Hadzieha et al. Iron in neurodegeneration Neuropathology and Applied Neurobiology This article is protected by copyright. All rights reserved. 1 Iron metabolism and the role of iron in neurodegenerative disorders 1 -review article- Revised version Maya Hadzhieva 1 , Elmar Kirches 1,2 and Christian Mawrin 1,2# 1 Department of Neuropathology and 2 Center for Behavioural Brain Sciences (CBBS), Otto-von-Guericke-University Magdeburg, Germany; Running title: Iron in neurodegeneration Key words: neurodegeneration, ROS, iron, mitoferrin, TfR, DMT1, ferritin, frataxin, IscU # Correspondence: Christian Mawrin, MD Department of Neuropathology Otto-von-Guericke University Leipziger Strasse 44 D-39120 Magdeburg Tel: +49 391 67 158282 Fax. +49 391 6713300 e-mail: [email protected]This article has been accepted for publication and undergone full peer review but has not been through the copyediting, typesetting, pagination and proofreading process, which may lead to differences between this version and the Version of Record. Please cite this article as doi: 10.1111/nan.12096 Accepted Article
Transcript
Regulatory Toxicology and Pharmacology 67 (2013) 456–467
Stephen Wessels a,1, Hanne Ingmer b,⇑a DHI, Agern Allé 5, DK-2970 Hørsholm, Denmarkb Department of Veterinary Disease Biology, Faculty of Health and Medical Sciences, University of Copenhagen, Stigbøjlen 4, DK-1870 Frederiksberg C, Denmark
a r t i c l e i n f o
Article history:Received 14 February 2013Available online 27 September 2013
Keywords:DisinfectantBiocideMode of actionAntimicrobial resistanceQuaternary ammoniumPeracetic acidPHMBEuropean regulationEuropean directive
a b s t r a c t
This review deals with three categories of active substances for disinfectant products, their modes ofaction (MOA), and how MOA can help predict propensity for resistance in microorganisms. Within theEuropean Union applications for approval of disinfectants of all kinds must be submitted in a few years,and documentation on MOA and resistance must be part of those applications. Peracetic acid is an unspe-cific, pervasive oxidizer of C–C double bonds and reduced atoms. This MOA would imply poor chance fordevelopment of resistance in microorganisms, as borne out by the absence of such reports in the litera-ture. The quaternary ammonium compounds (QAC’s) are much more specific in their antimicrobial mech-anism. Even very low concentrations cause damage to the cytoplasmic membrane due to perturbation ofthe bilayers by the molecules’ alkyl chains. Development of microbial resistance to QAC’s, as well ascross-resistance to antibiotics, are particularly well documented. The polymer PHMB is antimicrobialbecause it disturbs the cell membrane’s bilayer by interacting with it along the surface of the membrane.Resistance to the polymer appears not to develop despite many years of use in many fields. However,PHMB’s toxicity to humans upon inhalation dictates great caution when deploying the substance.
� 2013 Elsevier Inc. All rights reserved.
1. Introduction propensity for either resistance or tolerance. The literature of the
This review relates to disinfectants, to how they kill or inhibitunwanted microorganisms, and to the resistance that may ariseas a consequence of use of disinfectants. In particular, this articlereviews the scientific documentation that in the not-too distant fu-ture must be part of an application for approval of any disinfectantin any European Union (EU) member state (European Council &European Parliament, 1998; European Parliament & EuropeanCouncil, 2012). The past decades there have been legions of litera-ture on the modes of action (MOA) of, and resistance to, disinfec-tants, and there has been rising scientific debate on thecontribution of disinfectants to the increased frequency of antibi-otic resistant microorganisms (White and McDermott, 2001;Maillard, 2002; Davies and Davies, 2010; Russell, 2003b;McDonnell and Russell, 1999; Bridier et al., 2011). Also the termresistance is being debated. In the present review resistance de-scribes a situation where bacterial cells are not killed or inhibitedby a concentration of antimicrobial substance that acts upon themajority of cells in that culture (European Commission, 2009).Thus, in this review, resistance is defined as a greater than 4-foldincrease in the minimal inhibitory concentration (MIC), andtolerance covers increases in susceptibility less than 4-fold. Thisreview will illustrate how knowledge of MOA can help predict
past three decades is summed up on the MOA’s of peracetic acid(PAA), of the quaternary ammonium compounds (QAC’s), and ofpoly(hexa methylenebiguanide) hydrochloride (PHMB). The resultis an overview of the current understanding of the details ofthese MOA’s that may be incorporated in applications for approvalof disinfectant products containing these three active substances,as described in the next section.
