Carbonate apatite artificial bone

ABSTRACT

Bone apatite is not hydroxyapatite (HAp), it is carbonate apatite (CO₃Ap), which contains 6–9 mass% carbonate in an apatitic structure. The CO₃Ap block cannot be fabricated by sintering because of its thermal decomposition at the sintering temperature. Chemically pure (100%) CO₃Ap artificial bone was recently fabricated through a dissolution–precipitation reaction in an aqueous solution using a precursor, such as a calcium carbonate block. In this paper, methods of fabricating CO₃Ap artificial bone are reviewed along with their clinical and animal results. CO₃Ap artificial bone is resorbed by osteoclasts and upregulates the differentiation of osteoblasts. As a result, CO₃Ap demonstrates much higher osteoconductivity than HAp and is replaced by new bone via bone remodeling. Granular-type CO₃Ap artificial bone was approved for clinical use in Japan in 2017. Honeycomb-type CO₃Ap artificial bone is fabricated using an extruder and a CaCO₃ honeycomb block as a precursor. Honeycomb CO₃Ap artificial bone allows vertical bone augmentation. A CO₃Ap-coated titanium plate has also been fabricated using a CaCO₃-coated titanium plate as a precursor. The adhesive strength is as high as 76.8 MPa, with excellent tissue response and high osteoconductivity.

Graphical abstract

1. Bone apatite

Animals with a skeleton are classified as invertebrates or vertebrates. The skeleton of invertebrates is composed of calcium carbonate (CaCO₃) and is likely derived from elements found in seawater. In contrast, vertebrates, including humans, have a skeleton of carbonate apatite [CO₃Ap: Ca_10-a(PO₄)_6-b(CO₃)_c] instead of CaCO₃ [1–5].

A key difference between CaCO₃ and CO₃Ap skeletons is phosphorous. Phosphorous or phosphate is important in energy metabolism in the process of generating energy or adenosine triphosphate (ATP). Invertebrates can use phosphates present in seawater, even though their concentrations are low (1.3 μmol/L). However, vertebrates living on the land need to store phosphorous in the body. During evolution from invertebrates to vertebrates, bone became the storage organ of phosphorous in vertebrates. In other words, CO₃Ap was chosen as bone apatite as a result of evolution from invertebrates to vertebrates.

CO₃Ap or bone apatite should be amenable for use as artificial bone. However, sintered hydroxyapatite (HAp) instead of CO₃Ap has been used as a typical artificial bone since the 1970s. This is because the thermal decomposition of CO₃Ap begins at approximately 400°C, which prevents the fabrication of sintered CO₃Ap. Recently, 100% chemically pure CO₃Ap blocks were fabricated through a dissolution–precipitation reaction in an aqueous solution using a CaCO₃ block as a precursor.

2. CO₃Ap fabrication though dissolution–precipitation reaction using a precursor

There are three key requirements for compositional transformation through the dissolution–precipitation reaction. First, the solubility of the precursor should be higher than that of the final product. Second, any component that is lacking must be supplied from the aqueous solution. Third, precipitates or crystals of the final product should have the ability to interlock with one another to maintain the shape of the block.

CO₃Ap is a thermodynamically stable phase under physiological conditions. It is not soluble under physiological conditions. Moreover, CO₃Ap crystals can interlock. Many compounds are more soluble than CO₃Ap, and thus can be precursors for the fabrication of CO₃Ap through dissolution–precipitation reactions including CaCO₃ [6–18], α-tricalcium phosphate [19–24], dicalcium phosphate dihydrate [25,26], and CaSO₄ [27–30]. Moreover, a component that is lacking can be supplied from aqueous solution.

One of the ideal precursors is CaCO₃. It has low solubility in aqueous solution at neutral pH and contains both calcium and carbonate. Chemically pure CaCO₃ blocks can be easily fabricated by simply exposing calcium hydroxide [Ca(OH)₂] compact to carbon dioxide (CO₂) [eq. (1)].

$$\mathrm{C}\mathrm{a}{\left(\mathrm{O}\mathrm{H}\right)}_2+\mathrm{C}{\mathrm{O}}_2\to \mathrm{C}\mathrm{a}\mathrm{C}{\mathrm{O}}_3+{\mathrm{H}}_2\mathrm{O}$$ (1)

The compositional transformation of CO₃Ap from CaCO₃ requires phosphate. The CaCO₃ block needs to be immersed in a phosphate salt solution for this compositional transformation.

