Bactericidal properties of a titanium-peroxy gel obtained from metallic titanium and hydrogen peroxide ~~~~
~
~
~
P. Tengvall, E. G . Hornsten, H. Elwing, and I. Lundstrom Department of Physics and Measurement Technology, Linkoping lnsfitute of Technology, S-581 83 Linkiiping, Sweden
A stable titanium-peroxy-radical complex is formed when metallic titanium interacts with hydrogen peroxide. The radical appears as one component in an aqueous gel formed when excess peroxides have been (catalytically)decomposed. The interaction between titanium and hydrogen peroxide may be of importance also in vivo during an inflammatory response at the implant. We report in this paper on the bactericidal effects of the titanium gel in the lacto- and myeloperoxidase-halogen systems. Escherichia coli viable count was used to evaluate the bactericidal properties of the gel and of H202for comparison. The gel
had only small or no toxic properties at high dilutions. Higher concentrations of the gel had bactericidal properties similar to those of H202.The results indicate that at physiological pH, the decomposition products of the gel are titanium hydroxide (Ti(IV)(OH-)4)and hydrogen peroxide (H202).It was found that the gel probably oxidizes glutathione directly in contrast to H202,which needs a peroxidase to do so. A model for the interaction between titanium and hydrogen peroxide is suggested. Its consequences for the properties of titanium in vivo are also discussed.
INTRODUCTION
It has been observed that an oxide like layer grows on titanium and stainless steel implants in viv~.’-~ The oxide growth was largest in tissue with high metabolic activity. Model experiments in vitro have shown that the interaction between hydrogen peroxide and a titanium surface may be a source for this ~xidation.~” It also has been observed that titanium-peroxy and peroxide complexes are formed when metallic, spontaneously oxidized, titanium interacts with hydrogen peroxide. Furthermore, both the hydrogen peroxide and titanium complexes are catalytically decomposed at the titanium surface. An aqueous titanium gel is formed when excess peroxides have been c o n s ~ m e dSince . ~ ~ one important component of this gel may be the titanium-peroxy-radical, we call the gel a Ti-peroxy gel. The composition of the gel is suggested to be:5 Ti(IV)0,2-(OH-), - Ti(IV)O; (OH-), - Ti(IV)(OH-), - Ti(IV)O, - n - H,O The gel thus contains a superoxide or perhydroxyl radical, which may be formed in the gel and is stable for a long period of time.3,4The gel has oxidizJournal of Biomedical Materials Research, Vol. 24, 319-330 (1990) 0 1990 John Wiley & Sons, Inc. CCC 0021-9304/90/030319-120$04.00
TENGVALL ET AL.
320
ing properties with a redox potential larger than that of the Fe(II)/Fe(III)couple (0.77 V). The fraction of the different components in the formula above may vary from gel to gel. In a typical gel, the total Ti concentration is determined to be about 40 mM and oxidizing component(s) to be about 10 mM5. Since the gel is produced from metallic titanium and hydrogen peroxide in distilled water, it is very clean and does not contain any counter ions in contrast to gels produced from titanium chloride and ammonium hydroxide solution^.^,^ The formation of complexes between hydrogen peroxide and titanium ions from metallic titanium is of particular interest in connection with the known biocompatibility of titanium.&'' An inflammatory response is normally triggered when a foreign material is inserted into living tissue. Hydrogen peroxide and oxygen radicals are thereby formed. The interaction between an implant surface and hydrogen peroxide is therefore not only of theoretical interest; it may have a direct connection with the behavior of an implant. One speculation is that the interaction between a titanium implant and hydrogen peroxide may lead to a modification of the initial inflammatory response. Since the Ti-peroxy gel described above has oxidizing properties in itself, and furthermore, most probably degrades under the formation of H,O,, we found it of interest to investigate the properties of the gel in a situation where H,O, is known to play an important biological role. The present communication deals therefore with the bactericidal properties of the titanium-peroxy gel in two systems where H,O, is active, namely the myeloperoxidase (MPO) + chloride, and the lactoperoxidase (LPO) + iodide-systems. In the presence of sufficiently large concentrations of H,O,, these systems are known to have toxic effects on microorganism^.^^-'^
MATERIALS AND METHODS
Prior to use, all glassware, iodide, and chloride (from Merck) diluted in buffer, buffers, soy agar (Pharmacia), 0.35% agarose gel (Pharmacia), and distilled water were autoclaved for 30 min. Taurine (2-aminoethanesulfonic acid) (Sigma) was sterile filtered prior to use. The E . coli ATCC 12932 was incubated at 37°C overnight in lactose peptone broth, Oxoid CM 305, 42 g/L. Prior to use, the culture was washed twice in the buffer used in the actual experiment. The bacterial densities obtained during the experiments ranged from lo6 cells/mL to 10' cells/mL. The bacterial suspension was used in experiments within 30 min of washing. Incubation during the experiments was performed at 37°C in a water bath with agitation at 160 osc/min. Serial dilutions of incubated bacterial suspension with 50-pL samples in 4.5 mL in the buffer used were made in three steps. Three drops from each step were dropped on nutrient agar plates. Viability was assayed by viable count (colony forming units on nutrient agar; Oxoid CM 32, 28 g/L) after overnight incubation. Sodium-phosphate buffer 0.1 M pH 7 was used in all experiments
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
321
with MPO and 0.1 M sodium-acetate buffer pH 5 was used in the LPOexperiments. In some MPO-experiments, 0.67 M sodium sulphate was used to maintain tonicity when chloride was omitted. All experiments, except those in Figures 1and 2, were performed in duplicate or triplicate. The Ti-gel was prepared from 710-pm Ti-powder (Johnson Matthew) and 30% hydrogen peroxide (Merck) according to a procedure described el~ewhere.~ In brief 4 g of Ti-powder was incubated in 100 mL of 30% H,Oz in a water bath at room temperature. During 1 week, excess peroxide was decomposed and the solution formed a Ti-peroxy gel, with a pH of 4, at the same time as oxygen evolution ceased. When needed, the gel was diluted in deionized water. Myeloperoxidase (E.CI1.11,1.7)in lyophilized solid form (Calbiochem, lot 801094) was used without further purification. The enzyme was diluted in buffer and used within 45 min. The activity of the enzyme was determined with a method described by Klebanoff et a1.I' Ten microliters of the MPO dissolved in sodium phosphate buffer pH 7 was added to a 1-cm light path cuvette containing 3 mL of the fresh assay solution made from 26.9 mL H,O, 3 mL of 0.1M sodium phosphate buffer pH 7, 0.1 mL of 0.1 M H,O,
Figure 1. E. coli viability as function of iodide and hydrogen peroxide concentrations in the LPO-system. Bactericidal activity is detectable at less than 1 p M iodide. Minimal viability is achieved at about 50 p M H 2 0 2 concentration.
TENGVALL ET AL.
