Journal of Hospital Infection (2009) 71, 149e156

Available online at www.sciencedirect.com

www.elsevierhealth.com/journals/jhin

Photocatalytic degradation of prions using thephoto-Fenton reagent

I. Paspaltsis a, C. Berberidou b, I. Poulios b,*, T. Sklaviadis a,c

a Prion Disease Research Group, Laboratory of Pharmacology, School of Pharmacy,Aristotle University of Thessaloniki, Thessaloniki, Greeceb Laboratory of Physical Chemistry, Department of Chemistry,Aristotle University of Thessaloniki, Thessaloniki, Greecec Centre for Research and TechnologyeHellas, Institute of Agrobiotechnology, Thessaloniki, Greece

Received 19 May 2008; accepted 19 September 2008Available online 17 November 2008

KEYWORDSAdvanced oxidationprocess (AOP);Photocatalysis; Photo-Fenton reagent; Priondegradation;Transmissiblespongiformencephalopathies(TSEs)

* Corresponding author. Address: Labistry, Department of Chemistry, Aristoniki, 54124, Thessaloniki, Greece. Tel.10 997784.

E-mail address: poulios@chem.aut

0195-6701/$ - see front matter ª 200doi:10.1016/j.jhin.2008.09.015

Summary Prions are proteinaceous infectious agents postulated to be thecausative agents of a group of fatal neurodegenerative diseases knownas transmissible spongiform encephalopathies (TSEs). A known iatrogenictransmission route of TSEs to humans occurs via prion-contaminatedsurgical instruments or biological materials. Prions, unlike most commonpathogens, exhibit an extraordinary resistance to conventional decontami-nation procedures. We have recently demonstrated that the application ofTiO2-based heterogeneous photocatalytic oxidation is able to significantlyreduce prion infectivity. The present study investigates the potential of a ho-mogeneous photocatalytic method, based on the photo-Fenton reagent, todegrade prion proteins. We show that the photo-Fenton reagent efficientlydegrades not only recombinant prion proteins, but also the total proteinamount from brain preparations of naturally or experimentally infected spe-cies and PrPSc (PrP scrapie) contained in sheep scrapie brain homogenates.ª 2008 The Hospital Infection Society. Published by Elsevier Ltd. All rightsreserved.

oratory of Physical Chem-tle University of Thessalo-: þ23 10 997785; fax: þ23

h.gr

8 The Hospital Infection Socie

Introduction

Prion diseases or transmissible spongiform enceph-alopathies (TSEs) comprise a group of fatal de-generative diseases of the central nervous systemaffecting both humans and animals.1 Human TSEsare categorised into acquired, sporadic and

ty. Published by Elsevier Ltd. All rights reserved.

150 I. Paspaltsis et al.

genetic disorders. Nosocomial transmission hasbeen observed after medical treatments withprion-infected surgical tools or by the use of in-fected biological materials from affectedindividuals.

Prions contained in liquid waste or on contam-inated surgical instruments and devices cannot befully inactivated by procedures commonly used toinactivate viruses and bacteria.2 The World HealthOrganization guideline about the use of single-usedisposable surgical items in patients with con-firmed or suspected TSEs can only be seen asa transient solution.3 First, it is cost-ineffective;second, asymptomatic carriers cannot be excludedas potential sources of the infectious agent. This isof importance, considering the long incubationtime of the disease and the variety of tissues whichmay harbour infectivity.4 Although prion diseasesin humans are rare, the risk of an iatrogenic trans-mission cannot be tolerated by any health author-ity. Thus, prion inactivation or decontamination ofsurgical tools remains a prominent topic of priondisease research.

Previous studies demonstrated that prion in-fectivity persists over years even upon exposure toenvironmental conditions.5 Effective treatment ofliquid waste originating from facilities that handleprion-infected or possibly infected materials suchas biochemical laboratories, hospitals, abattoirsand farms limits the bioavailability of the pathogenand significantly decreases the risk of interspeciestransmission.