When CaCO₃ is immersed in an aqueous solution, it dissolves and supplies Ca²⁺ and CO₃^2- [eq. (2)]. If other ions are absent, the water becomes saturated with CaCO₃. However, the situation is different when water contains PO₄³⁻. In that case, the phosphate salt aqueous solution can be supersaturated with respect to CO₃Ap when both Ca²⁺ and CO₃^2- are supplied by the dissolution of CaCO₃. Thus, Ca²⁺, PO₄^3−, and CO₃^2- precipitate as CO₃Ap [eq (3)], and the precipitated CO₃Ap crystals interlock with one another. Continuous dissolution–precipitation reactions lead to the compositional transformation from CaCO₃ to CO₃Ap, maintaining the macroscopic structure of the precursor.

$$\mathrm{C}\mathrm{a}\mathrm{C}{\mathrm{O}}_3\to \mathrm{C}{\mathrm{a}}^{2+}+\mathrm{C}{{\mathrm{O}}_3}^{2-}$$ (2)

$$\mathrm{C}{\mathrm{a}}^{2+}+\mathrm{P}{{\mathrm{O}}_4}^{3-}+\mathrm{C}{{\mathrm{O}}_3}^{2-}\to \mathrm{C}{\mathrm{a}}_{10-\mathrm{a}}{\left(\mathrm{P}{\mathrm{O}}_4\right)}_{6-\mathrm{b}}{\left(\mathrm{C}{\mathrm{O}}_3\right)}_{\mathrm{c}}$$ (3)

Other useful precursors are CaSO₄ · 2H₂O, α-tricalcium phosphate [α-TCP: Ca₃(PO₄)₂], and dicalcium phosphate dihydrate [DCPD: CaHPO₄ · 2H₂O]. However, sulfate ions tend to remain in the apatitic structure when CaSO₄ · 2H₂O is used as a precursor. CO₃Ap containing a small amount of HAp tends to form when α-TCP or DCPD are used as precursors because of the competitive reaction to form CO₃Ap and HAp. Therefore, CaCO₃ appears to be an ideal precursor. Fabrication of CO₃Ap based on compositional transformation through a dissolution–precipitation reaction using CaCO₃ as a precursor may have arisen during the evolution of vertebrates from invertebrates.

3. CO₃Ap and bone remodeling

3.1. Bone remodeling process

Bone remodeling involves the replacement of old bone and autograft by new bone. Osteoclasts resorb the old bone or autograft, followed by the formation of new bone by osteoblasts (Figure 1). Apatite is osteoconductive. Therefore, osteoblasts are active on their surfaces, although the degree of activity can differ depending on the type of apatite. The activity of osteoclasts also differs according to the type of apatite. Figure 2 displays representative scanning electron microscopy (SEM) images of osteoclasts incubated on the surfaces of bone, HAp, and CO₃Ap. The absence of osteoclastic resorption with HAp [6] is evidence that HAp cannot be replaced with new bone because osteoclast resorption does not occur. In contrast, osteoclastic resorption occurs for bone and CO₃Ap.

Figure 1.

Graphical image of bone remodeling performed by osteoclasts and osteoblasts

Figure 2.

SEM images of bone, sintered HAp, and CO₃Ap when osteoclasts were incubated on their surfaces [6]

Osteoclasts form Howship’s lacunae and decrease the pH inside the lacunae to pH 3–5, leading to the dissolution of bone apatite. Thus, osteoclasts resorb apatite by dissolving it using a weak acid (Figure 3).

Figure 3.

Graphical image of osteoclasts

Figure 4 summarizes the solubilities of HAp and CO₃Ap as a function of pH. At physiological pH or pH 7.4, CO₃Ap is thermodynamically the most stable phase. This may explain why bone apatite is CO₃Ap. However, under weakly acidic conditions produced by osteoclasts or at a pH of 3–5, the solubility of CO₃Ap is higher than that of HAp. Therefore, CO₃Ap dissolves in weakly acidic conditions and is resorbed by osteoclasts, whereas HAp does not appreciably dissolve under weakly acidic conditions and is not resorbed by osteoclasts.

Figure 4.

Solubility of carbonate apatite and hydroxyapatite in body fluid in terms of calcium concentration as a function of pH [14]

3.2. Differentiation of osteoblasts

Osteoblastic activity is the counterpart of osteoclastic activity in bone remodeling (Figure 3). One of the parameters of osteoblastic activity is differentiation. Osteoblastic differentiation markers include type I collagen, alkaline phosphatase, osteopontin, and osteocalcin (Figure 5) [31]. Human bone marrow cells (hBMCs) incubated on CO₃Ap demonstrated much higher expression than HAp. Upregulation of osteoblast differentiation is likely one of the causes of the higher osteoconductivity of CO₃Ap artificial bone, in addition to the activation of osteoblasts through cell-cell interactions with osteoclasts.

