Structural Effects of Sulfur-Containing Functional Groups on Apatite Formation on Ca2+-Modified Copolymers in a Simulated Body Environment.

Chemical modification with specific functional groups has been the conventional method to develop bone-bonding bioactive organic-inorganic hybrids. These materials are attractive as bone substitutes because they are flexible and have a Young's modulus similar to natural bone. Immobilization of sulfonic acid groups (-SO3H) onto the polymer chain is expected to produce such hybrids because these groups induce apatite formation in a simulated body fluid (SBF) and enhance the activity of osteoblast-like cells. Sulfinic acid groups (-SO2H), which are derivatives of -SO3H, can also induce apatite nucleation. However, the structural effects of such sulfur-containing functional groups on apatite formation have not been elucidated. In the present study, apatite formation on Ca2+-modified copolymers containing -SO2H or -SO3H was investigated in a simulated body environment. The copolymer containing Ca2+ and -SO3H promoted Ca2+ release into the SBF and formed apatite faster (1 day) than the copolymer containing Ca2+ and -SO2H (14 days). In contrast, when they were not modified with Ca2+, the copolymer containing only -SO2H deposited the apatite faster (7 days) than that containing only -SO3H (>7 days) in the solution with Ca2+ concentration 1.5 times that of SBF. The former adsorbed larger amounts of Ca2+ than the latter. The measured stability constant of the complex indicated that the interaction of -SO2-···Ca2+ was more stable than that of -SO3-···Ca2+. It was found that both the release and adsorption of Ca2+ governed by the stability played an important role in induction of the apatite formation and that the apatite-forming ability of sulfur-containing functional groups drastically changed by the coexistence of Ca2+.

Introduction

Modifying with specific functional groups on the surface of implant materials to repair bone defects provides the implants with bone-bonding ability (i.e., bioactivity). Bone bonding is required for the longevity of the implant. Hench first showed that Bioglass bonds to the living bone without a foreign body reaction.[1] Subsequently, the glass ceramic A-W[2] and sintered hydroxyapatite[3,4] were developed as bioactive ceramics. When glass-based bioactive materials are implanted, the formation of low crystalline apatite induced by the surface Si–OH group and the release of Ca2+ is an important factor in their bonding to the living bone.[5]

Research has revealed that functional groups other than Si–OH allow the bonelike apatite formation on the material surface. Ti–OH,[6] Zr–OH,[7] Ta–OH,[8] carboxyl (−COOH),[9] phosphate (−PO4H2),[9] and sulfonate (−SO3H)[10] groups induced heterogeneous nucleation of apatite in a simulated body fluid (SBF). SBF contains inorganic ions similar to those found in human blood plasma but does not contain the proteins.[11,12] Modifications of these functional groups have been used to fabricate bioactive metals,[13,14] organic–inorganic hybrids,[15,16] and polymer materials.[17] In particular, bioactive organic–inorganic hybrids and organic polymers are expected to be good bone substitutes as they can form into a defect shape and have a Young’s modulus similar to the living bone.

To improve the biological compatibility of the hybrids and polymers, the effects of sulfonic acid group modification on apatite formation in SBF and activity of osteoblast-like cells on the surface have been examined. Miyazaki et al. created organic–inorganic hybrids consisting of poly[vinylsulfonic acid-co-2-hydoxyethyl methacrylate (HEMA)] and Ca2+.[18] These hybrids formed apatite within 7 days in SBF and exhibited plastic deformation under mechanical property tests. Chaterji and Gemeinhart demonstrated that the increased sulfate group content in the poly(sulfopropyl acrylate-co-acrylamide) hydrogel improved the adhesion and proliferation of MG-63 osteoblast-like cells as a result of increased serum protein uptake.[19] Modification by plasma treatment also increased the activity of the osteoblasts on the scaffold because it allowed vitronection adsorption.[20] Therefore, incorporating sulfonic acid groups is potentially a way to prepare bioactive flexible materials.