322
I
1.5-10-'
I
1.5*10-6
I
1.5-10-5
log (Ti-PEROXVGEL vhr]
1.5-10-4
1.5-10-3
[Ti [lV)O; lOY-l,M] IN ATYPICAL GE
Figure 2. Dependence of E . coli viability on the concentration of the Tiperoxy gel in the MPO-chloride-Ti peroxy gel system.
and 0.048 mL guaiacol (Sigma). The absorbance of the solution was determined at 470 nm during 3 min. The change in absorbance per minute was calculated from the initial rate of change. The tetraguaiacol product formed has an extinction coefficient of 26.6 mM-' cm-' at 470 nm. It is formed due to oxidation of quaiacol. The following simple formula was used to calculate the activity of MPO. units/mL = 45.1
X
AABS/min
The numerical factor was determined from the results in ref. (11)and the used volumes of the reagents. The activity was determined to be 40 mu/mL. Lactoperoxidase (E.C.1.11.1.7.) lyophilized powder (Sigma) was used without further purification. Prior to use, 0.1 mg of the enzyme was diluted in 1mL distilled water and used within 45 min. The activity was assayed in the following way. l4 A 2.5-mM sample of 4-aminoantipyrine (Sigma) was mixed with 810 mg phenol (Sigma) in 40 mL distilled water. Then, another 25 mg 4-aminoantipyrine was added and the mixture stored in dark. A 1.4mL volume of the phenol/aminoantipyrine solution was then mixed with 1.5 mL of 1.7 mM H,O,-solution in distilled water. A 0.1-mL sample of the
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
323
enzyme solution was added and the change in light absorption of the solution at 510 nm was recorded for 2 min, The activity was calculated according to Units/mL =
AABS/minute 6.58 x mg enzyme/mL reaction mixture
The numerical factor was obtained from ref. [14] and the used volumes of the reagents. The activity was determined to be 200 mu/mL. Glutathione (GSH) (Calbiochem) oxidation experiments were performed with 5.5-dithio-bis (2-nitrobenzoic acid) (DTNB) (Sigma) as a sulfhydryl reagent. 0.14 M sodium phosphate buffer pH 7 was used during these experiments. Except for DTNB, 300 p L of each additive, dissolved in the buffer were added into a 1-cm light path cuvette. The gel was diluted to 50% (v/v) in the buffer before addition. The samples were incubated for 30 min in dark at room temperature. Thereafter, 300 pL of the DTNB-solution was added and the formation of a colored sulfhydryl-DTNB product was measured within 30 s spectrophotometrically at 412 nrn.I5 The final sample volume was 1.8 mL and, when necessary, buffer was added to keep the volume constant. The final concentrations of the reagents are given in Figure 3a.
RESULTS
To learn about the chosen test systems, extensive experiments were performed with the LPO-H,O,-I--system. Optimal bactericidal action was sought in this system by varying the H 2 0 2and I- concentrations (see Fig. 1).The dilution series show that the E . coli viability was minimal at 50 pM H,02. Bactericidal action was observed in the LPO-Ti-peroxy gel-I--system as found from Table I. When LPO, I- or both were omitted, bactericidal effects were not observed either for H20, or for the Ti-peroxy gel. Furthermore, when C1- was used as the halide in the LPO-system neither H202nor the Ti-peroxy gel showed bactericidal effects. In the MPO-Ti peroxy gel or H20, system bactericidal effects were obtained with C1- as the chosen halide (see Table 11). Toxicity was not observed when myeloperoxidase, chloride or both were excluded. The bactericidal effect of the gel decreased when the gel was diluted (see Fig. 2). Different mechanical treatments of the gel or change of order of addition of the substances did not alter the bactericidal properties of the gel. Addition of catalase, an H,O,-scavanger, did not abolish the bactericidal effect of the gel (see Table I). Addition of taurine (2-aminoethane sulfonic acid) in the MPO-CI- system abolished the bactericidal activity completely both of the H,O, and the Tiperoxy gel (Table 11).
TENGVALL ET AL.
324 412 nrn
1.8
1 6 . 1.4
.
1.2
.
1.0 . 0.8 .
0.6 . 0.4
.
02 .
0
-
G s H G s H G s H G s H G s H G DTNB Ti-gel Ti-gel Lpo Ti-gel DTNB Cat IL p o L DTNB DTNB I-
DTNB
s H G s H G s H Ti-gel p o
ICat DTNB
33 3%
I-
DTNB
Figure 3a. Glutathione (GSH) oxidation in the Ti-peroxy gel or H20zperoxidase-halogen systems. The experiments show that GSH is oxidized by the Ti-peroxy gel but not by H202and that the peroxidase-halogen systems potentiate this oxidation. The concentrations used were [GSH] = 0.15 mM, [DTNB] = 0.15 mM, [Ti-peroxy gel, oxidizing equivalent] = 2.5 mM, [I-] = 25 pM, [H202]= 1 mM, [LPO] = 200 mu/mL and [cat] = 2 u/mL. The components were incubated for 30 min, whereafter DTNB was added and the change in absorbance measured within 30 s. The change in adsorbance is a measure of the amount of unoxidized sulfhydryls.
Figure 3a shows the results of glutathione oxidation experiments for several combinations of gel, H,02, LPO, and I-. In contrast to H,Oz, addition of Tiperoxy gel into a reduced glutathione sample caused chemical alteration of glutathione in such a way that the formation of the colored sulfhydrylDTNB product formed as a reaction product between reduced glutathione and DTNB decreased. Since the Ti-peroxy gel per se appeared to oxidize (or influence) glutathione directly the effect of incubation time in the gel on the number of unoxidized glutathione molecules was also studied (Fig. 3b). DISCUSSION
The results for H20, behavior in peroxidase systems is in accordance with earlier findings.'2,'6The results in Figure 1 support, e.g., recent observations of an inactivation of LPO at H,O, concentrations above 50 pM.16
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
325
1.6
1.2
0.8
Time (min) 0.4 0
I
I
I
20
40
60
Figure 3b. Glutathione oxidation versus incubation time in 40% v/v Tiperoxy gel in sodium phosphate buffer pH 7. The glutathione concentration was 0.25 mM.
It can be concluded that the Ti-peroxy gel and degradation products from it are not themselves deleterious to E. coli. The gel had at high concentrations, however, together with both LPO and MPO and a proper halide bactericidal properties similar to those of hydrogen peroxide. It was observed that the viscosity of the gel increases when it is chemically reduced. Diffusion limitations or enzyme immobilization may therefore explain why catalase, a known H,02 scavanger, did not abolish the bactericidal effects of the gel. H,O, is known to form hypochlorous acid, HOCl, in the MPO-C1- system17,18
H,O,
+ Cl- + H+ 3 HOCl + H,O
Taurine forms N-chloramines from HDCl which are not liposoluble but oxidize extracellular t h i ~ l s . ' ~The , ~ ' fact that taurine abolished the bactericidal properties of both H202and the gel (Table 11) indicates that (a) the bactericidal compounds in the peroxidase-halogen-H,O, or Ti peroxy gel system are formed outside the bacteria, (b) oxidation of thiol groups outside the cell is not enough to kill E. coli, (c) the bactericidal mediator most probably is a hypohalous acid. The hypohalous acid reacts with ammonium to form liposoluble ammonium chloride, which is a powerful oxidant and N-chloramines which are particularly reactive with thiols (see, e.g., ref [20]). Oxidation of E . coIi by
TENGVALL ET AL.
