Evaluation of Apatite-Forming Ability and Antibacterial Activity of Raw Silk Fabrics Doped with Metal Ions

Abstract

Raw silk can be doped with metal elements such as calcium and zinc due to the high affinity of sericin, which forms its outer layer. Raw silk doped in this manner is expected to possess various favourable properties as biomaterials. In this study, we investigated metal-doped raw silk fabric’s apatite-forming ability in simulated body fluid (SBF), as well as its antibacterial activity against Escherichia coli. The samples were prepared by soaking the fabric in aqueous solutions containing calcium, copper, or zinc ions. Both Cu-doped and Zn-doped raw silk fabric showed antibacterial activity, suggesting that antibacterial agents released from the samples killed the bacteria. Additionally, Ca-doped raw silk fabric showed both apatite-forming ability and antibacterial activity. The apatite formation on fabric might be because calcium ions released from the sample increased the degree of supersaturation of SBF with respect to apatite, and accelerated apatite formation. Additionally, the release of calcium ions caused local pH increases, resulting in bacterial hardly survival at the sample surface. Therefore, Ca-doped, Cu-doped, and Zn-doped raw silk fabrics may have applications as antibacterial biomaterials. Furthermore, Ca-doped raw silk fabric has the potential to bind to living bone.

 

This Paper was Originally Published in Japanese in J. Jpn. Soc. Powder Powder Metallurgy 65 (2018) 495–501.

Fig. 9 Results of quantitative antibacterial tests for R-Ca, R-Cu, and R-Zn. (**P < 0.01)

1. Introduction

Raw silk is a natural fiber constituting the cocoon shell of silkworms and takes a two-layer structure of protein with fibroin covered by sericin.¹⁾ In many cases, sericin is removed from the surface of raw silk by alkaline solution or soap water and is discarded. However, recent studies have reported that sericin possesses beneficial properties such as wound healing effects,²⁾ antioxidant action,³⁾ and promotion of human skin fibroblast adhesion,⁴⁾ and hence is expected to be useful as a biomaterial. In addition, since serine and aspartic acid in sericin contain hydroxy (OH) or carboxyl (COOH) groups in their side chain, they have high affinity for metallic elements such as calcium (Ca) and zinc (Zn).⁵⁾ Therefore, if a metal element is introduced on the raw silk’s surface, various useful functions as a biomaterial can be achieved.

Osteoconductive material forms bone-like apatite on its surface, which allows it to bind to living bone.^6,7) As a method for evaluating the osteoconductivity of artificial materials in vitro, they can be immersed in simulated body fluid (SBF), which has inorganic ion concentrations almost equal to those of human blood plasma.⁸⁾ Takeuchi et al. found that silk fabric formed apatite on its surface in the solution, which has 1.5 times the ion concentration of a normal SBF (1.5SBF).⁹⁾ However, since the Ca/P ratio and lattice constant of apatite formed on the surface of artificial materials in 1.5SBF differ greatly from those of human apatite,¹⁰⁾ silk fabric might not show excellent osteoconductivity in human body. On the other hand, some artificial materials such as CaO–SiO₂–P₂O₅ type glass and ethylene-vinyl alcohol polymer treated with calcium silicate form apatite even in SBF through the release of Ca ions.^11,12) Hence, if Ca ions can be introduced into raw silk, apatite formation on the surface can be expected, even in SBF. Furthermore, as a woven fabric, raw silk can deform flexibly, and possess excellent mechanical strength due to the presence of fibroin in its inner layer.¹⁾ Consequently, raw silk fabric can be expected to be used as a substitute material in, for example, tendons and ligaments, which are subject to high load application.

Furthermore, the surgical site infection (SSI) at implantation of artificial material is clinically problematic. In the field of orthopedic surgery, SSI onset is reported in the range of 0.6 to 11.9% in spinal surgery, 0.2 to 2.9% in initial artificial joint replacement surgery, and 0.5 to 17.3% in artificial joint replacement surgery.¹³⁾ Once SSI occurs, because the implanted artificial material should be replaced, placing a heavy burden on the patient. Therefore, it is desirable that artificial materials themselves exert antibacterial effects. In recent years, many methods have been attempted in order to induce antibacterial activity on material surfaces, for instance, by supporting silver, copper, or zinc as an antibacterial element,^14–16) or by using the oxidizing action of titanium oxide as photocatalyst.¹⁷⁾ Especially, antibacterial-element-supported materials are expected to exhibit continuous antibacterial effects in vivo.