Recently, our group demonstrated that theapplication of a novel oxidation process, in partic-ular the titanium dioxide (TiO2)-mediated photo-catalytic oxidation, can significantly reduce prioninfectivity.6 Heterogeneous photocatalysis belongsto the advanced oxidation processes (AOPs),a group of physicochemical oxidation methodsused mainly for the decontamination of waterand air polluted with organic compounds. The dif-ferent AOPs share a basic characteristic, the gen-eration and use of powerful transitory species,principally the free hydroxyl radicals (OH$), thestrongest oxidative oxygen-based species in na-ture.7 Together with other highly oxidant species(e.g. peroxide radicals) they easily attack organicpollutants non-selectively, resulting in their com-plete mineralisation. Moreover, photocatalytic ox-idation mainly mediated by TiO2, in the presenceof artificial or solar light, has been reported in sev-eral cases to successfully inactivate viruses, andpathogenic bacteria both in air and liquidmedia.8,9

The photocatalytic degradation of proteins in-volves several chemical and photocatalytic stages

and a great number of intermediates. Previousstudies have proposed mechanisms of oxidativecleavage mediated by OH$, which are able toattack the main and the side chains of proteins,producing molecules with lower molecular weightand finally CO2.

10,11

The application of Fenton or Fenton-like re-actions to protein degradation and particularlyprions has been reported.12,13 We aimed to deter-mine whether a modification of the classical Fen-ton system, the so-called ‘photo-Fenton reagent’(Fe3þ/H2O2/UV-A, visible light), which also belongsto the AOPs, has the potential to oxidise prions.

Methods

Substrates for photo-Fenton treatment

Bovine serum albumin (BSA) was obtained fromSigmaeAldrich (Athens, Greece). Recombinantproteins, i.e. 10�His (10 histidine)-tagged bovineand ovine prion proteins (PrPs) were expressed inE. coli BL21 (DE3) cells. Denatured proteins werepurified from inclusion bodies by affinity chroma-tography. Before photocatalytic treatment, ali-quots of each protein were precipitated bymixing with excess of methanol. The pellets wereresuspended by vortexing in the photo-Fentonmixture described below.

Prion-infected brain tissue from different spe-cies including human (sporadic CreutzfeldteJakobdisease, sCJD), cow (bovine spongiform encepha-lopathy, BSE), sheep (scrapie), hamster (263Kscrapie) and mouse (79A scrapie and 301 V BSE)was also used in the study. Brain homogenates(10% w/v) were prepared in phosphate-bufferedsaline (pH 7.4), 0.5% w/v sodium deoxycholate,0.5% w/v Igepal (a non-ionic detergent) and5 mM phenylmethanesulphonylfluoride. A 20%w/v sheep scrapie brain homogenate was similarlyprepared. The initial total protein amount of therecombinant proteins and of the different brainhomogenates was estimated by the Bradfordassay.14

Photocatalytic experiments

Analytical grade FeCl3$6H2O and H2O2 were pur-chased from Alfa-Aesar (Karlsruhe, Germany) andMerck Chemicals (Athens, Greece) respectively.Experiments were performed at room temperaturein uncapped disposable 1.5 mL plastic tubes andthe reaction mixture was gently stirred during thephotocatalytic treatment. Irradiation was carriedout using two parallel 8 W blacklight blue

Photo-Fenton reagent for prion protein degradation 151

fluorescent tubes, mounted in standard 8 W fluores-cent tube holders (TLD 8 W/08, Phillips, Athens,Greece), 5 cm above the irradiated solutionsurface. The light intensity at this position, in theUV-A region of 340e400 nm, was measured usinga Photometer/Radiometer PMA 2100 (Solar LightCo., Glenside, Pensilvania) equipped with a UV-Adetector and found to be 6.1� 0.3 mW/cm2.

The final reaction volume for each tube was50 mL. BSA or recombinant PrPs were photocata-lytically treated for 150 min in the presence ofUV-A with 12 mg/mL Fe3þ and 500 mg/mL H2O2 atpH 3.0 adjusted with 1 N HClO4. After the treat-ment the equivalent of 3 mg of the initial proteincontent was loaded per lane on a 12% sodium do-decyl sulphate (SDS) acrylamide gel and stainedwith Coomassie Brilliant Blue R250.15 Samples ofwhole-brain homogenates were diluted in distilledwater and treated with 500 mg/mL Fe3þ and5000 mg/mL H2O2 at pH 3.0 in the presence ofUV-A. Samples were analysed on a 12% SDS acryl-amide gel. Total proteins were stained with silvernitrate, whereas PrPs were detected by westernblot. Control reactions were handled similarly,but in combinations in which at least one ofthe photo-Fenton components (Fe3þ/H2O2/UV-A,visible) was absent.