326
TABLE I Bactericidal Effect of LPO, H202, Ti-Peroxy Gel and Iodide Viable cell count Organisms per mL
Supplements None Ti-peroxy gel Agarose Ti-peroxy gel + LPO Agarose t LPO LPO + H 2 0 2+ ILPO + Ti-peroxy gel + ILPO + agarose + 1Agarose + ILPO + H202 LPO + Ti-peroxy gel + H202 LPO + agarose + H,O, LPO + Ti-peroxy gel + H,02 + ILPO + agarose + H,02 + ILPO + Ti-peroxy gel + I- + catalase LPO + agarose + I- + catalase LPO + H,O, + agarose + I- + catalase LPO + H,O, + Ti-peroxy gel + 1- + catalase
4 x 106 3 x 106 4 x 106 1 x 106 3 x 106 < 100 < 100 400 x lo6 7 x lo6 10 x 106 2 x 106 7 x 106 < 100 < 100
< 100 7 x 106
< 100 < 100
The final sample volume was 2 rnL. 0.1 M Sodium acetate buffer p H 5 was used. 0.5 mL of each supplement solution was added. When a supplement was omitted, the same volume buffer was added. The final concentrations of the supplements and order of addition was as follows: LPO 200 mu/mL, E . coli suspended in the buffer, H202 50 pM or the gel, eventually including catalase (2 pg/mL), and I- 1 pM. Incubation 3045 rnin at 37°C in a shaking water bath.
TABLE I1 Bactericidal Effect of MPO H20,, Ti-Peroxy Gel and Chloride ~
Viable count Organisms per mL
Supplements None MPO MPO + Ti-peroxy gel MPO + H202 MPO + H202+ Ti-peroxy gel H,02 + Ti-peroxy gel Chloride MPO + chloride MPO + chloride + Ti-peroxy gel MPO + chloride + Ti-peroxy gel + taurine MPO + chloride + H 2 0 2 MPO + chloride + H20, + taurine Chloride + H,O, + Ti-peroxy gel MPO + chloride + H20,+ Ti-peroxy gel
I
I
I
5x 1x 0.3 x 1x 0.3 X 0.3 x 1x 1x
107 107 107 107 16
lo7
107 107 < 100 0.1 x lo7 < 100 7 x 107 0.4 x 107 < 100
The final sample volume was 1 mL. Sodium phosphate buffer 0.1 M p H 7 was used. The final concentrations of the supplements land order of addition was as follows: E. coli suspended in buffer, chloride 0.1 M, sodium sulphate when chloride was omitted 0.15 M, taurine 10 mM, H 2 0 220 p M , Ti-perqxy gel 0.2 mL/sample and MPO 40 mu/mL. Incubation 30-45 min at 37°C in a shaking water bath.
BACTERlCIDAL PROPERTIES OF TITANIUM-PEROXY GEL
327
hypochlorous acid also gives rise to hydrolysis of cytosolic nucleotide phosphoanhydride bonds and loss of metabolic energy.21 It was observed that gluthatione was oxidized (or structurally altered) directly by the Ti-peroxy gel but not by H,O, (Fig. 3a). The Ti-peroxy gel has thus oxidizing properties differing from those of H,O,. It was found that the LPO-I- system potentiated the oxidation of glutathione by both H202and the Ti-gel, Tlus is in accordance with earlier results which have shown sulfhydryl oxidation in the lactoperoxidase-hydrogen peroxide-iodide system.22In order to check that the gel actually interacted with the glutathione, some experiments were made where glutathione was incubated in the gel for different lengths of time before the addition of the color reagent (DTNB) (Fig. 3b). The experiment shows that the longer the time of incubation of glutathione in the Ti-peroxy gel, the smaller amount of oxidizable compound there will be left for the sulfhydryl-DTNB reaction. The result also indicates that the decreasing sulfhydryl-DTNB formation is not caused by the gel-DTNB interaction. The Ti-peroxy complex which is a strong oxidant is thus most probably initially reduced by the biological environment to form Ti-peroxide, Ti(IV)O;-(OH-),. The Ti-peroxide appears to possess bactericidal properties in the presence of LPO or MPO and the proper halogen. Several alternative reaction pathways may explain this observation. Hydrogen peroxide incorporated in the gel may be released during the degradation of the gel. Another possibility is that LPO or MPO may act directly on the Ti-peroxide to form (oxidizing) halide compounds. It is also possible that the Ti-peroxide gel reacts with water molecules to form Ti-hydroxide, Ti(IV)(OH-),, and free or weakly bound H,02,
-
+ H,O Ti(IV)HO,(OH), Ti(IV)HO,(OH-), & Ti(IV)(OH-), + H,O,
Ti(IV)O;-(OH-),
Increasing pH shifts the reactions to the right. These reactions are suggested to take place at physiological pH. The H20, is now used as a substrate by the peroxidases. If the HO; anion leaves the Ti-coordination sphere, it may through proton uptake form H,02 at neutral or weakly acidic conditions. Other reactions are also possible with the HOT anion, which eventually lead to H,O, format i ~ n . In ’ ~summary, the two most probable ways for the formation of hydrogen peroxide from a degrading Ti-peroxy gel at weakly acidic or neutral pH are 1) HO, abstraction from the Ti-coordination sphere with subsequent proton uptake; 2) water uptake by the Ti-peroxide giving rise to free or weakly bound hydrogen peroxide. Figure 4 is a tentative summary of the possible reactions between hydrogen peroxide and (oxidized) metallic titanium. H,O, takes part in the oxidation of the metal surface; participate in the formation of Ti-peroxy radicals and other Ti-complexes. Both Ti(1V)- and Ti(II1) may interact with H,O, during this process. H,02 is also catalytically decomposed at the metal surface. If the reaction conditions are proper, a titanium peroxy gel may be formed when
328
TENGVALL ET AL.
Figure 4. Suggested reaction pathways during interaction between titanium and hydrogen peroxide: When Ti-peroxide (1) is formed, it may, at neutral or weakly acidic conditions, spontaneously decompose to Ti-hydroxide and hydrogen peroxide (2) (main pathway). The titanium complexes as well as hydrogen peroxide are catalytically decomposed at the oxidized surface. This may form a Ti(N)O;(OH-),-radical which in turn may complex with other Ti-adducts (chelation as end result) or decompose to Ti-peroxide (4) through an oxidation of the surroundings. The Ti-peroxy gel eventually formed is degraded through reduction to Ti-peroxide (4). Alternatively, the radical may be formed from Ti(1II) and O2 as the catalytic Hp02-decomposition leads to a local increase of the oxygen pressure (5). During the processes above, the oxide layer on the titanium surface grows.