In this study, basic conditions for a raw silk fabric to exhibit osteoconductivity and antibacterial properties were investigated by evaluating the apatite-forming ability in SBF and antibacterial activity against Escherichia coli (E. coli) of Ca-, Cu-, and Zn-doped raw silk fabrics.

2. Experimental Procedure

2.1 Sample preparation

Special grade reagent of calcium chloride (CaCl₂, Wako Pure Chemical Industries, Ltd., Osaka, Japan), copper (II) chloride dihydrate (CuCl₂·2H₂O, Wako Pure Chemical Industries, Ltd., Osaka, Japan), or zinc chloride (ZnCl₂, Kanto Chemical Co., Tokyo, Japan) were dissolved in distilled water to prepare 1.0 M CaCl₂, CuCl₂, and ZnCl₂ aqueous solutions, respectively. Of each prepared aqueous solution, 30 mL was added to a centrifuge tube, and a commercially available raw silk fabric (R-Silk, Oda Weave Co., Ltd., Kyoto, Japan) cut into 1.0 cm square was soaked therein. The centrifuge tube was placed in a thermostatic chamber (IC402, Yamato Scientific Co., Tokyo, Japan) and kept at 36.5°C for 24 hours. Then, the sample was removed from the centrifuge tube, washed three times with distilled water, and dried at room temperature. Samples soaked in 1.0 M CaCl₂, CuCl₂, or ZnCl₂ aqueous solution were named R-Ca, R-Cu, or R-Zn, respectively. For the antibacterial test described later (Sections 2.5 and 2.6), we used a raw silk sample of 1.5 cm square subjected to the above treatment, and soaked in 67.5 mL of CaCl₂, CuCl₂, or ZnCl₂ aqueous solution to ensure that the volume of aqueous solution per unit area of sample was the same as that for a sample of 1.0 cm square soaked in the aqueous solution.

2.2 Morphology and elemental analysis of sample surface

The surfaces of R-Silk and of the sample obtained in section 2.1 were observed with a scanning electron microscope (SEM, VE-8800, Keyence Corporation, Osaka, Japan). Further, the abundance of elements on the sample surfaces were investigated using X-ray photoelectron spectroscopy (XPS, AXIS-ULTRA, Shimadzu Corporation, Kyoto, Japan). The samples were lyophilized under vacuum for 5 days using a freeze dryer (FD-1000, Tokyo Science Instrument Co., Ltd., Tokyo, Japan) prior to XPS measurement. The conditions of the measurement were AlKα X-ray source, 15 kV of tube voltage, 5 mA of tube current, and 700 × 300 µm of analysis area.

2.3 Dissolution test

Tris-HCl buffer solution (TBS) was prepared as follows: 6.118 g of tris (hydroxymethyl) aminomethane (Nacalai Tesque, Kyoto, Japan) was dissolved in 1,000 mL of distilled water, and the pH adjusted to 7.4 using 1.0 M hydrochloric acid (Nacalai Tesque, Kyoto, Japan). After preparation of TBS, R-Ca, R-Cu, or R-Zn were soaked in 30 mL of TBS in a centrifuge tube and kept in a thermostatic chamber at 36.5°C. The samples were removed from TBS after 1, 3, or 7 days of soaking and the concentration of Ca, Cu, or Zn ions were measured using inductively coupled plasma atomic emission spectroscopy (ICP-AES, iCAP6500, Thermo Fisher Scientific Co., Ltd., Kanagawa, Japan). In addition, the presence of elements on the surface of the sample was examined using XPS and the amount of Ca, Cu, and Zn on the sample after 0 and 7 days soaking in TBS were compared. The release ratio of Ca, Cu, or Zn from the sample was calculated using eq. (1).   

\begin{equation} \text{Release ratio (%)}=(C_{0}-C_{7})/C_{0}\times 100 \end{equation} (1)

• C₀: The value obtained by normalizing the abundance ratio of Ca, Cu, or Zn with that of carbon (C) on the sample surface after 0 days soaking in TBS.
• C₇: The value obtained by normalizing the abundance ratio of Ca, Cu or Zn with that of C on the sample surface after 7 days soaking in TBS.

