Characterization by Electron Spin Resonance Spectroscopy of Reactive Oxygen Species Generated by Titanium Dioxide and Hydrogen Peroxide

Abstract

The influence of reactive oxygen species (ROS) on the surface modification of titanium implants and osseointegration is unclear. The aim of this study was to evaluate the ability of titanium dioxide (TiO₂) to generate ROS in the presence of H₂O₂ and to determine whether any ROS thus generated play a role in osseointegration, as measured by electron spin resonance (ESR) spin-trapping with 5,5-dimethyl-1-pyrolline-N-oxide (DMPO). We demonstrate that TiO₂ together with H₂O₂ generated hydroxyl radicals (HO^•), as shown by a time-dependent increase in the spin concentration of the ESR signal for the DMPO-OH spin adduct, indicating HO^• generation. Interestingly, irradiated TiO₂ with H₂O₂ generated the superoxide (O₂ ^•-), as shown by an increase in the spin concentration of the signal for the DMPO-OOH spin adduct, indicating O₂ ^•- generation during the period of irradiation (0–5 min). These results suggest that ROS generated from the TiO₂ layer may be involved in creating appropriate conditions for the osseointegration of dental implants into alveolar bone tissues.

INTRODUCTION

The biocompatibility of pure titanium dental implants placed in contact with alveolar bone has been considered to be involved in osseointegration, where the development of bone occurs in close apposition to the metal surface (LeGeros and Craig, 1993). When implanted, titanium is subjected to complex and variable molecular events that may significantly affect osseointegration (Gristina, 1994). Changes in the local tissue pH and the generation of biologically relevant molecules—such as reactive oxygen species (ROS) and reactive nitrogen species (RNS) during wound healing, tissue remodeling, or both—may alter the surface oxide of the titanium (Sundgren et al., 1985; Gristina, 1994; Taylor et al., 1996; Yang et al., 2003, 2004; Lin and Bumgardner, 2004). However, the role played by ROS generation at the titanium surface in osseointegration remains to be elucidated. Thus, we need to determine the details of the ROS generation occurring at the bone-titanium interface, a subject currently under intense investigation.

Following the insertion of an implant, the subsequent inflammatory response is accompanied by an oxidative burst and production of ROS, which could modify the biological titanium dioxide (TiO₂) layer (Gristina, 1994; Taylor et al., 1996; Lin and Bumgardner, 2004). The ROS include superoxide (O₂ ^•-), hydrogen peroxide (H₂O₂), hydroxyl radical (HO^•), and singlet oxygen (¹O₂) that are generated from an oxidative burst by the enzymatic pathways of the inflammatory cell (Babior, 1984; Lee et al., 2000a) in alveolar bone. The powerful oxidant ROS, HO^•, could be produced via the Fenton reaction in the presence of biological free iron (Halliwell and Gutteridge, 1992), and could lead to various pathophysiological phenomena such as periodontitis (Chapple, 1997) or temporomandibular disease (Kawai et al., 2000).

The successful osseointegration of a titanium implant surgically placed in bone is accompanied by several biochemical and transport processes that can lead to adhesion at the implant-tissue surface. The implant surface consists of a TiO₂ layer, which forms spontaneously, onto which biomolecules are adsorbed and desorbed, possibly followed by modification or fragmentation (Kasemo, 1983; Tengvall et al., 1989a). The interaction between alveolar bone and ROS from the TiO₂ layer may be critical in osseointegration, such as in biological fixation of titanium dental implants. However, the role of ROS in osseointegration is not fully understood, and it remains to be determined if the generation of ROS occurs during osseointegration with titanium dental implants in vitro or in vivo.

