Synthesis and enhanced bone regeneration of carbonate substituted octacalcium phosphate

Abstract

Using a wet method, we have synthesized octacalcium phosphate carbonate, in which HPO₄²⁻ in octacalcium phosphate is replaced with CO₃²⁻. The physical, crystal, and chemical properties of this new material were compared to octacalcium phosphate, Ca-deficient hydroxyapatite, and Ca-deficient carbonate apatite using X-ray diffraction, Fourier-transform infrared spectroscopy, inductively coupled plasma spectroscopy, and scanning electron microscopy. Surface roughness and morphology were also characterized, along with the ability to support proliferation and differentiation of MG63 cells, as measured by MTT and alkaline phosphatase assay. We found that octacalcium phosphate carbonate enhanced osteoblast proliferation more strongly than all other materials tested. Similarly, Ca-deficient carbonate apatite, a hydrolysate of octacalcium phosphate carbonate, stimulated osteoblast differentiation to a better extent than Ca-deficient hydroxyapatite, a carbonate-free hydrolysate of octacalcium phosphate. These results indicate that octacalcium phosphate carbonate has good biocompatibility and osteoconduction, and incorporation of carbonate into octacalcium phosphate and apatite enhances bone regeneration.

Keywords

Octacalcium phosphate calcium-deficient hydroxyapatite carbonate substitution osteoblast bone regeneration

1. Introduction

Among bone substitutes, materials based on calcium phosphate are widely used, because of good biocompatibility and osteoconductivity [1–4]. Examples of such materials include hydroxyapatite and β-tricalcium phosphate [2,5,6]. Hydroxyapatite is well established to be osteoconductive, while β-tricalcium phosphate was found to have good biodegradability following nucleation of new bone [5–7]. Recently, octacalcium phosphate (OCP) has attracted increasing attention because of good osteoconductivity, biocompatibility, and biodegradability [7–11]. In particular, OCP is considered to be a better scaffold for bone regeneration in comparison to hydroxyapatite, which is almost non-absorbable, or to β-tricalcium phosphate, which is absorbed too quickly [6,7,9,12]. Thus, the physicochemical properties of OCP have been extensively investigated, with a view to enhancing its biological properties.

OCP was first suggested by Brown et al. [13,14] to be a precursor to bones and teeth. Synthetic OCP may have thin, plate-like, ribbon, strip-like, and acicular morphology, depending on the environment in which the crystals grow, pH, mode of synthesis, and presence of various inorganic ions and organic molecules [15–19]. In any case, the crystal structure of OCP is similar to that of hydroxyapatite, and includes a water and an apatite layer [20]. OCP is converted to apatite by hydrolysis, which decreases HPO₄²⁻ content, and increases the Ca/P molar ratio [8,21]. Importantly, Suzuki et al. [22,23] found that synthetic OCP enhances osteoblast differentiation and osteoclast formation. Indeed, OCP enhanced bone regeneration better than insoluble unstinted hydroxyapatite or Ca-deficient hydroxyapatite, a product of OCP hydrolysis [12,24]. Moreover, OCP is more resorbable than β-tricalcium phosphate after hydrolysis, a process that may dissolve OCP in situ, and/or deposit hydroxyapatite [7].

Octacalcium phosphate carbonate (OCPC), a kind of octacalcium phosphate that contains carbonate, was first observed by Hayek et al. [16] during a reaction of CaHPO₄ with CaCO₃ at 37°C, in which CO₃²⁻ was suggested to replace HPO₄²⁻ in apatite as the reaction progressed. The final carbonate-substituted product was then considered to be much more similar to bone. Notably, Pellegrino and Biltz [25] detected OCPC during formation of renal calculi in a patient with primary hyperparathyroidism, and proposed that after dicalcium phosphate dihydrate is hydrolyzed to OCP, OCP reacts with CO₃²⁻ to form OCPC. In addition, Brown et al. [26] concluded that carbonate was incorporated as OCP is hydrolyzed into hydroxyapatite. Indeed, OCP is believed to promote incorporation of impurities into hydroxyapatite [26].

The cellular response to a substrate depends on its physical and chemical characteristics, particularly crystallinity, particle size, surface structure, and chemical composition [27]. For example, formation of new bone around the typical OCP crystal was found to be higher than around a highly elongated OCP crystal [28]. In addition, Suzuki et al. [24] found that low-crystallinity OCP enhanced bone formation better than OCP or fully hydrolyzed OCP. On the other hand, carbonate apatite has been suggested to have significantly greater impact than hydroxyapatite on cell attachment, proliferation, and differentiation. Thus, incorporation of carbonate into OCP may affect its biological activity.

The physical and chemical properties of OCPC, including crystal structure, morphology, and chemical properties, have not been characterized. For instance, the osteoblast response to carbonate apatite has been investigated, but not to OCPC. OCP and hydrolyzed OCPC have also been evaluated, but not the physicochemical and biological effects of CO₃²⁻ incorporation. We have now synthesized OCPC through a wet method, and characterized its morphology, crystal structure, and chemical composition. We then used MG63 cells to investigate the osteoblastic response to OCPC, OCP, and their corresponding hydrolysates, Ca-deficient carbonate apatite and Ca-deficient hydroxyapatite.