Semiquantification of PrP immunodetection

In order to assess the sensitivity limit of PrPdetection in this study, serial dilutions of thesheep scrapie homogenate were subjected toimmunobloting as described below. The initialload of 1800 mg was serially diluted in 2-fold stepsdown to 0.88 mg brain equivalents (Figure 1A).

Electrophoresis and immunoblotting

SDSepolyacrylamide gel electrophoresis (PAGE)-analysed samples of BSA and recombinant prionproteins were stained with Coomassie BrilliantBlue. Gels were scanned and analysed densito-metrically as described below.

Degradation of proteins contained in brainhomogenates was visualised by the more sensitivesilver staining. PrP degradation was monitored bywestern blotting. Sheep scrapie brain homogen-ates were analysed by SDSePAGE and proteinswere transferred on to polyvinylidene fluoride(PVDF) membranes. Blots were probed with theanti-PrP monoclonal antibody 8G8 (recognisingamino acids 111e118 of the human PrP) diluted1:3000 from 1 mg/mL stock in blocking buffer (PBScontaining 0.1% v/v Tween 20 and 5% w/vskimmed milk). As secondary antibody, a rabbit

antimouseehorseradish peroxidase conjugate(Pierce, Athens, Greece) diluted 1:5000 in block-ing buffer was used. Prion proteins were visualisedboth with the enhanced chemiluminescence (ECL,Pierce) and with the more sensitive SuperSignalWest Femto Substrate (Pierce), able to detectfemtograms of protein, according to the manufac-turer’s instructions.

Densitometry

Densitometric measurements were performedwith the ImageJ software (version 1.37v, availableat http://rsb.info.nih.gov/ij/) through the analy-sis of multiple film exposures to ensure that com-parisons were made within the linear range of thefilm. As relative band density (R) we define:

R¼ aðt¼xÞ

aðt¼0Þ

where a is the absolute value of the area under thecurve created by the protein band at the investi-gated time point.

Results

Photo-Fenton treatment of pure proteins

Initially, we tested the effect of the photo-Fentonreactivity on BSA, bovine and ovine PrP at an initialprotein concentration of 150 mg/mL. After 150 minof photocatalytic treatment with 12 mg/mL Fe3þ

and 500 mg/mL H2O2, proteins were undetectableon Coomassie Brilliant Blue-stained SDSePAGEgels (Figure 2A, C). Control experiments underthe same conditions but in the absence of thephoto-Fenton reagent showed no or partial degra-dation, as expected, in the case of Fe3þ/H2O2 inthe absence of light (Figure 2B).

Photo-Fenton treatment of infectedbrain homogenates

Protein oxidation, mediated by the Fe3þ/H2O2/UV-A system, was also confirmed in experiments inwhich the whole protein content of 1% w/v brainhomogenates from different species was exposedto the oxidative environment. The initial total pro-tein amount of the different 10% w/v homogenatesranged between 1.8 and 2.1 mg/mL.

Figure 2D shows that after 2 h of UV-A illumi-nation in the presence of 500 mg/mL Fe3þ and5000 mg/mL H2O2 (pH 3.0), complete degradationof the total protein load was accomplished, as

1800

.090

0.0

450.

022

5.0

112.

5

56.3

28.1

14.0

7.0

3.5

1.8

0.9

Rel

ativ

e B

and

Den

sity

time (min)0.0

0.0

0.2

0.4

0.6

0.8

1.0

0.5 50 100 150 200

t: min

ECL Substrate West Femto Substrate

ECL Substrate West Femto Substrate

ECL Substrate West Femto Substrate

0 0.5 1 2 3 4 5 0 0.5 1 2 3 4 5

t: min 0 15 30 60 90 120 180 0 15 30 60 90 120 180

t: min 0 15 60 120 180 240 300 0 15 60 120 180 240 300

(A) (B)

(C) (D)

(E) (F)

(G) (H)

Figure 1 (A) 2-fold serial dilutions of the untreated 20% w/v sheep scrapie homogenate. The amount of brain tissueloaded in each lane (in mg) is indicated on the top. Blot was developed with the West Femto substrate. (B) Degradationof PrP in relation to illumination time after densitometric analysis of proteins detected with the ECL substrate in (-) 1,(�) 5 and (:) 10% w/v sheep scrapie homogenates. (C-H) Western blots of photo-Fenton (500 mg mL�1 Feþ3, 5000 mg mL�1

H2O2) treated1 (C, D), 5 (E, F) and10% w/v (G, H) sheep scrapie homogenates for the indicated time points. Theuntreatedsample (t: 0 min) contains 0.35, 0.9 and 1.8 mg brain equivalent respectively. Arrow: 32.5 kDa molecular mass standard.