2.4 SBF soaking test

SBF with pH 7.40 was prepared according to ISO 23317:2014. R-Ca, R-Cu, or R-Zn were soaked in 30 mL of SBF in a centrifuge tube and kept in a thermostatic chamber at 36.5°C. After 3, 5, or 7 days from the start of soaking, the sample was removed, washed with distilled water, and dried at room temperature.

The surface structure of the sample was investigated using SEM and X-ray diffractometer (XRD, MiniFlex600, Rigaku, Tokyo, Japan). In the XRD measurement, CuKα was used as a radiation source, using a scanning speed of 4°/min, a step size was 0.02°, tube voltage of 40 kV, and tube current of 15 mA. Elemental analysis of the surface of treated samples was carried out using an energy dispersive X-ray spectrometer (EDX, Japan Electronic Co., Tokyo, Japan). Changes in the concentrations of Ca and phosphorus (P) ions in SBF while soaking samples were measured using ICP-AES.

2.5 Antibacterial test by halo method

Based on JIS L 1902:2015, antibacterial properties of samples were qualitatively evaluated by the halo method. To prepare a broth medium, 5.0 g of meat extract (Becton Dickinson and Company, Franklin Lakes, New Jersey, USA), 10.0 g of peptone (Becton Dickinson and Company, Franklin Lakes, NJ, USA), and sodium chloride (NaCl, Nacalai Tesque Inc., Kyoto, Japan) were mixed in distilled water 1000 mL. One colony of E. coli (Escherichia coli [Migula 1895] Castellani and Chalmers 1919, JCM No. 5491, National Institute for Science and Technology, Science Institute Bioresource Center, Ibaraki, Japan) was transferred to 20 mL of broth medium. Subsequently, the bacterial solution was incubated in a thermostatic chamber for 18 hours as a primary culture. Then, 0.4 mL of the solution was transferred to 20 mL of fresh broth medium and incubated for 4 hours as a secondary culture. The solution was then diluted 100-fold with broth medium which was diluted 20-fold with distilled water and used for the antibacterial test.

The prepared bacterial solution (1 mL) and 15 mL of nutrient agar medium were added to a Petri dish (Corning Inc., Corning, New York, USA) and mixed well. The sample was placed on the surface of the nutrient agar medium, following which the petri dish was cultured at 36.5°C for 24 hours. After incubation, the presence or absence of a halo (a zone of in which bacterial growth has been inhibited) around the sample was confirmed and the antibacterial effect of the sample against E. coli was evaluated.

2.6 Antibacterial test by bacterial liquid absorption method

Based on JIS L 1902:2015, the antibacterial properties of samples were quantitatively evaluated using the bacterial liquid absorption method. Nutrient broth (NB) was prepared by adding 3 g of meat extract, and 5 g of peptone to 1,000 mL of distilled water. Primary and secondary culture for E. coli were carried out with NB as described in section 2.5. The bacterial solution after the secondary culture was diluted 1000-fold with NB which was diluted 20-fold with distilled water and used for the antibacterial test.

Six raw silk specimens placed in 5 mL vial bottles were inoculated with 50 µL of the prepared bacterial solution in each of several places. Three samples were cultured at 36.5°C for 18 hours. The remaining three samples were added with 5 mL of saline supplemented with 2.0 w/v% polysorbate 80 (Wako Pure Chemical Industries, Ltd., Osaka, Japan) and shaken 30 times to extract bacteria from the samples. Subsequently, the dilution plate method was used to count a number of viable bacteria in the extracted bacterial solution. The bacterial solution at each magnification (1 mL) was mixed with 15 mL of agar medium (EA) and added to a Petri dish. EA was prepared by adding 2.5 g of dehydrated yeast extract (Becton Dickinson and Company, Franklin Lakes, NJ, USA), 5.0 g of casein Tryptone (Becton Dickinson and Company, Franklin Lakes, NJ, USA), 1.0 g of D (+)-glucose (Wako Pure Chemical Industries, Ltd., Osaka, Japan), and 15 g of agar to distilled water. After culture at 36.5°C for 24 hours, the number of colonies was counted.

For three specimens cultured with bacterial solution for 18 hours after inoculation, the number of E. coli colonies was counted in the same manner as above. Thereafter, the bacterial growth value (F) and the antibacterial activity value (A) were calculated according to JIS L1902:2015.

For the control, 1.5 cm square of R-Silk was used. For the evaluation of significant differences between samples, the method of Tukey was used, and the significance level was set to 5%.