2. Materials and methods

2.1. Synthesis and preparation of substrates

Reagent-grade chemicals were obtained from Wako Pure Chemical Industries, Ltd. OCP was synthesized by dropwise addition of calcium acetate (250 mL, 0.04 mol, and pH 4.8) into sodium dihydrogen phosphate (250 mL, 0.03 mol, and pH 7.4) at 60°C. The resulting mixture was incubated at 60°C for another 3 hours without stirring, and the final pH was 4.9. Ca-deficient hydroxyapatite was synthesized in a similar manner, except that the mixture was stirred at 70°C for 24 hours, and the final pH was 4.5. Calcium hydrogen phosphate dihydrate, calcium carbonate, and sodium hydrogen carbonate were mixed at molar ratio 5:2:1 to prepare OCPC and Ca-deficient carbonate apatite. The aqueous suspension was stirred at 60°C for 2 hours to prepare OCPC with final pH 6.9, or for 24 hours to prepare Ca-deficient carbonate apatite with final pH 7.6. All products were washed 3 times with distilled water, split into 6 tubes, oven-dried at 60°C for 3 days, and finally ground into powder. Pellets ( $d = 1 cm$ , $h = 0.3 cm$ ) were prepared by pressing 0.05 g powdered products at 200 MPa into cylindrical molds in a standard evacuable pellet die, and dried overnight in an oven at 100°C.

2.2. Physicochemical characterization

2.2.1. Scanning electron microscopy, X-ray diffraction, FTIR, ICP spectroscopy

Powdered products were imaged on an S-3400NX scanning electronic microscope (Hitachi). X-ray diffraction patterns were collected at 3–45° in 0.02° steps using a Ni-filtered $Cu K α$ diffractometer (D8 Advance, Bruker AXS) operating at 40 kV and 40 mA. Unit cell parameters were determined by refinement of X-ray diffraction data. For Fourier-transform infrared spectroscopy on an FT/IR-4100 (Jasco), powdered products (1 mg) were mixed with 200 mg KBr powder, and pressed into thin pellets in a stainless steel cast. FTIR spectra were recorded from 4000 to 400 cm⁻¹. Chemical composition was determined using an inductively coupled plasma-atomic emission spectrometer (ICPS-7000 ver 2, Shimadzu).

2.2.2. Surface roughness

Pelleted products were washed with 70% ethanol in a 10 cm dish, then with MilliQ water twice, and finally dried at room temperature on a clean bench. Surface roughness was measured 50 times with pitch 0.02 using a VK-8500 laser microscope (KEYENCE). Measurements were obtained in five pellets of each product, at five different spots per pellet.

2.2.3. Conversion test

OCP and OCPC pellets were sterilized, dried, immersed in a 24-well plate containing αMEM (Wako) supplemented with 10% fetal bovine serum and 1% penicillin-streptomycin (Wako), and incubated at 37°C and 5% CO₂. Culture media were refreshed every three days. After 11 days, pellets were quickly washed with phosphate-buffered saline and MilliQ water once each, and dried on a clean bench. X-ray diffraction patterns were then collected as described.

2.3. Cell culture

The human osteosarcoma cell line MG63 was cultured at 37°C, 95% air, and 5% CO₂ in αMEM containing 10% fetal bovine serum and 1% penicillin-streptomycin. Cells were seeded at $0.8 \times 10^{4}$ cells/well and $1.5 \times 10^{4}$ cells/well in 24-well plates containing pellets ( $d = 1 cm$ ) of OCP, OCPC, Ca-deficient hydroxyapatite, and Ca-deficient carbonate apatite. Glass sheets ( $d = 1.2 cm$ ) were used as control substrate. Culture media were refreshed twice a week in cultures seeded at $0.8 \times 10^{4}$ cells/well, and cell proliferation was evaluated by MTT assay. For cultures seeded at $1.5 \times 10^{4}$ cells/well, media were changed after one day to αMEM containing 10% fetal bovine serum, 1% penicillin-streptomycin, 0.05 mg/mL ascorbic acid, 5 mM β-glycerol phosphate, and $10^{- 6} M$ dexamethasone, and then refreshed twice per week. Differentiation of MG63 cells to osteoblasts was evaluated by alkaline phosphatase activity.

2.4. Cell proliferation

MTT was dissolved in phosphate-buffered saline at 5 mg/mL, and mixed at 1:10 v:v with αMEM containing 10% fetal bovine serum and 1% penicillin-streptomycin. Cells seeded at $0.8 \times 10^{4}$ cells/well and grown on ceramic pellets and glass sheets for 1, 3, and 6 days were labeled with $500 µ L$ MTT medium for 3 hours at 37°C and 5% CO₂. Substrates were then quickly washed with phosphate-buffered saline, moved to a new 24-well plate, and immersed in $100 µ L$ DMSO to dissolve purple precipitates. The optical density of the resulting solution was measured at 570 nm using microplate reader.

2.5. Cell differentiation

Pellets and glass sheets seeded with $1.5 \times 10^{4}$ cells/well and grown for 6, 10, and 14 days were quickly washed twice with phosphate-buffered saline, transferred to a new 24-well plate, and immersed in $150 µ L$ 0.2% TritonX. Cells were then lysed over two cycles of freezing at $- 4$ °C, thawing at room temperature, and shaking for 1 min on a microplate shaker. Lysates were then processed with Micro BCA™ Protein Assay Kit to determine total protein, as measured at 570 nm and in comparison to bovine serum albumin standards. Samples were also processed with LabAssay™ ALP kit to determine alkaline phosphatase activity, as measured at 415 nm and in comparison to p-nitrophenol standards. Relative alkaline phosphatase activity was calculated as $µ mol/ µ g protein = activity (µ mol/ µ L) / total protein (µ g/ µ L)$ .