Figure 5.

Relative gene expression levels of type I collagen, alkaline phosphatase, osteopontin, and osteocalcin on the surface of plastic well, CO₃Ap, and sintered HAp [31]

3.3. Histological findings

Figure 6 summarizes the representative results of Villanueva Goldner (VG) histologic staining comparison of CO₃Ap and sintered HAp (Neobone®) used for reconstruction of mandibular bone defects adjacent to dental implants in beagle dogs 4 weeks after reconstruction surgery [32]. In VG staining, mature bone is stained green. Reconstruction of a defect with HAp involves the formation of only a limited amount of bone at the defect site, on the surface of the bone defect, and adjacent to the dental implant. This is one reason why no artificial bones have been approved for implant-related bone defect reconstruction surgeries in Japan. In contrast, more bone forms even at the center of bone defects reconstructed using CO₃Ap. The surfaces of bone defects and dental implants become covered with bone. The documented bonding that results between bone, CO₃Ap, and dental implant clearly indicating the usefulness of CO₃Ap as an artificial bone for dental implants.

Figure 6.

VG-stained histological image of sintered HAp (Neobone®) and CO₃Ap 4 weeks after reconstruction surgery of beagle dogs’ mandibular bone defect adjacent to a dental implant [32]

3.4. Clinical trials

The first human clinical trial was performed at three university hospitals in patients requiring sinus floor augmentation [33,34]. Figure 7 illustrates the procedure for two-stage sinus floor augmentation [34]. After the sinus floor membrane was elevated with a mucosal elevator, CO₃Ap granules were placed in the elevated space (Figure 7(b)). Implant placement was planned for 8 ± 2 months after the augmentation. Prior to implantation, a bone biopsy can be performed using a trephine bur (Figure 7(c)).

Figure 7.

Diagram of two-stage sinus floor augmentation for dental implantation. (a) before operation, (b) filling CO₃Ap granule after elevation of sinus floor membrane, (c) before biopsy which was performed 8 ± 2 months after the surgery, (d) after dental implantation

Figure 8 summarizes the micro-computed tomography (μ-CT) images and the appearance of biopsy tissue stained with hematoxylin and eosin (H-E) or VG. CO₃Ap granules were replaced with new bone, even though a small amount of CO₃Ap granules remain at this stage [34]. The presence of both mature bone and osteoids indicated active bone remodeling. Few inflammatory cells or foreign-body giant cells were observed in the biopsy specimens. The mean preoperative residual bone height of 3.4 ± 1.3 mm was increased to 13.0 ± 1.9 mm by the sinus floor augmentation using the CO₃Ap granules. Since the height after sinus floor augmentation is sufficient for dental implants, all patients received dental implants without any problems [33,34]. Based on the trial results, CO₃Ap granules were approved for clinical use as Cytrans Granules (GC Co, Tokyo, Japan) by the Pharmaceuticals and Medical Devices Agency (PMDA) in 2017. The chemically pure CO₃Ap granules are the first to be commercially available globally and the first artificial bone that can be used for bone reconstruction aimed at dental implantation in Japan.

Figure 8.

Photo, μ-CT scanning, H-E and VG staining of a histological specimen biopsied from a patient who underwent two-stage sinus floor augmentation for dental implantation [34]

4. CO₃Ap honeycomb artificial bone

Not only composition but also architecture play important roles in governing the ability of artificial bone. In other words, regulation of architecture may be one of the important keys for artificial bone to demonstrate the similar osteoconductivity of autograft. One of the attractive architectures is the honeycomb. Fabrication of the CO₃Ap honeycomb by the dissolution–precipitation reaction requires a precursor. In general, a honeycomb is fabricated by extruding a raw material through a honeycomb die (Figure 9). An organic binder is necessary for extrusion. Therefore, the organic binder must be eliminated in the subsequent debindering.

Figure 9.

Diagram of honeycomb fabrication process and SEM images of CaCO₃ and CO₃Ap honeycombs

The CaCO₃ honeycomb becomes a CO₃Ap honeycomb by immersion in a phosphate salt aqueous solution based on compositional transformation through a dissolution–precipitation reaction that maintains the honeycomb structure. Figure 9 shows typical SEM images of CaCO₃ and CO₃Ap honeycombs. The macroscopic honeycomb structure was maintained during the compositional transformation. However, the microstructure is changed during the dissolution–precipitation reaction, indicating a dependence on the crystal structure of each honeycomb composition.