152 I. Paspaltsis et al.

160

160

140

140

120

120

100

100

80

80

60

60

40

40

20

200

0Illumination time (min)

Cpr

otei

n (µ

g m

L-1

)

BSA

bPrP

oPrP

BSA

bPrP

oPrP

t: min 0 20 40 60 90 120 150UV-A

H2O2

Fe3+ –

––

––

+ –

–+–

+

+

–

++

–

+

+

+ + + + ++ ––––––

CJD BSE Sheepscrapie

Mousescrapie

MouseBSE

Hamsterscrapie

(A) (B)

(C) (D)

Figure 2 Photo-Fenton treatment with 12 mg /mLFe3þ, 500 mg/mL H2O2 of bovine serum albumin (BSA, -), bovineprion protein (bPrP, C) and ovine prion protein (oPrP, :). (A) Degradation in relation to irradiation time by gel elec-trophoresis and Coomassie staining. (B) Control samples were treated as indicated for 150 min. (C) Decrease of pro-tein concentration in relation to the illumination time after densitometric analysis of the gels in (A). (D) Treatment of1% w/v brain homogenates including sporadic CreutzfeldteJakob disease (sCJD) and bovine spongiform encephalop-athy (BSE)TSE for 120 min in the presence (þ) or absence (e) of photo-Fenton. Proteins were visualised by silverstaining.

Photo-Fenton reagent for prion protein degradation 153

demonstrated by silver-stained SDSePAGE gels.Similarly to the recombinant protein control re-actions, in which at least one component ofthe photo-Fenton system was absent, thisshowed no or minor protein degradation (datanot shown).

Serial dilution immunoblotting was performedto estimate the minimum tissue amount requiredfor PrP detection. A total of 7 mg of brain tissuewas found to contain the minimum amount of PrPdetectable by the applied protocol (Figure 1A).

In order to monitor the degradation of PrPSc, 1%w/v sheep scrapie homogenate was treated with500 mg/mL Fe3þ and 5000 mg/mL H2O2 (pH 3.0).Samples at different time points were subjectedto PrP detection by immunoblotting. As demon-strated in Figure 2D, within 2 min of treatmentPrP is undetectable even when visualised withthe West Femto substrate. Sheep scrapie homoge-nates (5% and 10% w/v) were treated under the ex-act same conditions. After 1 or 4 h of illuminationrespectively, PrP was no longer detectable with

the West Femto substrate (Figure 1F, H), whereaswestern blots of control reactions showed no orminor decrease of the PrP signal under the sametreatment conditions (Figure 3).

Discussion

Numerous studies have dealt with the oxidation ofmodel compounds and real effluents by the Fentonreagent, a mixture of iron (II) salts and H2O2.

16 Inthis technology, the ferrous and/or ferric cationdecomposes H2O2, to generate powerful oxidisingagents, able to degrade organic and inorganicsubstances. The process involves a large numberof reactions including redox, complexation, pre-cipitation, equilibrium reactions, and so on. Fen-ton reagent is an attractive oxidative system,which produces OH$ radicals (Equation 1) ina very simple way for waste-water and soil treat-ment, since iron is an abundant, non-toxic elementand H2O2 is easy to handle and environmentallysafe.

Fe3+

H2O2

UV-A

Fe3+

H2O2

UV-A

–

–– – –

– –

––

––

+

+ + ++

+ + +

+

+

–

–– – – – –

––

––+

+ + +

+

+ +

+ +

+

ECL Substrate

ECL Substrate

West Femto Substrate

West Femto Substrate

–

––

–

– –– – –

–

++

+++

+++

–

––

–

– –– – –

–

++

+++

+++

A B

DC

Figure 3 Control samples of 5% w/v (A, B) and 10% w/v (C, D) sheep scrapie homogenates treated for 3 and 5 hrespectively under the indicated conditions. The Fe3þ and H2O2 concentrations, when present, were 500 and5000 mg/mL respectively. Arrow: 32.5 kDa molecular mass standard. ECL, enhanced chemiluminescence.