2.6. Statistical analysis

MTT, alkaline phosphatase, and surface roughness data were collected from five samples of each substrate. Substrates were then compared by one-way ANOVA, with $p < 0.05$ considered to indicate a statistically significant difference.

3. Results

3.1. Surface characterization

3.1.1. Morphology

SEM images of powdered OCP, OCPC, Ca-deficient hydroxyapatite, and Ca-deficient carbonate apatite are shown in Fig. 1. Synthetic OCP powder consists of long, thin crystals (Fig. 1(a)), while synthetic OCPC powder consists of short plates and flakes (Fig. 1(c)). On the other hand, synthetic Ca-deficient hydroxyapatite is composed of aggregated needle-like crystals (Fig. 1(b)). Aggregated needle-like crystals are also observed in synthetic Ca-deficient carbonate apatite powder, along with plate-like crystals (Fig. 1(d)).

Fig. 1.

Scanning electron micrographs of OCP (a), Ca-deficient hydroxyapatite (b), OCPC (c), and Ca-deficient carbonate apatite (d) powders. Scale bars, 10 µm.

3.1.2. X-ray diffraction

X-ray diffraction patterns collected from powdered samples of OCP (curve a), OCPC (curve b), Ca-deficient hydroxyapatite (curve c), and Ca-deficient carbonate apatite (curve d) are presented in Fig. 2. The diffraction patterns of synthetic OCP and OCPC were identical to the OCP diffraction pattern deposited with the ICDD No. 26-1056, as shown in Fig. 2(B), in which characteristic OCP diffraction peaks are marked with black arrows. However, the relative intensities of diffraction peaks were different between OCP and OCPC. On the other hand, the diffraction patterns of Ca-deficient hydroxyapatite and Ca-deficient carbonate apatite were identical to that of hydroxyapatite (ICDD No. 09-0432), although both seemed to be less crystalline.

Fig. 2.

X-ray diffraction patterns collected from powdered OCP (curve a), OCPC (curve b), Ca-deficient hydroxyapatite (curve c), and Ca-deficient carbonate apatite (curve d, A). Black arrows mark characteristic peaks of OCP/OCPC, and red arrows mark characteristic peaks of Ca-deficient hydroxyapatite/carbonate apatite. Panel (B) is a zoomed-in view of the red box in (A), encompassing $2 θ$ 24° to 36°.

Table 1

Main reflections for OCP, OCPC, Ca-deficient hydroxyapatite and Ca-deficient carbonate apatite

OCP/OCPC	OCP^∗	CDHA/CDCA	HA^∗
4.8	4.7
9.4	9.4
9.8	9.8
		10.8	10.8
25.9	25.5	25.9	25.9
26.0	26.0
31.5	31.5
31.7	31.7	31.7	31.8
32.1/32.2	32.2	32.2	32.2
		32.9	32.9
33.5	33.5
		34	34

OCP^∗: Pattern PDF 26-1056 Radiation: 1.54060 Quality: Star.

HA^∗: Pattern PDF 09-0432 Radiation: 1.54060 Quality: Indexed.

CDHA: Ca-deficient hydroxyapatite.

CDCA: Ca-deficient carbonate apatite.

The key reflections that separate OCP and OCPC from their Ca-deficient hydrolysates are listed in Table 1. In particular, the intensity of peaks 010, 020, 110, and 070 in OCP and OCPC decreased as hydrolysis progressed, although these peaks were still detectable just before total hydrolysis (Fig. 2). In addition, the doublet peaks 2–21 and 002 in OCP and OCPC transformed into the single peak 002 after hydrolysis. Further, peaks 260, 2–41, and −1–42 in OCP and OCPC disappeared after hydrolysis, while peaks 211, 112, and 300 appeared in the Ca-deficient hydrolysis products (Fig. 2(B)). The lattice constants of OCP were $a = 9.55 (1)$ Å, $b = 19.03 (1) \overset{˚}{A}$ , $c = 6.85 (1) \overset{˚}{A}$ , $α = 92.8 (1) °$ , $β = 90.2 (1) °$ , and $γ = 79.7 (1) °$ . In OCPC, the lattice constants were $a = 9.524 (7) \overset{˚}{A}$ , $b = 19.013 (6) \overset{˚}{A}$ , $c = 6.872 (4) \overset{˚}{A}$ , $α = 92.28 (6) °$ , $β = 89.93 (6) °$ , and $γ = 79.75 (4) °$ . The c-axis was slightly longer in OCPC than in OCP, while the a-axis was shorter. The lattice constants of Ca-deficient hydroxyapatite were $a = 9.434 (8) \overset{˚}{A}$ and $c = 6.879 (6) \overset{˚}{A}$ , and were $a = 9.438 (4) \overset{˚}{A}$ and $c = 6.889 (2) \overset{˚}{A}$ in Ca-deficient carbonate apatite. As can be seen, the lattice constants are nearly equivalent in Ca-deficient hydrolysates. The lattice constants for OCP, OCPC, Ca-deficient hydroxyapatite and Ca-deficient carbonate apatite are listed in Table 2.

Table 2

Lattice constants for OCP, OCPC, Ca-deficient hydroxyapatite and Ca-deficient carbonate apatite

	$a (\overset{˚}{A})$	$b (\overset{˚}{A})$	$c (\overset{˚}{A})$	$α (°)$	$β (°)$	$γ (°)$
OCP	9.55(1)	19.03(1)	6.85(1)	92.8(1)	90.2(1)	79.7(1)
OCPC	9.524(7)	19.013(6)	6.872(4)	92.28(6)	89.93(6)	79.75(4)
CDHA	9.434(8)	–	6.879(6)	90	90	120
CDCA	9.438(4)	–	6.889(2)	90	90	120

CDHA: Ca-deficient hydroxyapatite.