We previously reported the apatite-forming ability of copolymers modified with Ca2+ and the sulfinic acid (−SO2H) group in SBF.[21] The copolymers formed apatite, and its morphology was significantly different by the existence of −SO2H. These results mean that the −SO2H group also induced heterogeneous nucleation of the apatite in SBF. Although both −SO3H and −SO2H consist of the same elements, the number of bonding oxygen atoms in the structure is different. It is assumed that the structural differences affect the stability of the interaction with Ca2+ because the large number of oxygen atoms leads to stable ionized acidic functional groups because of the delocalization of the negative charge.[22] In fact, our previous research revealed that the stability of the interaction between the Ca2+ and −PO3H2 groups significantly changed the apatite-forming behavior of the Ca2+-modified copolymers in SBF, depending on the chemical structure of the phosphate.[23] However, the structural effects of sulfur-containing acidic functional groups on the apatite-forming ability of the copolymers remain unclear.

In this study, we prepared copolymers consisting of Ca2+ and an organic matrix modified with −SO3H or −SO2H groups using sodium 4-vinylbenzenesulfonate (NaVBSO3) or sodium 4-vinylbenzenesulfinate (NaVBSO2) and HEMA. The mineralization behavior on the surface of hybrids was examined to determine the structural effects using analytical and surface chemistry techniques.

Results

Figure shows the Fourier transform infrared (FT-IR) spectra of different copolymer specimens before being soaked in the CaCl2 solution. The HEMA peak in both spectra was observed at 1072 cm–1, which was attributed to stretching vibration of the O–C bond for alcohol.[24] In addition, peaks corresponding to the vibration of CH3 and torsion of the OH groups of HEMA were observed at 1146–1148 cm–1. For both specimens, in-plane deformation vibrations of C–H and stretching vibration of S=O bonds were observed at 1030 and 1008 cm–1,[25] respectively. They were derived from both types of monomer-containing sulfur. VBSO3 displayed a peak corresponding to the antisymmetric stretching vibration of O=S=O at 1122 cm–1,[26] but VBSO2 did not.

Figure 1

FT-IR spectra of the different copolymer specimens before being soaked in the CaCl2 solution (▲: γ(CH3) or τ(OH) of HEMA; △: asymmetric stretching of O=S=O; ◆: ν(O–C) alcohol of HEMA; □: S=O bond of VBSO2; ○: C–H bond of in-plane VBSO3 and VBSO2 deformation; ν is stretching, γ is rocking, and τ is torsion).

Figure shows the Ca and sulfur-containing acidic functional group content in the different specimens. VBSO2 and VBSO3 contained nearly equivalent levels of each group. However, the content of Ca in VBSO3 2.5Ca was slightly lower than in VBSO2 2.5Ca. Figure shows the change in Ca concentration in the tris(hydroxymethyl)aminomethane (Tris)–NaCl buffer after soaking the different specimens treated with the CaCl2 solution as a function of soaking time. The Ca concentration increased up to a soaking time of 1 day, slightly decreased, and then became almost constant for both specimens. The Ca concentration of VBSO3 2.5Ca after soaking was higher than that of VBSO2 2.5Ca regardless of soaking time.

Figure 2

Content of (A) sulfur-containing acidic functional groups and (B) Ca in the specimens (n = 3).

Figure 3

Change in Ca concentration in the Tris–NaCl buffer after soaking the different copolymer specimens for various time intervals.

Figure shows thin-film X-ray diffraction (TF-XRD) patterns of the copolymer specimens after soaking in SBF for various time intervals. The peaks corresponding to apatite (JCPDS #09-0432) were first observed at 2θ = 26° and 32° after 1 and 14 days in VBSO3 2.5Ca and VBSO2 2.5Ca, respectively.

Figure 4

TF-XRD patterns of the different copolymer specimens after soaking in SBF for various time intervals.

Figure A shows scanning electron microscopy (SEM) images of the powdered copolymer specimens without preliminary Ca2+ incorporation after soaking in SBF1.5Ca for various time intervals. The surface of VBSO3 0Ca was still smooth after 7 days, while hemispherical depositions were observed on VBSO2 0Ca. Figure B shows a magnified SEM image of VBSO2 0Ca after 7 days and the energy-dispersive X-ray (EDX) spectrum of the depositions detected by point analysis using SEM–EDX. Hemispherical depositions of primary particles with scalelike shapes were observed on its surface. The P, Ca, and Mg peaks were detected in the EDX spectra. The Ca/P molar ratio of the deposition was 1.60. The morphology and the molar ratio were similar to the apatite formed in a simulated body environment.[10]

Figure 5

(A) SEM images of the powdered copolymer specimens after soaking in SBF1.5Ca for various time intervals and (B) magnified SEM image of the deposition on VBSO2 0Ca after 7 days and EDX spectrum of the spot indicated with an asterisk.