1.1. Regulation of disinfectants by EU law
Disinfectant products are chemical mixtures that eliminatemany or all undesirable microorganisms, except bacterial spores,on inanimate objects (Rutala et al., 2008). Disinfectants are com-monly applied to biotic or abiotic surfaces such as directly on theskin, in bathrooms, kitchens, or in production facilities, but mayalso be added to for example drinking water or swimming poolwater. In the EU, disinfectants are a subset of biocidal products(European Parliament & European Council, 2012). Other biocidalproducts include rat poison, mosquito repellent, antifouling paintsfor boats and slimicides, i.e., products not already covered by other,existing legislation.
When the EU biocides law is fully implemented, only specifi-cally approved active substances will be allowed on the market,and only specifically approved products will be allowed (EuropeanCouncil & European Parliament, 1998; European Parliament &European Council, 2012). Active substances are evaluated and ap-proved on behalf of the whole EU, whereas each product is
S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467 457
evaluated and approved by the authority in the member statewhere the product is to be marketed. Product applications can onlybe submitted after their active substances have been approved. Thedossiers for the disinfectant active substances were submitted in2007, and as yet, only a few of these have been evaluated and ap-proved. The active substance dossiers were compiled by companiesor groups of companies with interest in the substances in question.These companies and groups will own the approvals of, and rightsto, their active substances if and when approval is granted. New ac-tive substance dossiers may be submitted, but the time for evalu-ation is in the range of years. The dossiers both for the activesubstances and for future products must document their chemicalidentity, their efficacy as disinfectants, and their safety for humansand for the environment. Part of the efficacy documentation is ascientifically sound description of the substance’s and the prod-uct’s MOA. The gestalt of the MOA description for an active sub-stance dossier is given in Box 1. The present review gives thedata needed for the MOA description for PAA, the QAC’s, andPHMB. Part of the safety documentation is considerations on thepropensity of the active substance(s) for development of resis-tance. The present review also provides introductions to resistanceissues for these three active substances.
1.2. Disinfectant risk and benefit
The overall technical objective of the EU law on biocides is to en-sure that the benefits of a biocide (e.g., a disinfectant) outweigh therisks (European Parliament & European Council, 2012). Accordingto the law, the risk–benefit balance must be conveyed to the userof the product via the product’s label, as well as being part of thedossier kept by the approving authority (European Parliament &European Council, 2012). The benefit of a disinfectant is that it isefficacious at killing or inhibiting unwanted microorganisms. A riskof a disinfectant can be its toxicity to humans or its propensity forallowing development of resistance to the active substance.
The benefit of efficacy can be promoted or inhibited by manyfactors, and it is knowledge of the active substance’s MOA on thetargeted microorganism that allows predictions of efficacy to bemade (European Commission, 2008b). For instance, knowledge ofhow the quaternary ammonium ion of benzalkonium chlorideinteracts with the bacterial cell membrane calls for caution whensuggesting use of a viscose wipe to apply the disinfectant. Theactive substance is a cation, and viscose is anionic and a solid,which thus can adsorb and ‘‘inactivate’’ the active substance.
The risk of development of resistance to biocides has been apoint of great concern for the European Commission for more than
Table 1Biocidal active substances in this review. See Fig. 1 for chemical structures. Section 3.2 presethese with that of the closely related DDAC.
a Product-types. Disinfectants for: 1, personal hygiene; 2, private and institutional usCommission, 2013).
b Annual production volume within the EU: low, 10–1000 tons; high, >1000 tons; n.i.database of production volumes: http://esis.jrc.ec.europa.eu/index.php?PGM=hpv.
c Stepan Company (2012c).d (Stepan Company 2012a).e (European Commission 2013).f (Stepan Company 2012b).
a decade (European Commission, 2001a, b). Indeed, the scientificliterature abounds with reports of resistance to disinfectant activesubstances, and many of these reports have been reviewed (Daviesand Davies, 2010; Russell, 2003a; McDonnell and Russell, 1999).Evaluation of the risk for the development of resistance is alsohelped by knowledge of the active substance’s MOA (EuropeanCommission, 2008a). The MOA’s for three disinfectant active sub-stances are the primary focus of the rest of this review.