CDCA: Ca-deficient carbonate apatite.

Fig. 3.

FTIR spectra of synthetic OCP (a), OCPC (b), Ca-deficient hydroxyapatite (c), and Ca-deficient carbonate apatite (d). Black frames mark vibrations due to HPO₄⁻. Red frames mark vibrations due to CO₃²⁻, and black arrows indicate type A and type B CO₃²⁻. Blue frames mark vibrations due to OH⁻.

3.1.3. Fourier-transform infrared spectroscopy

The FTIR spectra of synthetic OCP, OCPC, and their Ca-deficient hydrolysates are presented in Fig. 3, and all relevant vibrations are listed in Table 3. Characteristic bands at 912 and 1291 cm⁻¹ are due to HPO₄²⁻ (Fig. 3, black box), which were clearly observed in OCP and OCPC despite low intensity. However, the intensity of HPO₄²⁻ bands was weaker in OCPC than in OCP. Additionally, bands at 1419 and 1456 cm⁻¹, observed in OCPC and its Ca-deficient hydrolysate, are due to type B CO₃²⁻ (Fig. 3, red box), while bands at 1540 cm⁻¹ are attributed to type A CO₃²⁻ (Fig. 3, red box). OH⁻ bands at 3567 cm⁻¹ were also observed in both Ca-deficient hydrolysates (Fig. 3, blue box). In general, spectra were comparable between OCPC and OCP, and between Ca-deficient carbonate apatite and Ca-deficient hydroxyapatite, except that OCPC and its hydrolysate contain carbonate bands. Taken together, these results indicate that OCPC contains type B carbonate within an OCP-like structure, and that Ca-deficient carbonate apatite also contains carbonate (mainly type B and some type A) in a structure that is otherwise similar to Ca-deficient hydroxyapatite.

Table 3
FT–IR vibrations for OCP, OCPC, Ca-deficient hydroxyapatite and Ca-deficient carbonate apatite

Vibration Mode OCP OCPC CDHA CDCA

PO₄³⁻ comb. 471w 471 471

522sh

552

PO₄³⁻ 562 563 563

PO₄³⁻ $v_{4}$ 601 601 603

614

640sh 632

HPO₄²⁻ 862 862

CO₃²⁻ $v_{2}$ 872 872

HPO₄²⁻ 912 913w

P–O 960 960 960

1010 1024s 1024s

1025s 1022s

1035

1059s 1063s 1069

1105 1099 1105s 1101s

1134 1138sh

1183

1206sh

HPO₄²⁻ 1286vw 1291vw

CO₃²⁻ $v_{3}$ 1419 1417

CO₃²⁻ $v_{3}$ 1456 1455

CO₃²⁻ $v_{3}$ 1541 1540

H₂O $v_{2}$ 1636 1636 1635

H₂O 3410–3450b 3410–3450b

OH⁻ 3567 3567vw

Vibration	Mode	OCP	OCPC	CDHA	CDCA
PO₄³⁻	comb.	471w		471	471
	522sh
	552
PO₄³⁻			562	563	563
PO₄³⁻	$v_{4}$		601	601	603
	614
	640sh		632
HPO₄²⁻		862		862
CO₃²⁻	$v_{2}$		872		872
HPO₄²⁻		912	913w
P–O		960	960	960
	1010		1024s	1024s
	1025s	1022s
		1035
	1059s	1063s		1069
	1105	1099	1105s	1101s
	1134		1138sh
	1183
			1206sh
HPO₄²⁻		1286vw	1291vw
CO₃²⁻	$v_{3}$		1419		1417
CO₃²⁻	$v_{3}$		1456		1455
CO₃²⁻	$v_{3}$		1541		1540
H₂O	$v_{2}$	1636	1636		1635
H₂O		3410–3450b	3410–3450b
OH⁻				3567	3567vw

vw, very weak; w, weak; m, medium; s, strong; sh, shoulder; b, broad.

CDHA: Ca-deficient hydroxyapatite.

CDCA: Ca-deficient carbonate apatite.

3.1.4. ICP–AES, surface roughness, and conversion of OCP/OCPC

The Ca/P molar ratio in OCP, OCPC, Ca-deficient hydroxyapatite, and Ca-deficient carbonate apatite was 1.32, 1.45, 1.46, and 1.5, respectively. The Ca/P molar ratio of the OCP we synthesized is slightly lower than the theoretical ratio of 1.33. The results also confirm that the OCP/OCPC hydrolysates are calcium deficient, as the theoretical ratio in hydroxyapatite is 1.67. Sodium was not detected. The mean surface roughness ± SD of OCP, OCPC, Ca-deficient hydroxyapatite, and Ca-deficient carbonate apatite was comparable at $0.305 \pm 0.02$ , $0.344 \pm 0.02$ , $0.364 \pm 0.05$ , and $0.340 \pm 0.02 µ m$ , respectively ( $p < 0.05$ ). Finally, there were no obvious changes in intensity in X-ray diffraction patterns of unseeded OCP and OCPC immersed for 11 days in culture media.