Figure 10 displays a representative histological image one month after reconstruction of the femoral bone defect using the CO₃Ap honeycomb [13]. Tissue penetration into all pores of CO₃Ap honeycomb is evident. At higher magnification, the new bone formation at the pore surface of CO₃Ap honeycomb is evident. The presence of osteoclasts and osteoblasts on the surface of the newly formed bone indicates active bone remodeling. Osteocytes are also observed within the bone matrix. Interestingly, numerous vascular endothelial cells and red blood cells are found inside the pores, indicating the formation of blood vessels [12].

Figure 10.

H-E stained histological image 4 weeks after rabbit femoral bone defect was reconstructed using a CO₃Ap honeycomb [13]. The (b) is a locally magnified photograph of (a)

Vertical bone augmentation can be performed using the CO₃Ap honeycomb. Figure 11 summarizes the results of vertical bone augmentation on rabbit cranium. New bone formation was confirmed 4 weeks after implantation, even at the top of the CO₃Ap honeycomb (Figure 11(b)). Magnified images of the top (Figure 11(c)) and bottom (Figure 11(d)) parts of the CO₃Ap honeycomb along the pores, cross section of the CO₃Ap honeycomb at mid-height (Figure 11(e)), and new bone formation at the pore surface of the CO₃Ap honeycomb (Figure 11(f)) are shown. The same histological results were obtained for vertical bone augmentation as for femoral bone defect augmentation. Osteoclasts, osteoblasts, and osteocytes were also observed. In addition, numerous vascular endothelial cells and red blood cells were found inside the pores.

Figure 11.

Implantation procedure (a) and H-E stained histological images, (b)-(f), 4 weeks after vertical bone augmentation on rabbit cranium [13]. (c) top of the CO₃Ap honeycomb, (d) bottom parts of the CO₃Ap honeycomb, (e) CO₃Ap honeycomb at mid-height, (f) higher magnification of (e)

A comparison of the cell and tissue responses among the CO₃Ap, HAp, and β-TCP honeycombs has also been reported [35]. Alkaline phosphatase (ALP) activity is an index of osteoblast maturation. ALP activity of MC3T3-E1 cells 7 days following seeding was approximately two-fold higher for the CO₃Ap honeycomb than for β-TCP and HAp honeycombs (Figure 12).

Figure 12.

Effects of honeycomb’s composition on MC3T3-E1 cell differentiation in vitro. ALP activity of cells cultured for 7 days on CO₃Ap, HAp, and β-TCP honeycombs [35]

Figure 13 summarizes μ-CT and H-E stain histological findings 4 weeks after bone defects made at the distal epiphysis of rabbit femurs were reconstructed with CO₃Ap, HAp, and β-TCP honeycombs [35]. In the case of the CO₃Ap honeycomb, new mature bone formed along the walls surrounding the macropores. Osteoblasts were detected along the new bone, osteoclasts were detected on the material surface, and blood vessels were detected in the macropores (Figure 13(c),Figure 13(g)). In the HAp honeycomb, immature bone, but not mature bone, was observed in only a small portion of the macropores at 4 weeks post-operation (Figure 13(e,Figure 13(h)). Osteoblasts, osteoclasts, and blood vessels were not observed in any macropores in the HAp honeycomb, whereas these cells and tissues were present in every macropore examined in the CO₃Ap honeycomb. Even at 12 weeks after surgery, almost all macropores were occupied by immature bone, and osteoblasts and osteoclasts were not observed in the macropores. In the case of β-TCP honeycomb 4 weeks after surgery, almost all macropores were filled with mesenchyme, and no osteoblasts or osteoclasts were observed in the macropores (Figure 13(f,Figure 13(i)). New mature bone formation was observed within a small portion of macropores. The area of mature bone at 4 and 12 weeks after surgery is summarized in Figure 14 along with the remaining mineral area. The area of mature bone area was significantly larger for CO₃Ap honeycomb compared to that of HAp and β-TCP honeycombs. Remaining materials area was largest for HAp honeycomb followed by CO₃Ap honeycomb and β-TCP honeycomb for both 4 and 12 weeks after surgery.

Figure 13.