Box 1Data required on mode of action in dossier for approvalof a biocidal active substanceExcerpt of Technical Notes for Guidance on DataRequirements, Common core data set for active substancesand biocidal products (European Commission, 2008c).Italicized text in brackets: reference to text in relevantannex of BPD (European Commission, 2008b; EuropeanCouncil & European Parliament, 1998).
nt
e;
, n
5.4
s differ
3, vete
ot in d
Mode of action (including time delay) [Ann. IIA, V.5.4.]
� The mode of action in terms, where relevant, of thebiochemical and physiological mechanism(s) andbiochemical pathways involved should be stated. Whereavailable, the results of experimental studies must bereported. � Where it is known that in order to exert its intendedeffect the active substance must be converted into ametabolite or degradation product following applicationor use of a preparation containing it, justification shouldbe submitted for why this metabolite or degradationproduct is not considered to be the active substance. Inaddition, available information relating to the formationof reactive metabolites or reaction products must beprovided. This information must include: �the chemical name, empirical and structural formula,molecular mass, and CAS and EC (EINECS, ELINCS or NoLonger Polymers list) numbers if available; �the processes, mechanisms and reactions involved; �kinetic and other data concerning the rate ofconversion and if known the rate limiting step; and �environmental and other factors effecting the rate andextent of conversion. � Indicate also if the actual active substance is the resultof a combined action of different products (i.e., whensuch a combination is necessary to achieve the intendedeffect).
ences among the antimicrobial efficacies of the three ADBAC’s and compares
Product-typea Relative production volumeb
1–5 High
1–4 Low1–4 Low1–4 Low1–4 Low1–5
n.i.n.i.
rinary hygiene; 4, food and feed production; 5, drinking water (European
atabase (European Chemicals Agency, 2012; European Council, 1993). See
458 S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467
1.3. Substances in this review
Table 1 presents the substances whose MOA are described inthis review. PAA was chosen for this review because of its veryextensive use in very large volumes for the disinfection in the foodand feed industries. The QAC’s in Table 1 were chosen as a groupfor the present review because they are used very widely in disin-fection in the production of both food and feed, in institutional anddomestic hygiene, and in personal hygiene, and their propensityfor resistance is well known (European Commission, 2009; Hegstadet al., 2010). PHMB was chosen for this review because its use is onthe rise, particularly in disinfectant wipes. In addition, the polymerwas also chosen for this review because its cationic, amphipathicstructure resembles that of the QAC’s but does not appear to elicitresistance. Differences in modes of action might help explain dif-ferences in propensity for resistance.
2. Peracetic acid
2.1. Introduction
PAA is widely appreciated as a strong oxidizing agent and verysuitable for disinfection (Kitis, 2004; Rajala-Mustonen et al., 1997;Finnegan et al., 2010). The substance is produced in a very high vol-ume, at more than 1000 tons a year in the EU alone (EuropeanChemicals Agency, 2013a; European Commission, 2013). Althoughmuch of the produced PAA may be used as a chemical reactant forother purposes, the shear production volume implies that verymuch is used for disinfection, and thus implies how important itis for the disinfectant user to understand its MOA. PAA is consid-ered more potent than hydrogen peroxide, functioning at lowerconcentrations (McDonnell and Russell, 1999; Kitis, 2004;Ceragioli et al., 2010; Finnegan et al., 2010). Generally, its efficacyas a disinfectant towards target organisms can be ranked as bacte-ria > viruses > bacterial spores > protozoan cysts. As with alldisinfectants, the efficacy is also influenced by the matrix in whichthe microorganism is lodged, such as in its own extracellularpolysaccharide, in a food, or in the soil of an efficacy test (Martinet al., 2008).
Because of the apparently large number of free radicals to arisefrom the reactions of the peroxide with organic compounds, andbecause free radicals are highly reactive, PAA probably inhibits orkills microorganisms by several mechanisms. However, the exactmechanism by which PAA oxidizes and kills a microorganism isstill controversial because of the complexity of the reaction path-way (Rokhina et al., 2010). In the following, PAA reactions will bementioned that result in simple oxidation, in di-hydroxylation ofdouble bonds, and in free-radical formation.
Table 2Summary modes of action and mechanisms of resistance presented in this review.
Substance Mode of action Mechanism ofresistance
PAA Non-specific oxidation, esp. of C–C doublebonds and reduced atoms, e.g., S
None known
QACs Physical disruption of cytoplasmicmembrane, causing i.a. leakage of cations,disruption of membrane functions
Upregulation ofefflux pumps.