Figure shows transmission electron microscopy (TEM) images of the surface of powdered specimens after being soaked in SBF1.5Ca for various time intervals. Depositions with a diameter around 20 nm were observed on the surface of VBSO2 0Ca after 3 days, and the diameter increased to 50–100 nm after 5 days. Fine particles with a diameter of about 20 nm were observed on the surface of VBSO3 0Ca after 5 days.

Figure 6

Bright-field TEM images of the powdered copolymer specimens after soaking in SBF1.5Ca for various time intervals.

Figure shows the change in Ca and P concentrations in SBF1.5Ca after soaking the powdered copolymer specimens for various time intervals. Decreased concentrations of both were observed after soaking both specimens. The concentration decrease was higher for VBSO2 0Ca than for VBSO3 0Ca.

Figure 7

Change in Ca and P concentration in SBF1.5Ca after soaking the powdered copolymer specimens for various time intervals.

Table shows the zeta potential of the powdered copolymer specimens after soaking in the Tris–NaCl buffer and the stability constant of the Ca2+ complex composed of the ionized p-toluenesulfinic acid or p-toluenesulfonic acid. VBSO2 0Ca and VBSO3 0Ca had nearly equivalent negative potentials at the ion strength of physiological conditions. The stability constant of the p-toluenesulfinate–Ca2+ complex was higher than that of the p-toluenesulfonate–Ca2+ complex.

Table 1

Zeta Potential of the Powdered Polymer Specimens (n = 3) and Stability Constant of the Complexes Consisting of Each Acidic Functional Group and Ca2+ at μ = 0.16

	VBSO2 0Ca, −SO2H	VBSO3 0Ca, −SO3H
zeta potential/mV	–16.8 ± 2.3	–18.5 ± 5.4
stability constant log β	1.61	1.31

Discussion

Copolymers with nearly equivalent amounts of different acidic functional groups were successfully obtained using the same method (Figure A). This was confirmed by FT-IR spectra showing that only VBSO3 displayed the O=S=O peak characteristic to −SO3H (Figure ). The apatite formation in SBF was faster on VBSO3 than on VBSO2 when they were treated with CaCl2 solution (Figure ). This means that −SO3H on the copolymers was more effective for apatite formation than −SO2H with coexistence of Ca2+.

The induction of the apatite formation on a substrate depends on the surface conditions and supersaturation degree with respect to apatite in SBF.[27] The release of Ca2+ from the surface is especially effective for increasing the supersaturation degree.[5] VBSO3 2.5Ca was able to release larger amounts of Ca2+ than VBSO2 2.5Ca (Figure ), meaning that the former provided more favorable conditions for the apatite formation in SBF. However, the Ca content in VBSO3 2.5Ca was lower than that in VBSO2 2.5Ca after soaking in CaCl2 solution (Figure B). The Ca2+ release from the copolymers into the SBF is expected to be governed by the initial Ca content and its binding with surrounding anionic functional groups. It has been reported that Ca2+ release is inhibited by the increased −SO3H content in the copolymers via formation of a complex such as (−SO3–)2···Ca2+.[18]

In this study, the copolymers were composed of functional groups with different structures. The change of Gibbs free energy (ΔG) during the formation of the complex is related to the stability constant of the complex (β) as shown in eq where R is the gas constant and T is the temperature of the solution. On the basis of the measured stability constant of the Ca2+ complex (Table ), the formation of (−SO2–)2···Ca2+ gave a lower negative value of ΔG than (−SO3–)2···Ca2+ at the same temperature. The difference in ΔG value indicates that the interaction of (−SO2–)2···Ca2+ was thermodynamically more stable than (−SO3–)2···Ca2+ under physiological conditions. Thus, it is assumed that the ability of −SO3H to trap Ca2+ in the copolymers is weak, which results in the enhancement of the Ca2+ release. Modification with −SO3H has the advantage of apatite formation on the hybrids.