μ-CT images, (a–c), and HE-stained histological images, (d-i), of bone defects 4 weeks after reconstruction with CO₃Ap (a, d, g), HAp (b, e, h), and β-TCP (c, f, i) honeycombs. ‘OB,’ ‘OC,’ ‘MB,’ ‘IB,’ ‘MC,’ ‘M,’ ‘BV,’ and ‘AC’ indicate osteoblast, osteoclast, new mature bone, immature bone, mesenchyme, material, blood vessel, and adipose cell, respectively [35]

Figure 14.

Mature bone area and remaining material area of bone defects 4 and 12 weeks after reconstruction with CO₃Ap, HAp, and β-TCP honeycombs. *p < 0.05 [35]

5. CO₃Ap coated titanium plate

Titanium (Ti) is used as an implant device because of its high mechanical strength and ease of osseointegration [36,37]. In particular, surface-roughed Ti implants demonstrate greater osseointegration and a higher implantation success rate. However, osteogenesis of Ti, even with a roughened surface, is poorer than that of osteoconductive materials like apatite. Therefore, initial loosening of Ti implants is a problem for early loading [38, 39, 40].

Ti roughened surface coated with CO₃Ap may be an ideal Ti implant because of its pronounced osteoconductivity and replacement by new bone. CO₃Ap coated Ti (CO₃Ap–Ti) can be fabricated by compositional transformation through a dissolution–precipitation reaction using calcite-coated titanium (CaCO₃–Ti) as a precursor [39–41].

In this process, CaCO₃–Ti is established on the roughened surface Ti by wetting the Ti with an ethanol solution of Ca(NO₃)₂, followed by heating at 550°C in a CO₂ gas atmosphere. This results in the thermal decomposition of Ca(NO₃)₂ and its carbonation. CO₃Ap–Ti is then fabricated through a dissolution–precipitation reaction in an aqueous solution of Na₂HPO₄ using CaCO₃–Ti as a precursor.

The adhesion strengths of CaCO₃–Ti and CO₃Ap–Ti are as high as 56.6 and 76.8 MPa, respectively. Figure 15 summarizes the VG-stained histological results 4 weeks after implantation in the straight defect of the proximal tibia 2 cm away from the knee joint in rabbits. Abundant mineralized bone formed on the surface of CO₃Ap–Ti. In contrast, osteoid, rather than mineralized bone, was primarily present on the surface of the rough–Ti surface. The bone-implant contact percentages in CO₃Ap–Ti to bone (72.9% ± 6.4%) exceeded those of rough–Ti to bone (48.2 ± 7.8%) and calcite–Ti to bone (59.9 ± 15.4%).

Figure 15.

Villanueva-Goldner-stained histological sections: general view, and high-magnification view of the cortical bone of sections in rough-Ti, calcite-Ti, and CO₃Ap-Ti after 4 weeks of healing showing bone-implant contact (BIC) and osteoid (Os). Yellow arrows indicate BIC areas [41]

Figure 16 illustrates the detaching test used to measure the adhesion strength of the implant to the bone and the results. The increased bone-implant contact resulted in markedly greater adhesion strengths of CO₃Ap–Ti to bone (42.5 ± 14.7 N) than that of rough-Ti to bone (8.7 ± 4.3 N) and calcite–Ti to bone (24.0 ± 8.9 N). Most of the fracture occurred between at the bone in the case of CO₃Ap–Ti, whereas most of the fracture occurred at the interface between Ti plate and calcite for calcite–Ti, and interface between Ti plate and bone for rough-Ti. Very high bonding strength with bone and fracture at bone may be caused, at least in part, to the expansion of CO₃Ap in the roughened Ti during the dissolution–precipitation reaction.

Figure 16.

Illustration of the (a) detaching test in bone harvested at 4 weeks after implantation of Ti plates in the tibia defects of rabbits, and (b) adhesion strength of bone-to-implant in rough-Ti, calcite-Ti, and CO₃Ap-Ti (n = 8). *p < 0.05 for comparisons between the indicated groups [41]

The documented excellent tissue response, higher osteoconductivity, and higher adhesion strength guarantee the usefulness of CO₃Ap–Ti for dental and orthopedic implants.

6. Conclusion

Chemically pure (100%) CO₃Ap artificial bone can be fabricated by compositional transformation through a dissolution–precipitation reaction using a precursor. Although the results obtained to date are encouraging, little is known about CO₃Ap artificial bone compared with other artificial bones, such as HAp and β-TCP.

Funding Statement

This work was supported by the Japan Agency for Medical Research and Development [JP21im0502004h, JP20lm0203123h].

Disclosure statement

No potential conflict of interest was reported by the author(s).