PHMB Non-specific adsorption to phospho lipidheadgroups in cell membrane;hexamethylenes disrupt hydrophobic interiorof membrane. Subsequent decline ofmembrane integrity.
None known
2.2. Mode of action
In the first place, the peroxide moiety itself oxidizes other mol-ecules readily, so that, for instance, any exposed SH-groups on thesurface of a microorganism would be oxidized, thus killing the cell(Schlegel, 1981; Dröge, 2002). In commercial biocide formulationsPAA exists in an equilibrium with concentrated hydrogen peroxideand acetic acid, all at concentrations of more than 1%, which forPAA is 130 mM. The di-hydroxylation of C–C double bonds pro-ceeds via a well characterized mechanism (Fig. 2). Depending onthe R groups, in Fig. 2, a single oxidation reaction can producetwo different products. The reaction endows a distinctly hydro-philic region to the product, where there, in the reactant, was def-inite hydrophobicity.
As well as hydroxylation of C–C double bonds, PAA also has thecapability of generating free-radicals. These subsequently oxidize
other compounds in order to resolve the radicals’ electron imbal-ance. The free radical state lets PAA participate in a chain of addi-tions to double bonds (Roberts and Caserio, 1965). This is notidentical to the hydroxylation described above. Free radicals ofPAA can arise by three means. Firstly, in acidic environments, theterminal oxygen of the peracid can be protonated due to its relativeelectron density (Roberts and Caserio, 1965). The resulting oxo-nium group is a suitable leaving group as water in organic reac-tions and results in an acetate ion with an excess electron.
Secondly, free radicals of PAA can also arise if the solution con-tains inorganic ions that can change their valence state by gainingor losing an electron. Such ions could be ferrous (Fe2+), ferric (Fe3+),cobaltic (Co3+), or other ions (Linley et al., 2012; Hendrickson et al.,1970). In this situation, PAA can reduce the inorganic ion by donat-ing a single electron from the peroxide group, while the acid pro-ton dissociates. This leaves a peracyl radical.
Thirdly, free radicals of PAA might also arise as a result of a pho-tochemical reaction (Hendrickson et al., 1970). Here the moleculeis imparted sufficient energy by UV light to break either the perox-ide or hydroxyl bond, followed by dissociation to give a radical. Theliterature does not contain specific reference to the formation ofsuch radicals. However, both the spectrum of the molecule andthe molecule’s photo-instability suggest that UV light can generatesuch radicals (Giguère and Olmos, 1956; Orlando, 2003).
Finnegan and co-workers have recently published an in vitrostudy of the results of the action of H2O2 and PAA on proteins atphysiological conditions (Finnegan et al., 2010). Especially PAA oxi-dized amino acids efficiently, degraded bovine serum albumin, andat millimolar concentrations reduced the efficiency of the enzymealkaline phosphatase. These multiple targets of PAA imply that atthese concentrations a microbial cell is less likely to mobilizeresistance.
The response of bacterial cells to PAA has also been investi-gated. By transcriptome analysis of Bacillus cereus ATCC14579,Ceragioli and co-workers found that the organism responded verysimilarly to hydrogen peroxide and to PAA (Ceragioli et al., 2010).The cells were exposed to both mild and lethal concentrations ofhydrogen peroxide (0.05 and 0.2 mM, respectively) and of PAA(10 and 100 lg ml�1, respectively). Of the genes whose expressiondiffered at the two concentrations, 85% of these were the same forthe two biocides. The remaining 15% showed similar expression ofgene whose functions seemed related. Genes related to DNA repairwere induced at both concentrations of both substances, lethalconcentrations inducing higher expression than mild. One of thegenes that was upregulated is a DNA polymerase that is describedas error-prone in Escherichia coli. Therefore, Ceragioli et al. assayedthe mutation rate in the B. cereus of the two peroxides. At the lethalconcentrations, both substances provoked significantly highermutation rates relative to untreated cells (Ceragioli et al., 2010).
PAA’s DNA-damaging effect when used as a disinfectant hasalso been documented recently in eukaryotic cells. PAA was
S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467 459
analyzed at 0.6 lg ml�1 as a disinfectant of lake water that was tobe used as drinking water and contained a heavy organic load andhoused carp fish (Villarini et al., 2011). As a means of monitoringdisinfection by-products that might be genotoxic, the bile of thefish was assayed in Caco-2 cells after 0, 10, and 20 days. Resultsshowed a significant time-dependent genotoxicity of the bile,which was absent for fish housed in non-PAA-containing water.Evidence that H2O2 kills microbial cells by DNA damage during dis-infection was recently reviewed by Linley et al. (Linley et al., 2012).These workers concluded that the correlation between reduction incell count and DNA damage was not yet proven, but that damage toother cell components may be the more probable cause of death bythat compound.