In contrast, −SO2H in VBSO2 0Ca induced the apatite formation faster than −SO3H in VBSO3 0Ca under the condition in which Ca2+ was not released to keep the supersaturation degree constant with respect to the apatite (Figure A,B). This phenomenon is opposite to what happens to the present copolymers modified with Ca2+. Dey et al. observed the mechanism of apatite formation on the monolayer of arachidic acid in SBF using cryo-TEM.[28] They showed that 50–80 nm diameter spherical nodules of amorphous calcium phosphate were formed by the aggregation of the calcium phosphate complex on the monolayer, resulting in conversion to apatite. Habraken et al. determined that the composition of the amorphous phase is [Ca2(HPO4)3]2–, which grows into Ca-defected octacalcium phosphate as a precursor of the apatite in the solution.[29] In the present study, both VBSO2 0Ca and VBSO3 0Ca also initially formed particles on their surfaces with diameters similar to the above reports (Figure ). Moreover, Ca and P consumption in SBF1.5Ca within 5 days suggested calcium phosphate formation on VBSO2 0Ca and VBSO3 0Ca (Figure ). Thus, it is assumed that −SO2H could induce formation of a larger amount of calcium phosphate as a precursor of the apatite at an earlier stage of soaking than −SO3H.

According to previous reports on the heterogeneous nucleation process of apatite, bioactive materials initially adsorb Ca2+, and then phosphate ions on their surface form calcium phosphate in the SBF.[30,31] Yamashita et al. reported that hydroxyapatite subjected to electric poling had a highly negatively charged surface and promoted apatite formation because the amount of adsorbed Ca2+ increased on its surface.[32] However, ion–ion interactions with Ca2+ were also considered as a factor of the Ca2+ adsorption for apatite formation.[9] Judging from the higher rate of decrease in Ca concentration at the initial stage of soaking (Figure ), it is assumed that VBSO2 0Ca had a more favorable surface for Ca2+ adsorption compared with VBSO3 0Ca. There was no significant difference in the zeta potential on each specimen (Table ). Therefore, the different stability constants of the two complexes should be the main factor affecting the apatite-forming ability, although statistical significance remains to be confirmed in the future.

On the basis of the discussion described above, the detailed mechanism of apatite formation on the copolymer in SBF1.5Ca is expected to progress as follows. (1) Chemical interaction of −SO3–···Ca2+ or −SO2–···Ca2+ was constructed on the surface of each specimen at initial stage of soaking. (2) HPO42– in SBF1.5Ca was adsorbed on the surface because the surface potential increases from negative to positive after the initial adsorption of Ca2+.[30,31] (3) HPO42– reacted with the Ca2+ bonded to the functional groups to form the spherical particles of amorphous calcium phosphate (Figure ). (4) The amorphous particles transformed and grew into apatite (Figure ) because of the consumption of the ions in SBF1.5Ca (Figure ). Regarding the reaction in the process (3), our previous research suggests that formation of Ca2+···HPO42– interaction increases the concentration of the calcium phosphate complexes which aggregate to the [Ca2(HPO4)3]2– cluster on the copolymer.[23] Moreover, Kawai et al. reported that the crystal growth of apatite is governed by the supersaturation degree regardless of the type of anionic functional groups on the surface in a simulated body environment.[27] It means that difference in the chemical structure of functional groups changes the behavior of apatite formation during the process of (1–3) without Ca2+ release from the copolymer. Therefore, it is assumed that −SO2H adsorbed a large amount of Ca2+ and accelerated the Ca2+···HPO42– formation and the subsequent aggregation, resulting in enhancement of the apatite deposition on VBSO2 0Ca.

