCM and CCFM, followed by CaCO₃ microspheres, showing a significant difference from the control group. The ARS staining data showed more calcium deposition in CCM and CCFM than in CaCO₃ microspheres with PAA and the control group [Figs. 7(e)–7(h)]. Therefore, it can be inferred that casein may enhance the expression of ALP and the formation of calcium sediment.

In this study, magnetic casein/CaCO₃/Fe₃O₄ microspheres (CCFM) were fabricated with osteoinductive casein micelles as a template and osteoconductive CaCO3 as the backbone and loaded with osteogenic Fe₃O₄@SiO₂–NH₂ nanoparticles. The appropriate concentration of CCFM stimulated the osteogenic differentiation of MC3T3-E1 cells and showed no apparent cytotoxicity. The formation of apatite on the surface after microspheres were soaked in SBF confirmed that CCFM have the potential to promote bone formation. Therefore, this new type of magnetic casein/CaCO₃/Fe₃O₄ microspheres is expected to converge around the lesion location and contribute to bone repair.

This work was supported by the National Natural Science Foundation of China (Grant Nos. 81571020, 81771124, 61376014, 61775242, and 61835013). The authors gratefully acknowledge the technical assistance from Dr. Jiahong Lv, Dr. Jingyi Xue, Dr. Qianli Li, and Mr. Shaohang Xie.

The authors have no conflicts to disclose.

Mingjie Zhang: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Validation (equal); Visualization (equal); Writing – original draft (lead); Writing – review & editing (equal). Xiaolei Li: Conceptualization (equal); Data curation (equal); Formal analysis (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Validation (equal); Visualization (equal); Writing – original draft (equal); Writing – review & editing (equal). Han Lin: Conceptualization (equal); Data curation (equal); Funding acquisition (equal); Investigation (equal); Methodology (equal); Project administration (equal); Resources (equal); Software (equal); Supervision (equal); Writing – original draft (equal); Writing – review & editing (equal).

The data that support the findings of this study are available from the corresponding author upon reasonable request.