the excess peroxide has been d e c ~ m p o s e d .Through ~-~ degradation of the Ti-peroxide part of the gel, the compounds will then act as a source of H,O,. The reactions in Figure 4 may also occur in vivo at an implant surface with H,O, supplied from a respiratory burst. The concentrations of H,O, and Ticomplexes will, however, be too small to cause a Ti-peroxy gel formation although complexed hydrated Ti-oxides may be found in the surroundings of an implant. The MPO-H,O,-C1-system is of large importance during an inflammatory response. 20,24,25 The hypochlorous acid produced in this system forms ammonium chloride and chloramines under physiological conditions.” These oxidize thiol containing residues both extra- and intracellularly and in the cell membranes. Some of the reactions lead to an enhancement of the inflam-
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
329
matory response,” whereas others inactivate inflammatory mediators.26The interaction between H,O, and titanium leads to a smaller concentration of H,O, and to Ti-compounds which slowly release (a small concentration of) H,O, close to the implant surface. This interaction will most probably have pronounced effects on the local inflammatory response at the implant. Another interesting observation is that the Ti-peroxy gel interacts with glutathione by itself whereas hydrogen peroxide does that only in presence of peroxidases. Apparently, the Ti-peroxy and/or Ti-peroxide interact directly with thiols (see Fig. 3a, b), which also may influence an inflammatory response. Although we have no independent evidence, it is most likely that the gel oxidizes the sulfhydryl groups of glutathione. A hypothesis is, therefore, that one of the positive properties of titanium as an implant material in addition to small or no systemic effects from the end products, there may be a down regulation of the initial inflammatory response due to the effects discussed above. Preliminary model studies of the interaction between the titanium gel and polymorphonuclear granulocytesz7as well as some observations on titanium implants in inflamed tissue support such a hypothesis. More information has, however, to be collected before the role of titanium implants during the inflammatory response has been fully elucidated. Dr. Lars-Magnus Bjursten, Department of Anatomy, University of Gothenburg, is gratefully acknowledged for stimulating discussions on the properties of the titaniumperoxy gel. The research on the biocompatibility of titanium is supported by grants from the National Swedish Board for Technical Development and the Institute for Applied Biotechnology, Gothenburg, Sweden.
References 1. J. E. Sundgren, P. Bodo, and I. Lundstrom, ”Auger electron spectro-
2. 3.
4.
5.
6. 7.
8.
scopic studies of the interface between human tissue and implants of titanium and stainless steel,” J. Colloid Interface Sci., 110, 9-20 (1986). J.E. Sundgren, P. Bodo, I. Lundstrom, A. Berggren, and S. Hellem, “Auger electron spectroscopic studies of stainless-steel implants,” 1. Biomed. Muter. Res., 19, 663-671 (1985). P. Tengvall, H. Elwing, L. Sjokvist, L.M. Bjursten, and I. Lundstrom, ”Interaction between hydrogen peroxide and titanium: A possible role in the biocompatibility of titanium,” Biomateriuls, 10, 118-120 (1989). P.Tengvall, I. Lundstrom, L. Sjokvist, H. Elwing, and L. M. Bjursten, ”Titanium-hydrogen peroxide interaction, Model studies of the influence of the inflammatory response on titanium implants,” Biomuferials, 10, 166-180 (1989). P. Tengvall, H. Elwing, and I. Lundstrom, ”Titanium gel made from metallic titanium and hydrogen peroxide,” J. Colloid Interface Sci., 130, 405-413 (1989). J. Ragai, “Trapped radicals in titania gels,” Nature, 325, 703-705 (1987). J. Muhlebach, K. Muller, and G. Schwarzenbach, “The peroxo complexes of titanium,” Inorg. Chem., 9, 2381-2390 (1970). P.I. Brbnemark, ”Introduction to osseointegration in tissue integrated prostheses in Osseointegration in Clinical Dentistry, P. I. Brhemark,
TENGVALL ET AL.
330
9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
23. 24. 25.
26. 27.
G. A. Zarb, and T. Albrektsson (eds.), Quentessence, Chicago, 1985, Chap. 4. P. Thomsen, L. M. Bjursten, and L. E. Ericson, “Implants in the abdominal wall of the rat,” Sand. J. Reconst. Surg., 20, 173-182 (1986). T. Albrektsson, H.-A. Hansson, and B. Ivarsson, “Interface analysis of titanium and zirconium bone implants,” Biomateriab, 6, 97-101 (1985). S.J. Klebanoff, A. M. Waltersdorph, and H. Rosen, Methods in Enzymology, vol. 105, L. Packer (ed.), Academic Press, New York, 1984, pp. 399-403. S. J. Klebanoff, “Iodination of bacteria: A bactericidal mechanism,” J. Ex?. Med., 126, 1063-1076 (1967). B. Reitner and G . Harnulv, ”Lactoperoxidase antibactericidal system: Natural occurence and practical applications,” J. Food Protect., 47, 724732 (1984). L. A. Decker (ed.), Worthington Enzyme Manual, Worthington Biochemical Corporation, Freehold, NJ 1977, pp. 66-67. E. Beutler, 0. Duron, and 8.M. Kelly, ”Improved method for determination of blood glutathione,” J. Lab. Clin. Med., 61, 882-888 (1963). H. Jenzer, W. Jones, and H. Kohler, “On the molecular mechanism of lactoperoxidase-catalyzed H2 0 2metabolism and irreversible enzyme inactivation,” J. Bid. Chem., 261, 15550-15556 (1986). J. A. Badwey and M. L. Kamovsky, ”Active oxygen species and oxygen functions of phagocytic leucocytes,” Ann. Rev. Biochem., 49, 695-726 (1980). M. V. Torrielli and M. U. Dianzani, “Free radicals in inflammatory disease,” in Free Radicals in Molecular Biology, Aging and Disease, D. Armstrong et al. (eds.), Raven Press, New York, 1984, pp. 355-379. R. K. Root and M. S. Cohen, ”The microbiocidal mechanisms of human neutrophils and eosinophils,” Rev. Infect. Dis., 3, 565-598 (1981). C. Deby, “Superoxide production in white cells in inflammation,” Proc. Conf. Cells in Inflammation, Puris, 33-39 (1984). W. C. Barrette, Jr., J. M. Albrich, and J. K. Hurst, “Hypochlorous acid promoted loss of metabolic energy in Escherichia coli,” Infect. ImrnunoI., 55, 2518-2525 (1987). E. L. Thomas and T. M. Aune, ”Oxidation of Escherichia coli sulfhydryl components by the peroxidase-hydrogen peroxide-iodide antimicrobiol system,” Antimicr. Agents and Chemother., 13, 1006-1010 (1978). G. Czapski and A. Samuni, “The kinetics of complexed and free H 0 2 radical in the reaction on Ce4* and H202,”Isr. J. Chem., 7, 361-373 (1969). R. C. Jandl, J. Andre-Schwartz, L. Borges-DuBois, R. S. Kipnes, and B. J . McMurrich ”Termination of the respirtory burst in human netrophils,” J. Clin. Invest., 61, 1176-1185 (1978). 0.Stendahl, B.-I. Coble, C. Dahlgren, J. Hed, and L. Molin, ”Myeloperoxidase modulates the phagocytic activity of polymorfonuclear neutrophil leucocytes. Studies with cells from a myeloperoxidasedeficient patient,” I. Clin. Invest., 73, 366-373 (1984). R. A. Clark and S. Szot, ”Chemotactic factor inactivation by stimulated human neutrophils mediated by myeloperoxidase-catalyzed methionine oxidation,” 1. Immunol., 128, 1507-1513 (1982). L. M. Bjursten, P. Tengvall, L. E. Ericson, C. Gretzer, and I. Lundstrom, “The anti-inflammatory effect of titanium-peroxy gel on polymorfonuclear granulocytes,” submitted for publication.