3. Results and Discussion

3.1 Structure and elemental analysis of samples

SEM images of each sample are shown in Fig. 1. There was no difference between the surface morphologies of R-Ca, R-Cu, and that of R-Silk, but precipitates were observed on the surface of R-Zn. The precipitates were described in the same section later. Table 1 shows the results of elemental analysis of sample surfaces by XPS. Slight amounts of Ca, Cu, or Zn were detected for R-Ca, R-Cu, or R-Zn, respectively, indicating that these elements were successfully doped into R-Silk. In all samples, silicon (Si) was detected. This might be attributed to minerals contained in raw silk fabrics.¹⁸⁾

Fig. 1

Scanning electron microscope (SEM) images of R-Silk (a), R-Ca (b), R-Cu (c), and R-Zn (d).

Table 1 Surface abundance of elements in samples as determined by X-ray photoelectron spectroscopy (XPS).

The Ca_2p, Cu_2p, and Zn_2p XPS spectrum of R-Ca, R-Cu, and R-Zn, respectively, are shown in Fig. 2. As R-Ca gave Ca_2p peaks around 347.6 and 351.1 eV, it is likely that Ca is present in R-Ca as Ca²⁺.^19,20) It is assumed that the Ca²⁺ associated with the COOH or OH groups of sericin by chelate bonding.⁵⁾ As the isoelectric point of sericin is approximately 4.3,²¹⁾ COOH is ionized to COO⁻ in the CaCl₂ aqueous solution, which is highly likely to attract Ca²⁺. Therefore, there is a high possibility that Ca in R-Ca is mainly localized to the side chain of aspartic acid. Additionally, as Cu_2p peaks of approximately 932.9, 934.8 and 944.6 eV were detected in R-Cu, it is probable Cu was present as Cu⁺ and Cu²⁺.²⁰⁾ Why some of the Cu²⁺ was reduced to Cu⁺ is unknown, but we speculate that these ions are supported mainly on the side chain of aspartic acid, similarly to Ca in R-Ca. Furthermore, as R-Zn gave a Zn_2p peak around 1022.0 eV, the precipitate on the surface of R-Zn (Fig. 1) is composed of zinc oxide (ZnO) and/or zinc hydroxide (Zn(OH)₂).^20,22)

Fig. 2

Ca_2p X-ray photoelectron spectroscopy (XPS) spectrum of R-Ca (a), Cu_2p XPS spectrum of R-Cu (b), and Zn_2p XPS spectrum of R-Zn (c).

3.2 Metal ion dissolution characteristics of samples

The concentrations of Ca, Cu, or Zn ions released from samples soaked in TBS for various periods are shown in Fig. 3. Ion concentration increased in each solution in which R-Ca, R-Cu, or R-Zn were soaked, indicating that the metal ions doped into each sample were released into TBS. Furthermore, the concentrations of released metal ions were almost the same between 1 day and 7 days soaking for all samples, indicating that their release is rapid. Table 2 shows the results of elemental analysis by XPS on the surface of samples soaked in TBS for 7 days. According to Tables 1 and 2, and eq. (1), approximately 73.3% of Ca in R-Ca, 33.2% of Cu in R-Cu, and 35.7% of Zn in R-Zn was released into TBS.

Fig. 3

Concentrations of Ca, Cu, and Zn released from R-Ca, R-Cu, and R-Zn into Tris-HCl buffer solution (TBS), respectively.

Table 2 Surface abundance of elements in samples after soaking in Tris-HCl buffer solution (TBS) for 7 days, as measured by X-ray photoelectron spectroscopy (XPS).

3.3 Apatite-forming ability of samples

Figure 4 shows SEM images of R-Ca, R-Cu and R-Zn after SBF soaking. In the case of R-Ca, limited precipitation (arrows) occurred on the surface after 3 days of soaking in SBF, with the entire surface being covered with precipitates after 5 days to 7 days. For R-Cu, no differences in surface structure were observed before and after soaking in SBF. With R-Zn, the precipitates (Fig. 1) observed on the surface from before soaking in SBF remained even 7 days after soaking. The XRD patterns of R-Ca, R-Cu, and R-Zn following soaking in SBF are presented in Fig. 5. R-Ca soaked for 5 or 7 days gave diffraction peaks attributed to hydroxyapatite (HAp, Ca₁₀(PO₄)₆(OH)₂) at 2θ = 26, 32, 40, and 47°.^9,23) On the other hand, no change in XRD pattern was observed in either R-Cu or R-Zn up to 7 days after soaking. These results demonstrate that R-Ca shows apatite-forming ability in SBF, but that R-Cu and R-Zn do not.