154 I. Paspaltsis et al.

Fe2þ þH2O2/Fe3þ þOH� þOH$ ð1Þ

Fe3þ þH2Oþ hvð< 450 nmÞ/Fe2þ þHþ þOH$ ð2Þ

The reaction can be greatly enhanced by UV/vis-ible light (artificial or natural), producing addi-tional OH$ and leading to the regeneration of thecatalyst [Equation (2), photo-Fenton reaction].17

This interconversion of ferrous and ferric ionsmay continuously catalyse the generation of OH$,thus eliminating the need for further addition offerrous salts. These reactions are the primaryforces of the photochemical self-cleaning of atmo-spheric and aquatic environment.18

In the present study the use of the Fe3þ/H2O2

system, as previously mentioned, leads accordingto Equations (1) and (2) to the production ofOH$, which attack proteins and other organic

molecules such as nucleic acids, lipids and carbo-hydrates (data not shown) present in the brain ho-mogenate preparations.

Figure 2 demonstrates a significantly fasterdegradation rate of BSA compared with bovineor ovine PrP under the same experimental condi-tions. This result could be attributed to the na-ture of the three different protein molecules.BSA is completely soluble in aqueous media incontrast to the hydrophobic recombinant prionproteins which tend to form aggregates. Thus,the main chain and the side chains of BSA mole-cules are more likely to be approached and at-tacked by OH$ compared with the hydrophobicsites of the recombinant PrP molecules, whichlimit the interaction with the oxidising species.It must be noted that, whereas recombinant prionproteins are not glycosylated, as PrPSc they form

Photo-Fenton reagent for prion protein degradation 155

aggregates and are still a good indicator for priondegradability.

Silver-stained gels in Figure 2D suggest that thephotocatalytic treatment causes not only the elim-ination of PrP, but of the entire protein load. Thefact that OH$ non-specifically oxidises and de-stroys organic molecules present in a tissue ho-mogenate, suggests that it would cause theindistinguishable degradation of any kind of or-ganic component that might be involved in TSEpathogenesis. The extent of degradation dependsonly on the ratio of the initial organic load to theamount of the photo-Fenton reagent.

Photo-Fenton-mediated homogeneous photoca-talytic oxidation of PrP, present in sheep scrapiebrain homogenates, has been investigated (Fig-ure 1). Treatment of 1% w/v scrapie-infected braintissue with 500 mg/mL Fe3þ and 5000 mg/mL H2O2

(pH 3.0) eliminates PrP immumoreactivity in<2 min (Figure 1D), reducing the initial PrP con-tent at least 50-fold. Even in the case of the 10%w/v homogenate, PrP was efficiently degraded un-der the same conditions within 5 h of treatment. Inthis case, the reduction of initial PrP content ex-ceeded 250-fold (Figure 1H). Control reactions(Figure 3) verify that efficient degradation of PrPoccurs only when all components of the photo-Fen-ton system are present. Figure 3B and D shows theenhanced sensitivity of the West Femto substrate.It must be stressed that the amount of PrP and oforganic residues adsorbed on contaminated surgi-cal instruments, or even in liquid waste, is remark-ably lower than in a 10% w/v homogenate. Thisfact provides indications for even shorter treat-ment times in scaled-up applications.

Laboratory-scale photocatalytic treatment ofsheep scrapie homogenates is shown to be effec-tive, concerning prion degradation, as detected byimmunoblotting. Nonetheless, further research isrequired for the development and exploitation ofphoto-Fenton in a possible application, e.g. forroutine decontamination of surgical instruments orbiochemical waste. Certain advantages of thisphotocatalytic process render it a promising toolfor the future development of a mature applica-tion. Mild operational conditions such as roomtemperature, atmospheric pressure, user and en-vironmentally safe reagents are required. The costof equipment, reagents and energy is low, suitablefor routine use in healthcare facilities. It also hasthe potential to be successfully scaled up. Asshown in studies with other types of organicpollutants, the results obtained after scaling upare in good agreement with those in laboratoryscale.19 Moreover, the OH$ radicals can easily at-tack soluble, insoluble or even immobilised organic

targets.20 Some of the OH$ radicals are generated,according to Equation (1), in the absence of illumi-nation. The contribution of the illumination to theprocess is, mainly, the regeneration of the catalyst[Fe2þ, Equation (2)]. Thus, direct illumination ofcontaminated surfaces is not necessary to achievedegradation of the absorbed organic residues.

In conclusion, the current study demonstratesthe potential of the photo-Fenton reagent toeffectively degrade prions, through the oxidativeattack of powerful transitory species. Indeed,optimisation and validation of protocols is requiredbefore this process could evolve to a matureapplication, along with more sensitive detectionmethods for the routine assessment of instrumentcleaning.21 However, the effectiveness and thesimplicity of the technique, the mild operationalconditions and the low cost of the reagents andequipment, render it a promising tool for prioninactivation.