Received October 21, 1988 Accepted September 6, 1989
~
~
~
P. Tengvall, E. G . Hornsten, H. Elwing, and I. Lundstrom Department of Physics and Measurement Technology, Linkoping lnsfitute of Technology, S-581 83 Linkiiping, Sweden
A stable titanium-peroxy-radical complex is formed when metallic titanium interacts with hydrogen peroxide. The radical appears as one component in an aqueous gel formed when excess peroxides have been (catalytically)decomposed. The interaction between titanium and hydrogen peroxide may be of importance also in vivo during an inflammatory response at the implant. We report in this paper on the bactericidal effects of the titanium gel in the lacto- and myeloperoxidase-halogen systems. Escherichia coli viable count was used to evaluate the bactericidal properties of the gel and of H202for comparison. The gel
had only small or no toxic properties at high dilutions. Higher concentrations of the gel had bactericidal properties similar to those of H202.The results indicate that at physiological pH, the decomposition products of the gel are titanium hydroxide (Ti(IV)(OH-)4)and hydrogen peroxide (H202).It was found that the gel probably oxidizes glutathione directly in contrast to H202,which needs a peroxidase to do so. A model for the interaction between titanium and hydrogen peroxide is suggested. Its consequences for the properties of titanium in vivo are also discussed.
INTRODUCTION
It has been observed that an oxide like layer grows on titanium and stainless steel implants in viv~.’-~ The oxide growth was largest in tissue with high metabolic activity. Model experiments in vitro have shown that the interaction between hydrogen peroxide and a titanium surface may be a source for this ~xidation.~” It also has been observed that titanium-peroxy and peroxide complexes are formed when metallic, spontaneously oxidized, titanium interacts with hydrogen peroxide. Furthermore, both the hydrogen peroxide and titanium complexes are catalytically decomposed at the titanium surface. An aqueous titanium gel is formed when excess peroxides have been c o n s ~ m e dSince . ~ ~ one important component of this gel may be the titanium-peroxy-radical, we call the gel a Ti-peroxy gel. The composition of the gel is suggested to be:5 Ti(IV)0,2-(OH-), - Ti(IV)O; (OH-), - Ti(IV)(OH-), - Ti(IV)O, - n - H,O The gel thus contains a superoxide or perhydroxyl radical, which may be formed in the gel and is stable for a long period of time.3,4The gel has oxidizJournal of Biomedical Materials Research, Vol. 24, 319-330 (1990) 0 1990 John Wiley & Sons, Inc. CCC 0021-9304/90/030319-120$04.00
TENGVALL ET AL.
320
ing properties with a redox potential larger than that of the Fe(II)/Fe(III)couple (0.77 V). The fraction of the different components in the formula above may vary from gel to gel. In a typical gel, the total Ti concentration is determined to be about 40 mM and oxidizing component(s) to be about 10 mM5. Since the gel is produced from metallic titanium and hydrogen peroxide in distilled water, it is very clean and does not contain any counter ions in contrast to gels produced from titanium chloride and ammonium hydroxide solution^.^,^ The formation of complexes between hydrogen peroxide and titanium ions from metallic titanium is of particular interest in connection with the known biocompatibility of titanium.&'' An inflammatory response is normally triggered when a foreign material is inserted into living tissue. Hydrogen peroxide and oxygen radicals are thereby formed. The interaction between an implant surface and hydrogen peroxide is therefore not only of theoretical interest; it may have a direct connection with the behavior of an implant. One speculation is that the interaction between a titanium implant and hydrogen peroxide may lead to a modification of the initial inflammatory response. Since the Ti-peroxy gel described above has oxidizing properties in itself, and furthermore, most probably degrades under the formation of H,O,, we found it of interest to investigate the properties of the gel in a situation where H,O, is known to play an important biological role. The present communication deals therefore with the bactericidal properties of the titanium-peroxy gel in two systems where H,O, is active, namely the myeloperoxidase (MPO) + chloride, and the lactoperoxidase (LPO) + iodide-systems. In the presence of sufficiently large concentrations of H,O,, these systems are known to have toxic effects on microorganism^.^^-'^
MATERIALS AND METHODS
Prior to use, all glassware, iodide, and chloride (from Merck) diluted in buffer, buffers, soy agar (Pharmacia), 0.35% agarose gel (Pharmacia), and distilled water were autoclaved for 30 min. Taurine (2-aminoethanesulfonic acid) (Sigma) was sterile filtered prior to use. The E . coli ATCC 12932 was incubated at 37°C overnight in lactose peptone broth, Oxoid CM 305, 42 g/L. Prior to use, the culture was washed twice in the buffer used in the actual experiment. The bacterial densities obtained during the experiments ranged from lo6 cells/mL to 10' cells/mL. The bacterial suspension was used in experiments within 30 min of washing. Incubation during the experiments was performed at 37°C in a water bath with agitation at 160 osc/min. Serial dilutions of incubated bacterial suspension with 50-pL samples in 4.5 mL in the buffer used were made in three steps. Three drops from each step were dropped on nutrient agar plates. Viability was assayed by viable count (colony forming units on nutrient agar; Oxoid CM 32, 28 g/L) after overnight incubation. Sodium-phosphate buffer 0.1 M pH 7 was used in all experiments
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
321
with MPO and 0.1 M sodium-acetate buffer pH 5 was used in the LPOexperiments. In some MPO-experiments, 0.67 M sodium sulphate was used to maintain tonicity when chloride was omitted. All experiments, except those in Figures 1and 2, were performed in duplicate or triplicate. The Ti-gel was prepared from 710-pm Ti-powder (Johnson Matthew) and 30% hydrogen peroxide (Merck) according to a procedure described el~ewhere.~ In brief 4 g of Ti-powder was incubated in 100 mL of 30% H,Oz in a water bath at room temperature. During 1 week, excess peroxide was decomposed and the solution formed a Ti-peroxy gel, with a pH of 4, at the same time as oxygen evolution ceased. When needed, the gel was diluted in deionized water. Myeloperoxidase (E.CI1.11,1.7)in lyophilized solid form (Calbiochem, lot 801094) was used without further purification. The enzyme was diluted in buffer and used within 45 min. The activity of the enzyme was determined with a method described by Klebanoff et a1.I' Ten microliters of the MPO dissolved in sodium phosphate buffer pH 7 was added to a 1-cm light path cuvette containing 3 mL of the fresh assay solution made from 26.9 mL H,O, 3 mL of 0.1M sodium phosphate buffer pH 7, 0.1 mL of 0.1 M H,O,
Figure 1. E. coli viability as function of iodide and hydrogen peroxide concentrations in the LPO-system. Bactericidal activity is detectable at less than 1 p M iodide. Minimal viability is achieved at about 50 p M H 2 0 2 concentration.
TENGVALL ET AL.