Fig. 4

Scanning electron microscope (SEM) images of R-Ca, R-Cu, and R-Zn after soaking in SBF for various periods. Arrows indicate precipitates.

Fig. 5

X-ray diffraction (XRD) patterns of R-Ca (a), R-Cu (b), and R-Zn (c) after soaking in simulated body fluid (SBF) for various periods.

Changes in the concentrations of Ca and P in SBF following sample soaking are presented in Fig. 6. The concentration of Ca and P ions decreased after soaking of R-Ca, indicating that Ca and P ions were incorporated, and that HAp was formed on its surface. Since the concentration of P ions associated with R-Ca had already decreased, the precipitate observed in its surface (Fig. 4), which formed after 3 days soaking in SBF, is considered to be low-crystalline calcium phosphate. On the other hand, the Ca and P ion concentrations did not decrease after soaking of R-Cu, indicating that Ca and P ions were not incorporated into the surface in this case, possibly accounting for the lack of low-crystalline calcium phosphate or HAp formation on the surface of R-Cu. As in the case of R-Cu, no significant decreases in the concentration of Ca and P ions were observed after R-Zn soaking. R-Zn therefore also did not attract Ca and P ions, and thus did not form low-crystalline calcium phosphate or HAp. Therefore, the precipitates observed on the surface of R-Zn after SBF soaking (Fig. 4) are considered to be residues of ZnO and/or Zn(OH)₂ detected by XPS on the surface of R-Zn before SBF soaking (Fig. 2(c)). Figure 7 shows the EDX spectrum of each sample after 7 days soaking in SBF. For R-Ca, strong Ca and P peaks, possibly due to the formation of HAp, were detected. However, R-Zn did not give these peaks, but rather a small Zn peak. These results also support the conclusion that the precipitates on the R-Zn surface were a zinc compound. On the other hand, as a Cu peak was hardly observed for R-Cu, Cu doped into R-Cu might be almost released by soaking in SBF for 7 days.

Fig. 6

Change in calcium (a) and phosphorus (b) concentration of simulated body fluid (SBF) after soaking R-Ca, R-Cu, and R-Zn for various periods.

Fig. 7

Energy dispersive X-ray spectrometer (EDX) spectrum of R-Ca (a), R-Cu (b), and R-Zn (c) after soaking in simulated body fluid (SBF) for 7 days.

Based on these results, we concluded that Ca ions released from R-Ca into SBF, as in the case of TBS (Fig. 3), increase the degree of supersaturation of SBF with respect to apatite, thereby causing Ca ions in SBF to be incorporated into the sample surface, and HAp to be formed.^11,12) On the other hand, although it is presumed that R-Cu and R-Zn released Cu or Zn ions into SBF, as in the case of TBS (Fig. 3), but incorporation of Ca ions into the sample surface was very limited, and HAp consequently did not form. This might be because, even if ion exchanges between released Cu or Zn ions and Ca ions in SBF occur, other positive ions in SBF from elements such as hydrogen (H), sodium (Na), potassium (K), and magnesium (Mg)⁸⁾ inhibited the selective introduction of Ca ions to the sample surface.

3.4 Antibacterial properties

The results of the antibacterial test using the halo method and bacterial liquid adsorption method are presented in Fig. 8 and Fig. 9, respectively. Halos (arrows) were observed for R-Cu and R-Zn after 24 hours culture. These samples therefore exhibited antibacterial activity against E. coli. With the liquid test, R-Silk used as a control, and did not inhibit the growth of E. coli. However, for R-Ca, R-Cu, and R-Zn, the viable bacteria count decreased. Furthermore, comparison of viable bacteria counts among samples after culturing, those of R-Ca, R-Cu, and R-Zn were significantly lower than those of R-Silk. The bacterial growth value (F) and the antibacterial activity value (A) from the results shown in Fig. 9 were presented in Table 3. According to JIS 1902:2015, an antibacterial effect is recognized in a sample which satisfies 2.0 ≤ A < 3.0, and a strong antibacterial effect is recognized in a sample which satisfies A ≥ 3.0. Therefore, R-Ca, R-Cu and R-Zn exhibit strong antibacterial activity against E. coli.