In conclusion, it was found that the apatite formation induced by various sulfur-containing functional groups was significantly altered by the coexistence of Ca2+. Namely, the low stability of −SO3–···Ca2+ improved the apatite-forming ability of the Ca2+-modified copolymers, although −SO3H by itself had a lower ability of inducing heterogeneous apatite nucleation than −SO2H. This is because Ca2+ released from the −SO3–···Ca2+ decreased the activation energy for heterogeneous nucleation and increased the rate of crystal growth on the copolymers. It has been reported that apatite growth on the polymer substrate is also related to the supersaturation degree in the simulated body environment.[27]

The Ca2+ bound to the material surface can adsorb the proteins, leading to the osteoblast adhesion and osteoconductivity of the titanium.[33,34] In addition, it has been reported that suitable Ca2+ concentration in the medium impacts osteoblast differentiation.[35] Consequently, it is expected that the structural differences of these functional groups affect the osteoblast activities and the bone-bonding ability through the state of binding with Ca2+. To confirm this, the behavior of protein adsorption and the activities of the cells on the different hybrids should be investigated.

Conclusions

The apatite-forming behavior of copolymers modified with −SO2H or −SO3H was examined in a simulated body environment. When they were modified with Ca2+, −SO3H formed apatite on the surface faster than −SO2H. It was assumed that the low stability of −SO3–···Ca2+ had the advantage of Ca2+ release, which promoted both heterogeneous nucleation and crystal growth of the apatite. In contrast, the opposite phenomenon was observed when they were not modified with Ca2+. High stability of the −SO2–···Ca2+ interaction enhanced the heterogeneous nucleation of the apatite on the copolymers. It was found that the stability of the interaction with Ca2+ depended on the structure of these functional groups and was an important factor governing the apatite formation.

Materials and Methods

Specimen Preparation

NaVBSO2 (90%, Tokyo Chemical Industry Co. Ltd., Tokyo, Japan) and NaVBSO3 (90%, Sigma-Aldrich Co., Saint Louis, USA) were used to incorporate the −SO2H and −SO3H groups, respectively, into the copolymer. A total of 1.0 mmol of NaVBSO2 or NaVBSO3 and 9.0 mmol of HEMA (95%, Wako Pure Chemical Industries Ltd., Osaka, Japan) were dissolved in ultrapure water. Furthermore, 5.0 × 10–4 mol of N,N′-methylenebisacrylamide (99%, Wako Pure Chemical Industries Ltd.) and 1.5 × 10–5 mol of N,N,N′,N′-tetramethylethylenediamine (98%, Wako Pure Chemical Industries Ltd.) and 2,2′-azobis(2-methylpropionamidine)dihydrochloride (95%, Wako Pure Chemical Industries Ltd.) were added. The mixture of these regents was adjusted to a total volume of 10 cm3 in a volumetric flask. To copolymerize these monomers, 1 cm3 of the resulting solution was put in a polyethylene cup at 60 °C for 1 day.

The obtained copolymers were cut into ϕ 5 × 1 mm cylinders. They were soaked in 1 kmol·m–3 HCl solution for 1 day to remove the unreacted reagents and sodium ions, soaked in ultrapure water for 1 day, and then dried in vacuum. The obtained copolymers were then soaked in 30 cm3 of 0 or 2.5 mol·m–3 CaCl2 solution at 36.5 °C for 1 day. The copolymers prepared from NaVBSO2 and NaVBSO3 were abbreviated as VBSO2XCa and VBSO3XCa, respectively, where X indicates the concentration of the CaCl2 solution used. VBSO2 and VBSO3 indicate the specimens without CaCl2 treatment.

Soaking Specimens in SBF and SBF1.5Ca To Evaluate Apatite-Forming Ability

The cylindrical VBSO2 2.5Ca and VBSO3 2.5Ca specimens were soaked in 30 cm3 of SBF (Na+ 142.0, K+ 5.0, Mg2+ 1.5, Ca2+ 2.5, Cl– 147.8, HCO3– 4.2, HPO42– 1.0, SO42– 0.5 mol·m–3, pH 7.40) for various time intervals. In addition, 9 mg of powdered VBSO2 0Ca and VBSO3 0Ca specimens was soaked in SBF1.5Ca, which has Ca2+ concentration 1.5 times that of the normal SBF[36] with a pH 7.25 at 36.5 °C for various time intervals. The powder was prepared by cooling the bulk specimen using liquid N2 and pulverizing it with a porcelain mill and agate balls.