Our laboratories have been developing the in vitro biomedical application of electron spin resonance (ESR) for detecting ROS, such as O₂ ^•- (Lee et al., 2000a), H₂O₂ (Kiyose et al., 1999), ¹O₂ (Ishibashi et al., 1996; Yoshino et al., 2002), and HO^• (Lee et al., 2002). A recent study showed that ESR spectra detected ROS generation in irradiated TiO₂ (Konaka et al., 1999), suggesting that O₂ ^•- or ¹O₂ is produced by irradiated TiO₂. It is unclear if this occurs with inflammatory disease or is relevant to the interaction with other ROS. The generation of ROS has been shown to accompany the inflammation occurring during and after implant insertion in vivo. One critical biologically relevant event is the O₂ ^•- generation by polymorphonuclear leukocytes (Babior, 1984; Lee et al., 2000a) during the inflammatory process. Since the O₂ ^•- is dismuted to H₂O₂ by superoxide dismutase (SOD), the interaction between the implant titanium and H₂O₂ may be a crucial part of the inflammatory/healing process following implant insertion. The TiO₂ layer initially present may be partly reformed to a TiOOH matrix, due to the interaction with H₂O₂, suggesting that the fates of O₂ ^•-/H₂O₂ at the titanium surface are important processes that must be understood before the basis for the biocompatibility of titanium can be comprehended (Tengvall et al., 1989a,b; Lutz, 1990). However, the role of ROS in osseointegration is not well-understand. and it remains to be determined if ROS are generated by titanium in vitro and/or in vivo. Therefore, the aim of the present study was to evaluate the ability of TiO₂ to generate ROS in the presence of H₂O₂, and to determine whether the ROS generated thereby play a role in osseointegration. In this study, we detected ROS in both TiO₂ and irradiated TiO₂ in the presence of H₂O₂ using an ESR spin-trapping technique.

MATERIALS & METHODS

Materials

The following materials and chemicals were used: TiO₂ (anatase form, particle size of 20 nm, 99.0% pure) was purchased from Ishihara Sangyo Kaisha Ltd. (Tokyo, Japan), and hydrogen peroxide was purchased from Wako Pure Chemical Industries (Osaka, Japan). 5,5-Dimethyl-1-pyrolline-N-oxide (DMPO) was purchased from Labotec (Tokyo, Japan). All other reagents were of analytical grade.

ESR Measurement

The experimental protocol and time schedules are shown in Fig. 1. ROS were generated by TiO₂ (final concentration, 0.15% wt/vol) in 3.0% H₂O₂, or by photoexcitation (385 nm) of TiO₂ (Fig. 1B ) with H₂O₂ (final concentration, 3.0%). Test samples were directly irradiated in a microwave cavity with focused light from a UV lamp (Radical Research, SUPERCURE-203S, Tokyo, Japan) operating at 365 nm. The production of ROS by photoexcited TiO₂ was verified by ESR. ESR observations were performed with a JES-RE 3X, X-band, spectrometer (JEOL, Tokyo, Japan) connected to a WIN-RAD ESR Data Analyzer (Radical Research, Tokyo, Japan) at the following instrument settings: modulation amplitude, 0.063 mT; sweep width, 5 mT; sweep time, 1 min; time constant, 0.03 sec; microwave power, 8 mW; and magnetic field, 335.5 ± 5 mT. The component signals in the spectra were identified and quantified as reported (Lee et al., 2000a). We compared the double integrals of DMPO-OOH experimental spectra with those of a 1 μM 4-hydroxy-2,2,6,6-tetramethyl-piperidine-N-oxide (TEMPOL) standard measured under identical settings, to estimate the concentration of the O₂ ^•- adduct.

RESULTS

Hydroxyl Radical Generation from TiO₂ with H₂O₂

We investigated ROS generation measured by ESR spin-trapping with DMPO. After the addition of H₂O₂ to TiO₂, we obtained primarily a characteristic DMPO-OH spin adduct with hyperfine splitting constant (hfcc), giving rise to 4 resolved peaks (A_N = A_H ^β = 1.49 mT) (Fig. 2A). These hfcc values suggest that HO^• was generated from TiO₂ with H₂O₂ (Buettner, 1987). The spin concentration of the ESR signal for the DMPO-OH spin adduct increased in a time-dependent manner, and reached a maximum after 30 min (Fig. 2B). We did not detect any ESR spectrum during irradiation of DMPO solution with H₂O₂ in the absence of TiO₂ (data not shown).

Superoxide Generation from Irradiated TiO₂ with H₂O₂

Using ESR with DMPO as a spin trap, we observed a characteristic DMPO-OOH spin adduct with hfcc, giving rise to 12 resolved peaks (A_N = 1.43 mT, A_H ^β = 1.17 mT, and A_H ^γ = 0.125 mT) during irradiation of TiO₂ in the presence of H₂O₂ (Fig. 3A ). These hfcc values suggest that O₂ ^•- was generated during irradiation of TiO₂ in the presence of H₂O₂ (Buettner, 1987). The spin concentration of the signal for the DMPO-OOH spin adduct increased with the period of irradiation (0–5 min), reaching a maximum after 1 min of irradiation (Fig. 3B). Ten min after the irradiation was stopped, TiO₂ in the presence of H₂O₂ produced an ESR spectrum that changed completely from the DMPO-OOH adduct to the DMPO-OH adduct (Fig. 3A ). We did not detect any ESR spectrum during irradiation of the DMPO solution containing H₂O₂ in the absence of TiO₂ (data not shown).