Fig. 8

Results of qualitative antibacterial tests for R-Ca, R-Cu, and R-Zn before and after culture.

Fig. 9

Results of quantitative antibacterial tests for R-Ca, R-Cu, and R-Zn. (**P < 0.01)

Table 3 Bacterial growth value (F) and antibacterial activity value (A) for R-Ca, R-Cu, and R-Zn. R-Silk was used as control.

In the case of R-Cu and R-Zn, in which the formation of halos was observed, we considered that the observed antibacterial effects were due to the release of Cu or Zn ions, which are antibacterial elements, from the respective samples.¹⁴⁾ For R-Cu, Cu ions cause protein denaturation and inhibition of enzymatic function, inhibiting bacterial growth.^24,25) For R-Zn, released Zn ions might inhibit active transport, amino acid metabolism, and enzymatic function.²⁶⁾ Furthermore, in the case of R-Zn, not only the released Zn ions but also zinc compounds formed on the surface contribute to its antibacterial properties. Hirota et al. reported that ZnO produces reactive oxygen species (ROS) even in the dark²⁷⁾ and destroys the bacterial cell wall. Therefore, if R-Zn forms ZnO on its surface, it can contribute to antibacterial properties. Generally, ROS are unstable and highly reactive, which limits their diffusion range. It is assumed that precipitates containing ZnO diffuse into the agar medium, thereby allowing for ROS antibacterial effects to be exhibited against bacteria separated from the sample by some distance. This is also suggested from the fact that ZnO is insoluble in water and the halo observed around R-Zn is white and muddy (Fig. 8). Regarding R-Ca, antibacterial activity was not detected using the halo method, but strong activity was found when using the bacterial liquid absorption method. This might be because H⁺ present near the surface of R-Ca was introduced into the sample by ion exchange with Ca ion released into the medium and the pH near the sample surface increased. According to previous studies, the optimum pH for the growth of E. coli is between 6 to 8,²⁸⁾ therefore R-Ca can inhibit bacterial growth by alkalizing the medium near the sample surface.²⁹⁾

4. Conclusion

In this study, we investigated the antibacterial activity of raw silk fabric doped with Ca, Cu, or Zn ion against E. coli, as well as doped raw silk’s apatite-forming ability in SBF. Ca-doped raw silk fabric formed apatite on its surface in SBF. This was probably because the release of Ca ions from the sample increased the degree of SBF supersaturation with respect to apatite. In addition, Ca-, Cu-, and Zn-doped raw silk fabric showed strong antibacterial activity against E. coli. This antibacterial activity was likely due to (i) increased pH near the sample surface resulting from the release of Ca ions in the Ca-doped sample, (ii) release of antibacterial Cu ions from the Cu-doped sample, and (iii) release of antibacterial Zn ions and generation of reactive oxygen species in the Zn-doped sample.

Acknowledgments

This work was partially supported by a research grant from the Cosmetology Research Promotion Foundation, and Division for Interdisciplinary Advanced Research and Education, Tohoku University. We would like thank Editage (www.editage.jp) for English language editing.