322
I
1.5-10-'
I
1.5*10-6
I
1.5-10-5
log (Ti-PEROXVGEL vhr]
1.5-10-4
1.5-10-3
[Ti [lV)O; lOY-l,M] IN ATYPICAL GE
Figure 2. Dependence of E . coli viability on the concentration of the Tiperoxy gel in the MPO-chloride-Ti peroxy gel system.
and 0.048 mL guaiacol (Sigma). The absorbance of the solution was determined at 470 nm during 3 min. The change in absorbance per minute was calculated from the initial rate of change. The tetraguaiacol product formed has an extinction coefficient of 26.6 mM-' cm-' at 470 nm. It is formed due to oxidation of quaiacol. The following simple formula was used to calculate the activity of MPO. units/mL = 45.1
X
AABS/min
The numerical factor was determined from the results in ref. (11)and the used volumes of the reagents. The activity was determined to be 40 mu/mL. Lactoperoxidase (E.C.1.11.1.7.) lyophilized powder (Sigma) was used without further purification. Prior to use, 0.1 mg of the enzyme was diluted in 1mL distilled water and used within 45 min. The activity was assayed in the following way. l4 A 2.5-mM sample of 4-aminoantipyrine (Sigma) was mixed with 810 mg phenol (Sigma) in 40 mL distilled water. Then, another 25 mg 4-aminoantipyrine was added and the mixture stored in dark. A 1.4mL volume of the phenol/aminoantipyrine solution was then mixed with 1.5 mL of 1.7 mM H,O,-solution in distilled water. A 0.1-mL sample of the
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
323
enzyme solution was added and the change in light absorption of the solution at 510 nm was recorded for 2 min, The activity was calculated according to Units/mL =
AABS/minute 6.58 x mg enzyme/mL reaction mixture
The numerical factor was obtained from ref. [14] and the used volumes of the reagents. The activity was determined to be 200 mu/mL. Glutathione (GSH) (Calbiochem) oxidation experiments were performed with 5.5-dithio-bis (2-nitrobenzoic acid) (DTNB) (Sigma) as a sulfhydryl reagent. 0.14 M sodium phosphate buffer pH 7 was used during these experiments. Except for DTNB, 300 p L of each additive, dissolved in the buffer were added into a 1-cm light path cuvette. The gel was diluted to 50% (v/v) in the buffer before addition. The samples were incubated for 30 min in dark at room temperature. Thereafter, 300 pL of the DTNB-solution was added and the formation of a colored sulfhydryl-DTNB product was measured within 30 s spectrophotometrically at 412 nrn.I5 The final sample volume was 1.8 mL and, when necessary, buffer was added to keep the volume constant. The final concentrations of the reagents are given in Figure 3a.
RESULTS
To learn about the chosen test systems, extensive experiments were performed with the LPO-H,O,-I--system. Optimal bactericidal action was sought in this system by varying the H 2 0 2and I- concentrations (see Fig. 1).The dilution series show that the E . coli viability was minimal at 50 pM H,02. Bactericidal action was observed in the LPO-Ti-peroxy gel-I--system as found from Table I. When LPO, I- or both were omitted, bactericidal effects were not observed either for H20, or for the Ti-peroxy gel. Furthermore, when C1- was used as the halide in the LPO-system neither H202nor the Ti-peroxy gel showed bactericidal effects. In the MPO-Ti peroxy gel or H20, system bactericidal effects were obtained with C1- as the chosen halide (see Table 11). Toxicity was not observed when myeloperoxidase, chloride or both were excluded. The bactericidal effect of the gel decreased when the gel was diluted (see Fig. 2). Different mechanical treatments of the gel or change of order of addition of the substances did not alter the bactericidal properties of the gel. Addition of catalase, an H,O,-scavanger, did not abolish the bactericidal effect of the gel (see Table I). Addition of taurine (2-aminoethane sulfonic acid) in the MPO-CI- system abolished the bactericidal activity completely both of the H,O, and the Tiperoxy gel (Table 11).
TENGVALL ET AL.
324 412 nrn
1.8
1 6 . 1.4
.
1.2
.
1.0 . 0.8 .
0.6 . 0.4
.
02 .
0
-
G s H G s H G s H G s H G s H G DTNB Ti-gel Ti-gel Lpo Ti-gel DTNB Cat IL p o L DTNB DTNB I-
DTNB
s H G s H G s H Ti-gel p o
ICat DTNB
33 3%
I-
DTNB
Figure 3a. Glutathione (GSH) oxidation in the Ti-peroxy gel or H20zperoxidase-halogen systems. The experiments show that GSH is oxidized by the Ti-peroxy gel but not by H202and that the peroxidase-halogen systems potentiate this oxidation. The concentrations used were [GSH] = 0.15 mM, [DTNB] = 0.15 mM, [Ti-peroxy gel, oxidizing equivalent] = 2.5 mM, [I-] = 25 pM, [H202]= 1 mM, [LPO] = 200 mu/mL and [cat] = 2 u/mL. The components were incubated for 30 min, whereafter DTNB was added and the change in absorbance measured within 30 s. The change in adsorbance is a measure of the amount of unoxidized sulfhydryls.
Figure 3a shows the results of glutathione oxidation experiments for several combinations of gel, H,02, LPO, and I-. In contrast to H,Oz, addition of Tiperoxy gel into a reduced glutathione sample caused chemical alteration of glutathione in such a way that the formation of the colored sulfhydrylDTNB product formed as a reaction product between reduced glutathione and DTNB decreased. Since the Ti-peroxy gel per se appeared to oxidize (or influence) glutathione directly the effect of incubation time in the gel on the number of unoxidized glutathione molecules was also studied (Fig. 3b). DISCUSSION
The results for H20, behavior in peroxidase systems is in accordance with earlier findings.'2,'6The results in Figure 1 support, e.g., recent observations of an inactivation of LPO at H,O, concentrations above 50 pM.16
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
325
1.6
1.2
0.8
Time (min) 0.4 0
I
I
I
20
40
60
Figure 3b. Glutathione oxidation versus incubation time in 40% v/v Tiperoxy gel in sodium phosphate buffer pH 7. The glutathione concentration was 0.25 mM.
It can be concluded that the Ti-peroxy gel and degradation products from it are not themselves deleterious to E. coli. The gel had at high concentrations, however, together with both LPO and MPO and a proper halide bactericidal properties similar to those of hydrogen peroxide. It was observed that the viscosity of the gel increases when it is chemically reduced. Diffusion limitations or enzyme immobilization may therefore explain why catalase, a known H,02 scavanger, did not abolish the bactericidal effects of the gel. H,O, is known to form hypochlorous acid, HOCl, in the MPO-C1- system17,18
H,O,
+ Cl- + H+ 3 HOCl + H,O
Taurine forms N-chloramines from HDCl which are not liposoluble but oxidize extracellular t h i ~ l s . ' ~The , ~ ' fact that taurine abolished the bactericidal properties of both H202and the gel (Table 11) indicates that (a) the bactericidal compounds in the peroxidase-halogen-H,O, or Ti peroxy gel system are formed outside the bacteria, (b) oxidation of thiol groups outside the cell is not enough to kill E. coli, (c) the bactericidal mediator most probably is a hypohalous acid. The hypohalous acid reacts with ammonium to form liposoluble ammonium chloride, which is a powerful oxidant and N-chloramines which are particularly reactive with thiols (see, e.g., ref [20]). Oxidation of E . coIi by
TENGVALL ET AL.