3.2. Cell proliferation

As shown in Fig. 4, MTT values increased from day 1 to day 6, indicating that pellets of OCP, OCPC, and their Ca-deficient hydrolysates are biocompatible, and did not negatively affect the viability and proliferation of MG63 cells. Results were comparable at day 1 and day 3, but proliferation on OCPC was significantly higher at day 6 than on any other substrate. At this time point, proliferation on Ca-deficient carbonate apatite was higher than on OCP, and also higher on OCP than on Ca-deficient hydroxyapatite ( $p < 0.05$ ). These results indicate that proliferation of MG63 cells was enhanced by day 6 on OCPC, Ca-deficient carbonate apatite, OCP, and Ca-deficient hydroxyapatite, in this order.

Fig. 4.

Proliferation of MG63 cells cultured on OCP (a), OCPC (b), Ca-deficient hydroxyapatite (c), Ca-deficient carbonate apatite (d) pellets, as measured by MTT assay and using glass as control. Error bars are SD ( $n = 5$ ).

3.3. Cell differentiation

Cell differentiation was assessed by alkaline phosphatase activity in MG63 cells cultured for 6, 10, and 14 days on pellets of OCP, OCPC, and their Ca-deficient hydrolysates. As shown in Fig. 5, activity increased from day 6 to day 10, and decreased at day 14. At day 10, activity was higher in cultures grown on OCPC than on OCP, but not to a statistically significant extent. However, activity was significantly higher at days 6 and 10 in cells growing on OCP, OCPC, and Ca-deficient carbonate apatite than on Ca-deficient hydroxyapatite, suggesting enhanced MG63 differentiation.

Fig. 5.

Changes in alkaline phosphatase activity in MG63 cells cultured on OCP (a), OCPC (b), Ca-deficient hydroxyapatite (c), Ca-deficient carbonate apatite (d) pellets, using glass as control. Error bars represent SD ( $n = 5$ ).

4. Discussion

In this study, we synthesized OCPC, examined cellular proliferation and differentiation on OCP, OCPC, and their Ca-deficient hydrolysates, and compared their physicochemical characteristics. Results (Figs 4 and 5) confirmed that carbonate incorporation into OCP/apatite should promote bone regeneration.

Synthetic OCPC formed short plate-like crystals in powder form (Fig. 1(c)), while OCP formed long, plate-like crystals (Fig. 1(a)). The difference in length is likely due to environmental conditions during synthesis. In particular, OCPC was obtained as an intermediate product during hydrolysis of dicalcium phosphate dihydrate in the presence of a small amount of NaHCO₃, which changed the pH of the reaction, and thus altered the environment in which crystals grew, especially at the beginning of the reaction. Indeed, the effects of synthesis conditions on OCP morphology has been reported [29]. Notably, the OCPC crystals were similar to plate-like OCP crystals that have been previously obtained [2,8,29–31], as well as to OCPC described by Hayek [16]. Thus, we conclude that the difference in crystal length between OCPC and OCP is insignificant, and that the crystal morphologies are similar.

Accordingly, X-ray diffraction patterns suggested that the crystal structure of OCPC is similar to that of OCP. As shown in Fig. 2, there were no differences in X-ray diffraction patterns between OCP and OCPC, and between their Ca-deficient hydrolysates. However, reflections 002 and 300 shifted to lower angles in Ca-deficient carbonate apatite, in comparison to Ca-deficient hydroxyapatite. Similar shifts to lower angles from OCP to OCPC were also observed in reflections 010 and 002. These results indicate that incorporation of carbonate into OCP and Ca-deficient hydroxyapatite slightly changes the structure of the lattice. Indeed, a slight expansion of the c-axis and a contraction of the a-axis (Fig. 2) were observed in OCPC, as has been reported when PO₄³⁻ is replaced with CO₃²⁻ in crystals of carbonate apatite. Nevertheless, we consider the observed changes in lattice constants to be negligible.

These changes were also detectable by FTIR, through which carbonate bands were observed both in OCPC and its hydrolysate Ca-deficient carbonate apatite (Fig. 3), although absorbance bands at 1419, 1456, and 1540 cm⁻¹ indicate that some HPO₄²⁻ had been retained (Fig. 3, curve b). In hydroxyapatite, carbonate substitutions are classified as type A (CO₃²⁻ for PO₄³⁻) or type B (CO₃²⁻ for OH⁻), which are observed at 1450 and 1540 cm⁻¹ and at 1420 and 1460 cm⁻¹, respectively. As OCP contains apatite and water layers, the results indicate that carbonate may substitute for HPO₄²⁻ in the water layer or for PO₄³⁻ in the apatite layer, thereby generating OCPC. Notably, the FTIR spectra of Ca-deficient carbonate apatite (Fig. 3, curve d) contains a carbonate absorbance band at 1540 cm⁻¹ that is more definitive than that of OCPC, as well as an OH⁻ absorbance band less definitive than that of Ca-deficient hydroxyapatite. A similar phenomenon in carbonate apatite has led to the notion of a type AB substitution, in which carbonate substitutions occur not only at PO₄³⁻ sites but also at OH⁻ sites [32]. Hence, the results indicate that we have successfully synthesized a carbonate-substituted OCPC, as well as a Ca-deficient carbonate apatite with type AB substitutions.