REFERENCES

• 1)   G.H.  Altman,  F.  Diaz,  C.  Jakuba,  T.  Calabro,  R.L.  Horan,  J.  Chen,  H.  Lu,  J.  Richmond and  D.L.  Kaplan: Biomaterials 24 (2003) 401–416.
• 2)   P.  Aramwit and  A.  Sangcakul: Biosci. Biotechnol. Biochem. 71 (2007) 2473–2477.
• 3)   N.  Kato,  S.  Sato,  A.  Yamanaka,  H.  Yamada,  N.  Fuwa and  M.  Nomura: Biosci. Biotechnol. Biochem. 62 (1998) 145–147.
• 4)   K.  Tsubouchi,  Y.  Igarashi,  Y.  Takasu and  H.  Yamada: Biosci. Biotechnol. Biochem. 69 (2005) 403–405.
• 5)   M.  Sasaki,  H.  Yamada and  N.  Kato: Nutr. Res. 20 (2000) 1505–1511.
• 6)   M.  Neo,  S.  Kotani,  T.  Nakamura,  T.  Yamamuro,  C.  Ohtsuki,  T.  Kokubo and  Y.  Bando: J. Biomed. Mater. Res. 26 (1992) 1419–1432.
• 7)   L.L.  Hench: J. Am. Ceram. Soc. 81 (1998) 1705–1728.
• 8)   T.  Kokubo and  H.  Takadama: Biomaterials 27 (2006) 2907–2915.
• 9)   A.  Takeuchi,  C.  Ohtsuki,  T.  Miyazaki,  H.  Tanaka,  M.  Yamazaki and  M.  Tanihara: J. Biomed. Mater. Res. A 65 (2003) 283–289.
• 10)   H.M.  Kim,  K.  Kishimoto,  F.  Miyaji,  T.  Kokubo,  T.  Yao,  Y.  Suetsugu,  J.  Tanaka and  T.  Nakamura: J. Biomed. Mater. Res. A 46 (1999) 228–235.
• 11)   C.  Ohtsuki,  T.  Kokubo and  T.  Yamamuro: J. Non-Cryst. Solids 143 (1992) 84–92.
• 12)   A.  Oyane,  M.  Kawashita,  K.  Nakanishi,  T.  Kokubo,  M.  Minoda,  T.  Miyamoto and  T.  Nakamura: Biomaterials 24 (2003) 1729–1735.
• 13)   K.  Matsushita,  S.  Abe,  T.  Ishi,  S.  Kajiyama,  A.  Kotani,  M.  Saito,  T.  Masaoka and  T.  Suguro: Jpn. J. Chemother. 60 (2012) 319–326.
• 14)   D.H.  Nies: Appl. Microbiol. Biotechnol. 51 (1999) 730–750.
• 15)   T.N.  Kim,  Q.L.  Feng,  J.O.  Kim,  J.  Wu,  H.  Wang,  G.C.  Chen and  F.Z.  Cui: J. Mater. Sci. Mater. Med. 9 (1998) 129–134.
• 16)   C.  Prinz,  M.  Elhensheri,  J.  Rychly and  H.G.  Neumann: J. Biomater. Appl. 32 (2017) 139–149.
• 17)   K.  Gupta,  R.P.  Singh,  A.  Pandey and  A.  Pandey: Beilstein J. Nanotechnol. 4 (2013) 345–351.
• 18)   M.  Mondal,  K.  Trivedy and  S.N.  Kumar: Caspian J. Environ. Sci. 5 (2007) 63–76.
• 19)   W.  Gu,  D.W.  Bousfield and  C.P.  Tripp: J. Mater. Chem. 16 (2006) 3312–3317.
• 20)  J.F. Moulder, W.F. Stickle, P.E. Sobol and K.D. Bomben: Handbook of X-ray Photoelectron Spectroscopy, (Physical Electronics USA, Inc., Minnesota, 1995) pp. 68–69, 86–89.
• 21)   A.M.  Sookne and  M.  Harris: J. Res. Natl. Stand. 23 (1939) 299–308.
• 22)   J.  Duchoslav,  R.  Steinberger,  M.  Arndt and  D.  Stifter: Corros. Sci. 82 (2014) 356–361.
• 23)   A.  Takeuchi,  C.  Ohtsuki,  T.  Miyazaki,  M.  Kamitakahara,  S.  Ogata,  M.  Yamazaki,  Y.  Furutani,  H.  Kinoshita and  M.  Tanihara: J. R. Soc. Interface 2 (2005) 373–378.
• 24)   R.B.  Thurman,  C.P.  Gerba and  G.  Bitton: Crit. Rev. Environ. Control 18 (1989) 295–315.
• 25)   R.M.  Sterritt and  J.N.  Lester: Sci. Total Environ. 14 (1980) 5–17.
• 26)   A.  Sirelkhatim,  S.  Mahmud,  A.  Seeni,  N.H.M.  Kaus,  L.C.  Ann,  S.K.M.  Bakhori,  H.  Hasan and  D.  Mohamad: Nano-Micro Lett. 7 (2015) 219–242.
• 27)   K.  Hirota,  M.  Sugimoto,  M.  Kato,  K.  Tsukagoshi,  T.  Tanigawa and  H.  Sugimoto: Ceram. Int. 36 (2010) 497–506.
• 28)   E.  Padan,  D.  Zilberstein and  S.  Schuldiner: Biochim. Biophys. Acta 650 (1981) 151–166.
• 29)   J.F.  Siqueira and  H.P.  Lopes: Int. Endod. J. 32 (1999) 361–369.