Conflict of interest statementNone declared.

Funding sourcesThis work was supported by the NeuroPrion-Network of Excellence, FOOD-CT-2004-506579.

References

1. Prusiner SB. Prions. Proc Natl Acad Sci USA 1998;95:13363e13383.

2. Taylor DM. Inactivation of prions by physical and chemicalmeans. J Hosp Infect 1999;43(Suppl):S69eS76.

3. Anonymous. WHO infection control guidelines for transmis-sible spongiform encephalopathies: report of a WHO con-sultation. WHO/CDS/CSR/APH/2000.3. Geneva: WorldHealth Organization; 1999.

4. Wadsworth JD, Joiner S, Hill AF, et al. Tissue distribution ofprotease resistant prion protein in variant CreutzfeldteJakob disease using a highly sensitive immunoblottingassay. Lancet 2001;358:171e180.

5. Brown P, Gajdusek DC. Survival of scrapie virus after 3years’ interment. Lancet 1991;337:269e270.

6. Paspaltsis I, Kotta K, Lagoudaki R, Grigoriadis N, Poulios I,Sklaviadis T. Titanium dioxide photocatalytic inactivationof prions. J Gen Virol 2006;87:3125e3130.

7. Litter M. Introduction to photochemical advanced oxidationprocesses for water treatment. Environ Chem 2005;2:325e366.

8. Watts RJ, Kong S, Orr MP, Miller GC, Henry BE. Photocatalyticinactivation of coliform bacteria and viruses in secondarywastewater effluent. Water Res 1995;29:95e100.

9. Maness PC, Smolinski S, Blake DM, Huang Z, Wolfrum EJ,Jacoby WA. Bactericidal activity of photocatalytic TiO2 re-action: toward an understanding of its killing mechanism.Appl Environ Microbiol 1999;65:4094e4098.

10. Muszkat L, Feigelson L, Bir L, Muszkat KA. Titanium dioxidephotocatalyzed oxidation of proteins in biocontaminatedwaters. J Photochem Photobiol B 2001;60:32e36.

156 I. Paspaltsis et al.

11. Xu G, Chance MR. Hydroxyl radical-mediated modificationof proteins as probes for structural proteomics. Chem Rev2007;107:3514e3543.

12. Solassol J, Pastore M, Crozet C, Perrier V, Lehmann S. Anovel copperehydrogen peroxide formulation for prion de-contamination. J Infect Dis 2006;194:865e869.

13. Suyama K, Yoshioka M, Akagawa M, et al. Assessment ofprion inactivation by Fenton reaction using protein misfold-ing cyclic amplification and bioassay. Biosci Biotechnol Bio-chem 2007;71:2069e2071.

14. Bradford MM. A rapid and sensitive method for the quantita-tion of microgram quantities of protein utilizing the principleof protein-dye binding. Analyt Biochem 1976;72:248e254.

15. Laemmli UK. Cleavage of structural proteins during the as-sembly of the head of bacteriophage T4. Nature 1970;227:680e685.

16. Neyens E, Baeyens J. A review of classic Fenton’s peroxida-tion as an advanced oxidation technique. J Hazard Mater2003;98:33e50.

17. Oliveros E, Legrini O, Hohl M, Mueller T, Braun A. Industrialwastewater treatment: large scale development of a light-enhanced Fenton reaction. Chem Eng Process 1997;36:397e405.

18. Faust BC, Zepp RG. Photochemistry of aqueous iron (III) poly-carboxylate complexes: roles in the chemistry of atmosphericand surface waters. Environ Sci Technol 1993;27:2517e2522.

19. Garcia-Montano J, Perez-Estrada L, Oller I, Maldonado MI,Torrades F, Peral J. Pilot plant scale reactive dyes degrada-tion by solar photo-Fenton and biological processes. J Pho-tochem Photobiol A 2008;195:205e214.

20. Suketa N, Sawase T, Kitaura H, et al. An antibacterialsurface on dental implants, based on the photocatalyticbactericidal effect. Clin Implant Dent Relat Res 2005;7:105e111.

21. Lipscomb IP, Sihota AK, Keevil CW. Comparison betweenvisual analysis and microscope assessment of surgicalinstrument cleanliness from sterile service departments.J Hosp Infect 2008;68:52e58.