326
TABLE I Bactericidal Effect of LPO, H202, Ti-Peroxy Gel and Iodide Viable cell count Organisms per mL
Supplements None Ti-peroxy gel Agarose Ti-peroxy gel + LPO Agarose t LPO LPO + H 2 0 2+ ILPO + Ti-peroxy gel + ILPO + agarose + 1Agarose + ILPO + H202 LPO + Ti-peroxy gel + H202 LPO + agarose + H,O, LPO + Ti-peroxy gel + H,02 + ILPO + agarose + H,02 + ILPO + Ti-peroxy gel + I- + catalase LPO + agarose + I- + catalase LPO + H,O, + agarose + I- + catalase LPO + H,O, + Ti-peroxy gel + 1- + catalase
4 x 106 3 x 106 4 x 106 1 x 106 3 x 106 < 100 < 100 400 x lo6 7 x lo6 10 x 106 2 x 106 7 x 106 < 100 < 100
< 100 7 x 106
< 100 < 100
The final sample volume was 2 rnL. 0.1 M Sodium acetate buffer p H 5 was used. 0.5 mL of each supplement solution was added. When a supplement was omitted, the same volume buffer was added. The final concentrations of the supplements and order of addition was as follows: LPO 200 mu/mL, E . coli suspended in the buffer, H202 50 pM or the gel, eventually including catalase (2 pg/mL), and I- 1 pM. Incubation 3045 rnin at 37°C in a shaking water bath.
TABLE I1 Bactericidal Effect of MPO H20,, Ti-Peroxy Gel and Chloride ~
Viable count Organisms per mL
Supplements None MPO MPO + Ti-peroxy gel MPO + H202 MPO + H202+ Ti-peroxy gel H,02 + Ti-peroxy gel Chloride MPO + chloride MPO + chloride + Ti-peroxy gel MPO + chloride + Ti-peroxy gel + taurine MPO + chloride + H 2 0 2 MPO + chloride + H20, + taurine Chloride + H,O, + Ti-peroxy gel MPO + chloride + H20,+ Ti-peroxy gel
I
I
I
5x 1x 0.3 x 1x 0.3 X 0.3 x 1x 1x
107 107 107 107 16
lo7
107 107 < 100 0.1 x lo7 < 100 7 x 107 0.4 x 107 < 100
The final sample volume was 1 mL. Sodium phosphate buffer 0.1 M p H 7 was used. The final concentrations of the supplements land order of addition was as follows: E. coli suspended in buffer, chloride 0.1 M, sodium sulphate when chloride was omitted 0.15 M, taurine 10 mM, H 2 0 220 p M , Ti-perqxy gel 0.2 mL/sample and MPO 40 mu/mL. Incubation 30-45 min at 37°C in a shaking water bath.
BACTERlCIDAL PROPERTIES OF TITANIUM-PEROXY GEL
327
hypochlorous acid also gives rise to hydrolysis of cytosolic nucleotide phosphoanhydride bonds and loss of metabolic energy.21 It was observed that gluthatione was oxidized (or structurally altered) directly by the Ti-peroxy gel but not by H,O, (Fig. 3a). The Ti-peroxy gel has thus oxidizing properties differing from those of H,O,. It was found that the LPO-I- system potentiated the oxidation of glutathione by both H202and the Ti-gel, Tlus is in accordance with earlier results which have shown sulfhydryl oxidation in the lactoperoxidase-hydrogen peroxide-iodide system.22In order to check that the gel actually interacted with the glutathione, some experiments were made where glutathione was incubated in the gel for different lengths of time before the addition of the color reagent (DTNB) (Fig. 3b). The experiment shows that the longer the time of incubation of glutathione in the Ti-peroxy gel, the smaller amount of oxidizable compound there will be left for the sulfhydryl-DTNB reaction. The result also indicates that the decreasing sulfhydryl-DTNB formation is not caused by the gel-DTNB interaction. The Ti-peroxy complex which is a strong oxidant is thus most probably initially reduced by the biological environment to form Ti-peroxide, Ti(IV)O;-(OH-),. The Ti-peroxide appears to possess bactericidal properties in the presence of LPO or MPO and the proper halogen. Several alternative reaction pathways may explain this observation. Hydrogen peroxide incorporated in the gel may be released during the degradation of the gel. Another possibility is that LPO or MPO may act directly on the Ti-peroxide to form (oxidizing) halide compounds. It is also possible that the Ti-peroxide gel reacts with water molecules to form Ti-hydroxide, Ti(IV)(OH-),, and free or weakly bound H,02,
-
+ H,O Ti(IV)HO,(OH), Ti(IV)HO,(OH-), & Ti(IV)(OH-), + H,O,
Ti(IV)O;-(OH-),
Increasing pH shifts the reactions to the right. These reactions are suggested to take place at physiological pH. The H20, is now used as a substrate by the peroxidases. If the HO; anion leaves the Ti-coordination sphere, it may through proton uptake form H,02 at neutral or weakly acidic conditions. Other reactions are also possible with the HOT anion, which eventually lead to H,O, format i ~ n . In ’ ~summary, the two most probable ways for the formation of hydrogen peroxide from a degrading Ti-peroxy gel at weakly acidic or neutral pH are 1) HO, abstraction from the Ti-coordination sphere with subsequent proton uptake; 2) water uptake by the Ti-peroxide giving rise to free or weakly bound hydrogen peroxide. Figure 4 is a tentative summary of the possible reactions between hydrogen peroxide and (oxidized) metallic titanium. H,O, takes part in the oxidation of the metal surface; participate in the formation of Ti-peroxy radicals and other Ti-complexes. Both Ti(1V)- and Ti(II1) may interact with H,O, during this process. H,02 is also catalytically decomposed at the metal surface. If the reaction conditions are proper, a titanium peroxy gel may be formed when
328
TENGVALL ET AL.
Figure 4. Suggested reaction pathways during interaction between titanium and hydrogen peroxide: When Ti-peroxide (1) is formed, it may, at neutral or weakly acidic conditions, spontaneously decompose to Ti-hydroxide and hydrogen peroxide (2) (main pathway). The titanium complexes as well as hydrogen peroxide are catalytically decomposed at the oxidized surface. This may form a Ti(N)O;(OH-),-radical which in turn may complex with other Ti-adducts (chelation as end result) or decompose to Ti-peroxide (4) through an oxidation of the surroundings. The Ti-peroxy gel eventually formed is degraded through reduction to Ti-peroxide (4). Alternatively, the radical may be formed from Ti(1II) and O2 as the catalytic Hp02-decomposition leads to a local increase of the oxygen pressure (5). During the processes above, the oxide layer on the titanium surface grows.