The Ca/P molar ratio of synthetic OCP was 1.32, which is very close to the theoretical ratio of 1.33. Thus, we appear to have obtained a nearly stoichiometric OCP. Similarly, LeGeros et al. [33] synthesized OCP via precipitation, and obtained material with Ca/P molar ratio from 1.325 to $1.335 \pm 0.005$ . On the other hand, OCPC had Ca/P molar ratio 1.45, an increase that may be attributed to the loss HPO₄²⁻ through substitution with CO₃². Indeed, Hayek [34] reported that the Ca/P molar ratio of OCPC ranged from 1.48 to 1.67 depending on carbonate content. In addition, ICP–AES analysis did not detect sodium in OCPC, which, in theory, has chemical formula ${Ca}_{8} {({HPO}_{4})}_{1.52} {({PO}_{4})}_{4} {({CO}_{3})}_{0.48} \cdot 5 H_{2} O$ . On the other hand, the Ca/P molar ratio of Ca-deficient hydroxyapatite and Ca-deficient carbonate apatite was 1.46 and 1.5, respectively. Indeed, hydrolysis of OCP into calcium-deficient hydroxyapatite increases the Ca/P molar ratio from $\sim 1.23 – 1.26$ to $\sim 1.49$ [8,30,35] as a result of HPO₄²⁻ release and absorption of Ca²⁺ [31]. Similarly, the increase in Ca/P molar ratio for Ca-deficient carbonate apatite may be attributed to continuous CO₃²⁻ incorporation into OCPC and/or release of HPO₄²⁻. Nevertheless, all values are lower than the theoretical ratio of 1.66, confirming that these hydrolysates were calcium deficient.

Notably, OCPC enhanced osteoblast proliferation (Fig. 4) and differentiation (Fig. 5) in vitro, in comparison to all other substrates, and especially at day 6. Accordingly, Ca-deficient carbonate apatite was significantly more effective at stimulating cell proliferation than Ca-deficient hydroxyapatite (Fig. 4). Alkaline phosphatase activity was also higher in Ca-deficient carbonate apatite than in Ca-deficient hydroxyapatite, but was comparable between OCPC and OCP at day 6 and 10 (Fig. 5). Notably, cell proliferation is sensitive to surface roughness [36,37]. However, surface roughness was comparable among all materials tested, and surface morphology was comparable between materials with (OCPC and hydrolysate, Fig. 1, curves d and h) or without carbonate (OCP and hydrolysate, Fig. 1, curves b and f). Thus, differences in cell proliferation cannot be attributed to differences in crystal structure, surface roughness, and morphology, but perhaps to differences in chemical composition. For example, incorporation of magnesium and strontium into OCP increased cell proliferation and differentiation, and enhanced ALP activity [38], while carbonate apatite promote osteoblastic differentiation, and enhanced ALP activity earlier than sintered hydroxyapatite [39]. Indeed, various theories have been proposed to explain how carbonate may improve osteoconductivity. For instance, carbonate may increase solubility [40], enhance protein adsorption and early cell attachment [41], or upregulate bone-related genes such as collagen and osteocalcin. In addition, incorporation of CO₃²⁻ increases the ratio of calcium to phosphate, and thereby regulate osteoblast activity and bone formation [42,43]. Collectively, these observations indicate that carbonate apatite may stimulate bone growth [44]. However, further studies in vivo are necessary to test this hypothesis.

We note that OCPC may transform into Ca-deficient carbonate apatite during synthesis, given that OCP also tends to convert into apatite both in vitro and in vivo [8,45–47]. However, OCPC and OCP immersed in culture media did not appear to convert. This may be due to (1) in situ hydrolysis of OCP into hydroxyapatite, and/or (2) dissolution of OCP, followed by hydroxyapatite precipitation and formation of new bone, as observed in vivo [46–49].

Further, we note that physiological apatite is an ion-substituted nonstoichiometric form of hydroxyapatite, and contains 4–8 wt% carbonate ions. Nevertheless, hydroxyapatite has been widely used in the clinic as bone substitute, and the osteoconductivity and biocompatibility of carbonate apatite have also been investigated. Similarly, OCP has been shown to possess good osteoconductivity and biodegradability, and is thought to be a precursor of mineral crystals in bone and tooth [50]. Based on all data, we anticipate that OCPC will also become a bone substitute to treat bone defects.

5. Conclusions

We have demonstrated that carbonate incorporation into OCP/apatite may promote bone regeneration. In particular, results indicate that carbonate is not only incorporated into apatite, but is also incorporated into OCP to form the intermediate product OCPC, which enhances osteoblastic proliferation and differentiation in vitro. These studies may help elucidate the effect of carbonates incorporated into OCP/apatite, as well as characterize OCPC as bone substitute.

Footnotes

Acknowledgements

We acknowledge generous assistance from all teachers and students at Department of Inorganic Biomaterials and Department of Pediatric Dentistry at Tokyo Medical and Dental University.

Conflict of interest

The authors have no conflict of interest to report.

References

LeGeros

R.Z.

, Properties of osteoconductive biomaterials: Calcium phosphates, Clin Orthop395 (2002), 81–98. doi:10.1097/00003086-200202000-00009.

Suzuki

Nakamura

Miyasaka

Kagayama

and Sakurai

, Bone formation on synthetic precursors of hydroxyapatite, Tohoku J Exp Med164 (1991), 37–50. doi:10.1620/tjem.164.37.

Bucholz

R.W.

, Nonallograft osteoconductive bone graft substitutes, Clin Orthop395 (2002), 44–52. doi:10.1097/00003086-200202000-00006.

Damien

C.J.

and Parsons

J.R.

, Bone graft and bone graft substitutes: A review of current technology and applications, Journal of Applied Biomaterials2 (1991), 187–208. doi:10.1002/jab.770020307.

Bucholz

R.W.

Carlton

and Holmes

R.E.

, Hydroxyapatite and tricalcium phosphate bone graft substitutes, Orthop Clin North Am18 (1987), 323–334.

Metsger

D.S.