SBF and SBF1.5Ca were prepared by dissolution of NaCl, NaHCO3, KCl, K2HPO4·3H2O, MgCl2·6H2O, CaCl2, and Na2SO4 (Nacalai Tesque, Inc., Kyoto, Japan) in ultrapure water in that order. The pH for each resulting solution was adjusted by the addition of 1 kmol·m–3 HCl solution and Tris (Nacalai Tesque, Inc.).

Soaking Specimens in Tris–NaCl Buffer To Measure Ca2+ Release

The hybrid specimens were soaked in 30 cm3 of Tris–NaCl buffer (NaCl 142, Tris 50 mol·m–3, pH 7.40) at 36.5 °C for various time intervals to measure the amount of released Ca2+. The Tris–NaCl buffer was prepared by the addition of NaCl and Tris in ultrapure water, and the pH was adjusted with the addition of 1 mol·m–3 HCl.

Specimen Characterization

The chemical structure of the different copolymers was analyzed by FT-IR spectroscopy (FT/IR-6100, JASCO Co., Tokyo, Japan) using an attenuated total reflectance method. The content of acidic functional groups in the specimens was measured by neutralizing titration. A 30 cm3 of 0.16 kmol·m–3 NaCl solution with specimens soaked without Ca2+ was titrated with 10.0 mol·m–3 of the NaOH solution. The point of neutralization was determined by the change in pH of the solution measured by a pH meter (F-23IIC, Horiba Ltd., Kyoto, Japan). To determine the Ca content in the specimens, the decrease in Ca concentration in the CaCl2 solution after soaking of specimens was measured using inductively coupled plasma atomic emission spectrometry (ICP–AES; ICPE-9820, SHIMADZU Co., Kyoto, Japan).

After the specimens were soaked in SBF or SBF1.5Ca, they were analyzed by TF-XRD (MXP3V, Mac Science Co., Yokohama, Japan) and EDX (EMAX Energy, Horiba Ltd) equipped with a scanning electron microscope (S-3500N, Hitachi Co., Tokyo, Japan). In the TF-XRD analysis, the incident X-ray (monochromatic Cu Kα radiation) was fixed at 1° relative to the specimen surface with a scan rate of 0.02°·s–1. After the immersion in SBF1.5Ca, the powder specimens were deposited onto a Cu grid with 250 mesh (ELS-C10, Okenshoji Co., Ltd., Tokyo, Japan). Ultrapure water was then added several times to remove the SBF1.5Ca from the grid. The surface was then observed with a transmission electron microscope (JEM-3010, JEOL Ltd., Tokyo, Japan).

The surface zeta potential of the powdered copolymer specimens was measured by the zeta potential analyzer (ELS-Z, Otsuka Electronic Co., Osaka, Japan). The Tris–NaCl buffer with soaked specimens was injected into the quartz cell, and the electrophilic mobility was measured by the laser Doppler method to calculate the zeta potential. The Ca concentration in the Tris–NaCl buffer and Ca and P concentrations in the specimens soaked in SBF1.5Ca for various time intervals were measured using ICP–AES.

Determination of the Stability Constant of the Complexes

The stability constant of the complexes, which consisted of Ca2+ and the ligand of ionized p-toluenesulfinic acid or p-toluenesulfonic acid, was determined to compare the thermodynamic stability of the interaction between Ca2+ and each acid group. A 3.0–4.2 mmol of sodium p-toluenesulfinate (98%, Tokyo Chemical Industry Co., Ltd.) or p-toluenesulfonic acid monohydrate (99%, Wako Pure Chemical Industries Ltd.) was dissolved in 30 cm3 of mol·m–3 CaCl2 solution. The ion strength and pH were adjusted to 0.16 and 6.4–6.7, respectively, by the addition of NaCl and Tris. The free Ca2+ concentration in the solution was measured using a Ca2+-ion-selective electrode (6583-10C, Horiba Ltd.). The stability constant, β, of the Ca2+ complex was calculated with eq where [L2···Ca2+], [L], and [Ca2+] are molar concentrations of the complex, free ligand, and free Ca2+, respectively. [L]total and [Ca2+]total are the total concentration of ligand and Ca2+ in the solution, respectively.