the excess peroxide has been d e c ~ m p o s e d .Through ~-~ degradation of the Ti-peroxide part of the gel, the compounds will then act as a source of H,O,. The reactions in Figure 4 may also occur in vivo at an implant surface with H,O, supplied from a respiratory burst. The concentrations of H,O, and Ticomplexes will, however, be too small to cause a Ti-peroxy gel formation although complexed hydrated Ti-oxides may be found in the surroundings of an implant. The MPO-H,O,-C1-system is of large importance during an inflammatory response. 20,24,25 The hypochlorous acid produced in this system forms ammonium chloride and chloramines under physiological conditions.” These oxidize thiol containing residues both extra- and intracellularly and in the cell membranes. Some of the reactions lead to an enhancement of the inflam-
BACTERICIDAL PROPERTIES OF TITANIUM-PEROXY GEL
329
matory response,” whereas others inactivate inflammatory mediators.26The interaction between H,O, and titanium leads to a smaller concentration of H,O, and to Ti-compounds which slowly release (a small concentration of) H,O, close to the implant surface. This interaction will most probably have pronounced effects on the local inflammatory response at the implant. Another interesting observation is that the Ti-peroxy gel interacts with glutathione by itself whereas hydrogen peroxide does that only in presence of peroxidases. Apparently, the Ti-peroxy and/or Ti-peroxide interact directly with thiols (see Fig. 3a, b), which also may influence an inflammatory response. Although we have no independent evidence, it is most likely that the gel oxidizes the sulfhydryl groups of glutathione. A hypothesis is, therefore, that one of the positive properties of titanium as an implant material in addition to small or no systemic effects from the end products, there may be a down regulation of the initial inflammatory response due to the effects discussed above. Preliminary model studies of the interaction between the titanium gel and polymorphonuclear granulocytesz7as well as some observations on titanium implants in inflamed tissue support such a hypothesis. More information has, however, to be collected before the role of titanium implants during the inflammatory response has been fully elucidated. Dr. Lars-Magnus Bjursten, Department of Anatomy, University of Gothenburg, is gratefully acknowledged for stimulating discussions on the properties of the titaniumperoxy gel. The research on the biocompatibility of titanium is supported by grants from the National Swedish Board for Technical Development and the Institute for Applied Biotechnology, Gothenburg, Sweden.
References 1. J. E. Sundgren, P. Bodo, and I. Lundstrom, ”Auger electron spectro-
2. 3.
4.
5.
6. 7.
8.
scopic studies of the interface between human tissue and implants of titanium and stainless steel,” J. Colloid Interface Sci., 110, 9-20 (1986). J.E. Sundgren, P. Bodo, I. Lundstrom, A. Berggren, and S. Hellem, “Auger electron spectroscopic studies of stainless-steel implants,” 1. Biomed. Muter. Res., 19, 663-671 (1985). P. Tengvall, H. Elwing, L. Sjokvist, L.M. Bjursten, and I. Lundstrom, ”Interaction between hydrogen peroxide and titanium: A possible role in the biocompatibility of titanium,” Biomateriuls, 10, 118-120 (1989). P.Tengvall, I. Lundstrom, L. Sjokvist, H. Elwing, and L. M. Bjursten, ”Titanium-hydrogen peroxide interaction, Model studies of the influence of the inflammatory response on titanium implants,” Biomuferials, 10, 166-180 (1989). P. Tengvall, H. Elwing, and I. Lundstrom, ”Titanium gel made from metallic titanium and hydrogen peroxide,” J. Colloid Interface Sci., 130, 405-413 (1989). J. Ragai, “Trapped radicals in titania gels,” Nature, 325, 703-705 (1987). J. Muhlebach, K. Muller, and G. Schwarzenbach, “The peroxo complexes of titanium,” Inorg. Chem., 9, 2381-2390 (1970). P.I. Brbnemark, ”Introduction to osseointegration in tissue integrated prostheses in Osseointegration in Clinical Dentistry, P. I. Brhemark,
TENGVALL ET AL.
330
9. 10.
11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22.
23. 24. 25.
26. 27.
G. A. Zarb, and T. Albrektsson (eds.), Quentessence, Chicago, 1985, Chap. 4. P. Thomsen, L. M. Bjursten, and L. E. Ericson, “Implants in the abdominal wall of the rat,” Sand. J. Reconst. Surg., 20, 173-182 (1986). T. Albrektsson, H.-A. Hansson, and B. Ivarsson, “Interface analysis of titanium and zirconium bone implants,” Biomateriab, 6, 97-101 (1985). S.J. Klebanoff, A. M. Waltersdorph, and H. Rosen, Methods in Enzymology, vol. 105, L. Packer (ed.), Academic Press, New York, 1984, pp. 399-403. S. J. Klebanoff, “Iodination of bacteria: A bactericidal mechanism,” J. Ex?. Med., 126, 1063-1076 (1967). B. Reitner and G . Harnulv, ”Lactoperoxidase antibactericidal system: Natural occurence and practical applications,” J. Food Protect., 47, 724732 (1984). L. A. Decker (ed.), Worthington Enzyme Manual, Worthington Biochemical Corporation, Freehold, NJ 1977, pp. 66-67. E. Beutler, 0. Duron, and 8.M. Kelly, ”Improved method for determination of blood glutathione,” J. Lab. Clin. Med., 61, 882-888 (1963). H. Jenzer, W. Jones, and H. Kohler, “On the molecular mechanism of lactoperoxidase-catalyzed H2 0 2metabolism and irreversible enzyme inactivation,” J. Bid. Chem., 261, 15550-15556 (1986). J. A. Badwey and M. L. Kamovsky, ”Active oxygen species and oxygen functions of phagocytic leucocytes,” Ann. Rev. Biochem., 49, 695-726 (1980). M. V. Torrielli and M. U. Dianzani, “Free radicals in inflammatory disease,” in Free Radicals in Molecular Biology, Aging and Disease, D. Armstrong et al. (eds.), Raven Press, New York, 1984, pp. 355-379. R. K. Root and M. S. Cohen, ”The microbiocidal mechanisms of human neutrophils and eosinophils,” Rev. Infect. Dis., 3, 565-598 (1981). C. Deby, “Superoxide production in white cells in inflammation,” Proc. Conf. Cells in Inflammation, Puris, 33-39 (1984). W. C. Barrette, Jr., J. M. Albrich, and J. K. Hurst, “Hypochlorous acid promoted loss of metabolic energy in Escherichia coli,” Infect. ImrnunoI., 55, 2518-2525 (1987). E. L. Thomas and T. M. Aune, ”Oxidation of Escherichia coli sulfhydryl components by the peroxidase-hydrogen peroxide-iodide antimicrobiol system,” Antimicr. Agents and Chemother., 13, 1006-1010 (1978). G. Czapski and A. Samuni, “The kinetics of complexed and free H 0 2 radical in the reaction on Ce4* and H202,”Isr. J. Chem., 7, 361-373 (1969). R. C. Jandl, J. Andre-Schwartz, L. Borges-DuBois, R. S. Kipnes, and B. J . McMurrich ”Termination of the respirtory burst in human netrophils,” J. Clin. Invest., 61, 1176-1185 (1978). 0.Stendahl, B.-I. Coble, C. Dahlgren, J. Hed, and L. Molin, ”Myeloperoxidase modulates the phagocytic activity of polymorfonuclear neutrophil leucocytes. Studies with cells from a myeloperoxidasedeficient patient,” I. Clin. Invest., 73, 366-373 (1984). R. A. Clark and S. Szot, ”Chemotactic factor inactivation by stimulated human neutrophils mediated by myeloperoxidase-catalyzed methionine oxidation,” 1. Immunol., 128, 1507-1513 (1982). L. M. Bjursten, P. Tengvall, L. E. Ericson, C. Gretzer, and I. Lundstrom, “The anti-inflammatory effect of titanium-peroxy gel on polymorfonuclear granulocytes,” submitted for publication.
Received October 21, 1988 Accepted September 6, 1989