Effects of H₂O₂ on ROS Generation from Irradiated TiO₂ and H₂O₂

High concentrations of H₂O₂ (from 0.3% to 30%) generated predominantly DMPO-OOH spin adduct signals, indicating O₂ ^•- generation, while low concentrations (from 0.0003% to 0.03%) generated DMPO-OH spin adduct signals, indicating HO^• generation (Fig. 4 ). These results suggest that both ROS generation and ROS species from irradiated TiO₂ would be dependent on the H₂O₂ concentration.

DISCUSSION

Irreversible failure of the integration of biomaterials and implants can clearly be caused by the primary adhesion and colonization of biomaterial surfaces and adjacent damaged tissue by micro-organisms (Gristina, 1994). Equally importantly, however, aseptic (non-infectious) integration failure may also occur, due to the biomaterial surface or biomaterial particulate activation of host cellular and humoral immune system responses, including normal wound-healing inflammatory reactions (Barth et al., 1988; Myrvik et al., 1989; Muller et al., 1991; Gristina, 1994). It is of interest to know how titanium, including titanium dioxide, behaves under the conditions of dental implant insertion associated with an inflammatory response caused by surgical trauma.

It is therefore highly significant that biomaterials can trigger the generation of ROS upon initial contact with alveolar macrophages, which may result in protein oxidative damage in the adjacent tissue. This could decrease the ability of phagocytes to generate an oxidative burst when subsequently encountering phagocytosed bacteria during their response to infection, as well as their release of cytokines (Barth et al., 1988; Gristina, 1994; Lin and Bumgardner, 2004). Successful osseointegration may involve the gradual formation of a hydrated titanium peroxy gel layer on the TiO₂ layer of the implant. They provided an ideal environment for protein interactions involved in the process of osseointegration as well as material with the ability to act as a trap for O₂ ^•-, which have been identified within the gel matrix (Tengvall et al., 1989b). Therefore, the interaction between ROS generated from inflammatory cells (Babior, 1984) and ROS generated from the TiO₂ layer could be important in repair and in osseointegration with alveolar bone.

When titanium prosthetics are implanted into the jaw, the local concentrations and the duration of production of ROS during the inflammatory response are unknown. Therefore, questions remain regarding whether ROS generation from inflammatory cells and titanium-mediated ROS generation modulate osseointegration after implant surgery, since there has been no direct monitoring of the concentrations of titanium-mediated ROS generation. The present study clearly provides direct evidence that TiO₂ with H₂O₂ generated HO^• (Fig. 2), and that irradiated TiO₂ with H₂O₂ generated O₂ ^•- (Fig. 3 ), with DMPO as an ESR spin trap.

We have reported μmole levels of ROS generated by inflammatory cells such as PMNs after inflammatory stimulants (Lee et al., 2000a; Lee et al., 2002). Interestingly, the level of ROS generated from TiO₂ with H₂O₂ was at the same μmole level as was observed in our current study (Figs. 2, 3 ). It may be possible that fewer ROS are generated in vivo in comparison with our experimental system, while the ability of ROS to be generated from TiO₂ may partially contribute to the repair/wound-healing inflammatory response after titanium implant surgery. We have already reported our evaluation of oxidative stress induced by ROS in rodent maxillofacial regions using in vivo ESR (Lee et al., 2000b); now a similar study is needed to determine the extent of ROS generation in vivo after titanium implant treatment.

In an electrolyte solution, the electron and hole (h⁺) created by photoexcited (~ 400 nm) TiO₂ can reduce or oxidize chemical species on the surface of a TiO₂ layer. It is well-known that the hole oxidizes a water molecule to yield HO^•, and the electron reduces oxygen to give O₂ ^•- or H₂O₂. During such ROS generation, a question arises regarding how the interaction of HO^• with O₂ ^•- or H₂O₂ with TiO₂ or irradiated TiO₂ occurs. It has already been reported that no ESR-detectable HO^• formed in a TiO₂ and metallic Ti system (Tengvall et al., 1989a,b), while another study showed that ESR spectra detected ROS generation in irradiated TiO₂ (Konaka et al., 1999). This has led to the present controversy and confusion regarding the role of ROS generation in TiO₂ function, including its biocompatibility. Therefore, we have performed studies to assess the extent of ROS generation using ESR and monitor ROS spin concentrations directly. The present ESR study indicated that non-irradiated TiO₂ produced HO^• in the presence of H₂O₂ (Fig. 2), while O₂ ^•- was generated by irradiated TiO₂ (Fig. 3 ).