Driskell

and Paulsrud

, Tricalcium phosphate ceramic – A resorbable bone implant: Review and current status, J Am Dent Assoc105 (1982), 1035–1038. doi:10.14219/jada.archive.1982.0408.

Kamakura

Sasano

Shimizu

Hatori

Suzuki

Kagayama

and Motegi

, Implanted octacalcium phosphate is more resorbable than β-tricalcium phosphate and hydroxyapatite, J Biomed Mater Res59 (2002), 29–34. doi:10.1002/jbm.1213.

Suzuki

Kamakura

Katagiri

Nakamura

Zhao

Honda

and Kamijo

, Bone formation enhanced by implanted octacalcium phosphate involving conversion into Ca-deficient hydroxyapatite, Biomaterials27 (2006), 2671–2681. doi:10.1016/j.biomaterials.2005.12.004.

Suzuki

, Octacalcium phosphate: Osteoconductivity and crystal chemistry, Acta Biomaterialia6 (2010), 3379–3387. doi:10.1016/j.actbio.2010.04.002.

10.

Hak

D.J.

, The use of osteoconductive bone graft substitutes in orthopaedic trauma, J Am Acad Orthop Surg15 (2007), 525–536. doi:10.5435/00124635-200709000-00003.

11.

Suzuki

, Octacalcium phosphate (OCP)-based bone substitute materials, Japanese Dental Science Review49 (2013), 58–71. doi:10.1016/j.jdsr.2013.01.001.

12.

Imaizumi

Sakurai

Kashimoto

Kikawa

and Suzuki

, Comparative study on osteoconductivity by synthetic octacalcium phosphate and sintered hydroxyapatite in rabbit bone marrow, Calcif Tissue Int78 (2006), 45–54. doi:10.1007/s00223-005-0170-0.

13.

Brown

W.E.

Eidelman

and Tomazic

, Octacalcium phosphate as a precursor in biomineral formation, Adv Dent Res1 (1987), 306–313. doi:10.1177/08959374870010022201.

14.

Eidelman

and Eanes

, Role of OCP in biological processes, Monogr Oral Sci18 (2001), 50–76. doi:10.1159/000061648.

15.

Brown

W.E.

Lehr

J.R.

Smith

J.P.

and Frazier

A.W.

, Crystallography of octacalcium phosphate, J Am Chem Soc79 (1957), 5318–5319.

16.

Hayek

, Die Mineralsubstanz der Knochen, Journal of Molecular Medicine45 (1967), 857–863.

17.

Salimi

Heughebaert

and Nancollas

, Crystal growth of calcium phosphates in the presence of magnesium ions, Langmuir1 (1985), 119–122. doi:10.1021/la00061a019.

18.

Bigi

Bracci

Panzavolta

Iliescu

Plouet-Richard

Werckmann

and Cam

, Morphological and structural modifications of octacalcium phosphate induced by poly-L-aspartate. Crystal Growth & Design4 (2004), 141–146.

19.

Bigi

Gazzano

Ripamonti

and Roveri

, Effect of foreign ions on the conversion of brushite and octacalcium phosphate into hydroxyapatite, J Inorg Biochem32 (1988), 251–257. doi:10.1016/0162-0134(88)85004-9.

20.

Brown

W.E.

, Octacalcium phosphate and hydroxyapatite: Crystal structure of octacalcium phosphate, Nature196 (1962), 1048–1050. doi:10.1038/1961048b0.

21.

Chow

L.C.

and Eanes

E.D.

, Octacalcium Phosphate, Karger Medical and Scientific Publishers, 2001.

22.

Sasano

Kamakura

Homma

Suzuki

Mizoguchi

and Kagayama

, Implanted octacalcium phosphate (OCP) stimulates osteogenesis by osteoblastic cells and/or committed osteoprogenitors in rat calvarial periosteum, Anat Rec256 (1999), 1–6. doi:10.1002/(SICI)1097-0185(19990901)256:1<1::AID-AR1>3.0.CO;2-X.

23.

Takami

Mochizuki

Yamada

Tachi

Zhao

Miyamoto

Anada

Honda

Inoue

and Nakamura

, Osteoclast differentiation induced by synthetic octacalcium phosphate through receptor activator of NF-κB ligand expression in osteoblasts. Tissue Engineering Part A15 (2009), 3991–4000.

24.

Miyatake

Kishimoto

K.N.

Anada

Imaizumi

Itoi

and Suzuki

, Effect of partial hydrolysis of octacalcium phosphate on its osteoconductive characteristics, Biomaterials30 (2009), 1005–1014. doi:10.1016/j.biomaterials.2008.10.058.

25.

Pellegrino

E.D.

and Biltz

R.M.

, Calcium phosphate carbonate transformations in renal calculi, Calcif Tissue Int26 (1978), 191–193. doi:10.1007/BF02013255.

26.

Chickerur

Tung

and Brown

, A mechanism for incorporation of carbonate into apatite, Calcif Tissue Int32 (1980), 55–62. doi:10.1007/BF02408521.

27.

Kawahara

, Cellular responses to implant materials: Biological, physical and chemical factors, Int Dent J33 (1983), 350–375.

28.

Honda

Anada

Kamakura

Morimoto

Kuriyagawa

and Suzuki

, The effect of microstructure of octacalcium phosphate on the bone regenerative property. Tissue Engineering Part A15 (2009), 1965–1973.

29.

Mathew

and Takagi

, Crystal structures of calcium orthophosphates, Octacalcium Phosphate18 (2001), 1–16. doi:10.1159/000061646.