The photodynamic properties of TiO₂ are well-known. When exposed to UVA (320–400 nm), the reduction-oxidation (redox) activity of TiO₂ has a significant bactericidal activity in vivo (Ireland et al., 1993). This property of photoexcited TiO₂ has been predicted to be useful, eventually, for the photodynamic treatment of cancer (Cai et al., 1992; Kubota et al., 1994). The present ESR study indicated that non-irradiated TiO₂ produced HO^• in the presence of H₂O₂ (Fig. 2), while O₂ ^•- was generated by irradiated TiO₂ (Fig. 3 ). The spin concentration of O₂ ^•- increased with the time of irradiation (from 0 to 5 min) (Fig. 3B ), while HO^• generation was seen instead of O₂ ^•- after irradiation (Fig. 3A ). From these data, we can consider that this irradiated TiO₂ with H₂O₂ may actually produce predominantly O₂ ^•- in a biological system.

Furthermore, high concentrations of H₂O₂ led predominantly to the generation of O₂ ^•-, while low concentrations led to HO^•, indicating that the H₂O₂ concentration could be critical for generating the amounts and species of ROS from irradiated TiO₂ (Fig. 4 ). Thus, it may be possible that ROS generated from the TiO₂ layer may lead to appropriate conditions promoting osseointegration into alveolar bone tissues after dental implantation. Thus, future work will need to evaluate the influence of ROS formed by irradiated (O₂ ^•-) or non-irradiated (HO^•) TiO₂ on the TiO₂ layer of titanium. This would contribute to our understanding of the role of ROS in osseointegration and biocompatibility of titanium implant treatment. It may also provide a platform for the testing of novel clinical strategies for the treatment of titanium implants with irradiation.

In conclusion, we have demonstrated that TiO₂, together with H₂O₂, generated HO^•, and that irradiated TiO₂ with H₂O₂ generated O₂ ^•-, with DMPO as an ESR spin trap. Regarding the biocompatibility of titanium dental implants, the generation of ROS, including O₂ ^•-, H₂O₂, and HO^•, at the titanium surface may be involved in creating appropriate conditions for the osseointegration of dental titanium implants into alveolar bone tissues.

Figure 1.

Schematic diagram of the experimental protocols used for non-irradiated and irradiated TiO₂. Experimental protocol for non-irradiated TiO₂ (A) or for irradiated (365 nm) TiO₂ (B) in the presence of H₂O₂ for ESR measurement.

Figure 2.

Hydroxyl radical (HO^•) generation from TiO₂ in the presence of H₂O₂. (A) The time course of ESR spectrum from TiO₂ measured by EPR spin-trapping following the addition of H₂O₂ at time 0 and with DMPO as a spin trap. (B) Monitoring of the spin concentration of the HO^• generated from TiO₂ in the presence of H₂O₂. All data are presented as the mean (n = 3–5). Experimental conditions are described in “MATERIALS & METHODS” and illustrated in Fig. 1 .

Figure 3.

Superoxide (O₂ ^•-) generation from irradiated TiO₂ in the presence of H₂O₂. (A) The time course of ESR spectrum during and after irradiation (365 nm) of TiO₂ measured by EPR spin-trapping following the addition of H₂O₂ at time 0 and with DMPO as a spin trap. (B) Monitoring of the spin concentration of the O₂ ^•- generated from irradiated TiO₂ (5 min) in the presence of H₂O₂. Each column represents the mean (n = 5). Experimental conditions are described in “MATERIALS & METHODS” and illustrated in Fig. 1 .

Figure 4.

Dose-dependent effects of H₂O₂ on reactive oxygen species (ROS) generation from irradiated TiO₂. The ESR spectrum of irradiated (365 nm) TiO₂ (0.15%, wt/vol) measured by EPR spin-trapping following the addition of different concentrations of H₂O₂ (0.0003%–30%) at 1 min and with DMPO (880 mM) as the spin trap.