30.

Suzuki

, Adsorption of bovine serum albumin onto octacalcium phosphate and its hydrolyzates, Cell and Materials5 (1995), 45–54.

31.

Suzuki

Yagishita

Amano

and Aoba

, Reversible structural changes of octacalcium phosphate and labile acid phosphate, J Dent Res74 (1995), 1764–1769. doi:10.1177/00220345950740110801.

32.

Fleet

M.E.

and Liu

, Local structure of channel ions in carbonate apatite, Biomaterials26 (2005), 7548–7554. doi:10.1016/j.biomaterials.2005.05.025.

33.

LeGeros

R.Z.

, Preparation of octacalcium phosphate (OCP): A direct fast method, Calcif Tissue Int37 (1985), 194–197. doi:10.1007/BF02554841.

34.

Hayek

, Die Mineralsubstanz der Knochen, Journal of Molecular Medicine45 (1967), 857–863.

35.

Suzuki

Kamakura

and Katagiri

, Surface chemistry and biological responses to synthetic octacalcium phosphate, Journal of Biomedical Materials Research Part B: Applied Biomaterials77 (2006), 201–212. doi:10.1002/jbm.b.30407.

36.

Deligianni

D.D.

Katsala

N.D.

Koutsoukos

P.G.

and Missirlis

Y.F.

, Effect of surface roughness of hydroxyapatite on human bone marrow cell adhesion, proliferation, differentiation and detachment strength, Biomaterials22 (2000), 87–96. doi:10.1016/S0142-9612(00)00174-5.

37.

Lincks

Boyan

B.D.

Blanchard

C.R.

Lohmann

C.H.

Liu

Cochran

D.L.

Dean

D.D.

and Schwartz

, Response of MG63 osteoblast-like cells to titanium and titanium alloy is dependent on surface roughness and composition, Biomaterials19 (1998), 2219–2232. doi:10.1016/S0142-9612(98)00144-6.

38.

Boanini

Torricelli

Fini

Sima

Serban

Mihailescu

I.N.

and Bigi

, Magnesium and strontium doped octacalcium phosphate thin films by matrix assisted pulsed laser evaporation, J Inorg Biochem107 (2012), 65–72. doi:10.1016/j.jinorgbio.2011.11.003.

39.

Nagai

Kobayashi-Fujioka

Fujisawa

Ohe

Takamaru

Hara

Uchida

Tamatani

Ishikawa

and Miyamoto

, Effects of low crystalline carbonate apatite on proliferation and osteoblastic differentiation of human bone marrow cells, J Mater Sci Mater Med26 (2015), 1–8. doi:10.1016/j.msec.2015.07.028.

40.

Lajeunesse

Kiebzak

G.M.

Frondoza

and Sacktor

, Regulation of osteocalcin secretion by human primary bone cells and by the human osteosarcoma cell line MG-63, Bone Miner14 (1991), 237–250. doi:10.1016/0169-6009(91)90025-U.

41.

Spence

Patel

Brooks

and Rushton

, Carbonate substituted hydroxyapatite: Resorption by osteoclasts modifies the osteoblastic response, Journal of Biomedical Materials Research Part A90 (2009), 217–224. doi:10.1002/jbm.a.32083.

42.

Bellows

Heersche

and Aubin

, Inorganic phosphate added exogenously or released from β-glycerophosphate initiates mineralization of osteoid nodules in vitro, Bone Miner17 (1992), 15–29. doi:10.1016/0169-6009(92)90707-K.

43.

Dvorak

M.M.

Siddiqua

Ward

D.T.

Carter

D.H.

Dallas

S.L.

Nemeth

E.F.

and Riccardi

, Physiological changes in extracellular calcium concentration directly control osteoblast function in the absence of calciotropic hormones, Proc Natl Acad Sci USA101 (2004), 5140–5145. doi:10.1073/pnas.0306141101.

44.

Rupani

Hidalgo-Bastida

L.A.

Rutten

Dent

Turner

and Cartmell

, Osteoblast activity on carbonated hydroxyapatite, Journal of Biomedical Materials Research Part A100 (2012), 1089–1096. doi:10.1002/jbm.a.34037.

45.

Johnsson

M.S.

and Nancollas

G.H.

, The role of brushite and octacalcium phosphate in apatite formation, Crit Rev Oral Biol Med3 (1992), 61–82. doi:10.1177/10454411920030010601.

46.

Ban

Jinde

and Hasegawa

, Phase transformation of octacalcium phosphate in vivo and in vitro, Dent Mater J11 (1992), 130–140. doi:10.4012/dmj.11.130.

47.

LeGeros

R.Z.

Daculsi

Orly

Abergas

and Torres

, Solution-mediated transformation of octacalcium phosphate (OCP) to apatite, Scanning Microsc3 (1989), 129–137; discussion 137–138.

48.

Iijima

and Moriwaki

, Effects of ionic inflow and organic matrix on crystal growth of octacalcium phosphate; relevant to tooth enamel formation, J Cryst Growth198–199 (1999), Part 1, 670–676.

49.

Suzuki

, Adsorption of bovine serum albumin onto octacalcium phosphate and its hydrolyzates, Cell and Materials5 (1995), 45–54.

50.

Brown

W.E.

Smith

J.P.

Lehr

J.R.

and Frazier

A.W.

, Octacalcium phosphate and hydroxyapatite: Crystallographic and chemical relations between octacalcium phosphate and hydroxyapatite, Nature196 (1962), 1050–1055. doi:10.1038/1961050a0.