2.3. Reduced susceptibility to PAA
Although the literature contains reports of protection of themicrobial cell from the effects of PAA by an extracellular matrix,we found no reports of resistance of the cell itself to PAA (Martinet al., 2008). This absence of reports could be expected when recall-ing both PAA’s unspecific and pervasive oxidative MOA and PAA’swide spectrum of efficacy, against bacteria, bacterial spores,viruses, and protozoan cysts.
3. Quaternary ammonium compounds
3.1. Introduction
Structurally, the commercially available QACs are compoundswith a nitrogen atom with covalent bonds to four residues, makingthe nitrogen positively charged. Of the four residues, one is typi-cally an alkyl chain of length between about C5 and C18; one,either a benzyl moiety or one more alkyl chain; and two methylresidues. This cationic, quaternary headgroup is most often accom-panied by a halide as the anion, though in some preparations, bythe hydroxyl anion. See Fig. 1. Because the alkyl groups ultimatelyderive from natural sources, such as coconut and soya bean oil, thealkyl chains of commercially produced QACs are of heterogeneous
Fig. 1. Chemical structures of disinfectant active substances reviewed in thispublication. Top to bottom: peracetic acid (PAA), N-alkyldimethylbenzylammoni-um chloride (ADBAC) (also known as a benzalkonium chloride), didecyl dimethylammonium chloride (DDAC), and poly(hexamethylenebiguanide) hydrochloride(PHMB). Commercial formulations of PHMB have degrees of polymerization (n) of2–40, with an average value of about 12 (European Chemicals Agency, 2011).
lengths (Gilbert and Moore, 2005; European Commission, 2006). Inthe nomenclature, this can result in ambiguous, generic names forQACs that at best are only defined by a range of lengths of the alkylchain(s), as illustrated by the common use of ‘‘benzalkonium chlo-ride’’ and ‘‘dialkyl dimethyl ammonium chloride.’’ It is significantthat, at the start of the EU approval process of the QACs as biocidalactive substances, the European Commission was compelled topublish a document that clarified which range of chain lengthswas covered by which CAS number, and thus which range of chainlengths was covered by which application for EU-wide approval(European Commission, 2006). Taken together, the QAC’s are prob-ably produced at a total of more than 1000 tons of pure compoundsper year in the EU, several of the individual compounds being pro-duced at more than 10 tons per year (European Chemicals Agency,2013a).
3.2. Langmuirian adsorption and desorption
Strictly speaking, the MOA of the cationic antimicrobials isphysical and not chemical, as the molecules themselves are notreactants in chemical reactions but only act as ‘‘wedges’’ and‘‘ion exchangers’’ in the membrane’s lipid bilayer and on the sur-face (Gilbert and Moore, 2005). In light of such physical interac-tions, it would be useful to be able to predict the efficacy of agiven cationic antimicrobial against a potential target microorgan-ism. Such predictions might be about the speed of adsorption ofthe disinfectant molecule to the cell surface, or about the influenceof steric hindrance at adsorption sites, or about the rate of possibledesorption from the surface. These interactions involve the ex-change of molecules between two physical phases, i.e., the aqueoussolution and, to an approximation, the solid surface of the cell. Inthe case of disinfectant molecules that are both high molecularweight and polymers with different degrees of multiplicity, e.g.,PHMB, understanding the physical interaction between moleculeand cell can be made easier by applying models of adsorptionand desorption. Such models exist and are used in sundry fieldsto predict movement between two or three phases, such as inthe fields of heavy metals and of radionuclide waste in the environ-ment (Erichsen et al., 2010; Allison & Allison, 2005). One of themost common models that involves adsorption and desorption toa solid surface is the Langmuir adsorption isotherm (Butt et al.,2006).
The Langmuir adsorption isotherm model makes a first approx-imation possible of the coverage, or fraction of total binding sitesthat are occupied by the solute molecules. The model asserts thatthe adsorption rate in moles per second and per unit area is pro-portional to the number of vacant binding sites and to the concen-tration of the solute (Butt et al., 2006). Likewise, the model assertsthat the desorption rate is proportional to the number of adsorbedmolecules. At equilibrium, the adsorption rate equals the desorp-tion rate. These assertions allow the mathematical derivation ofcoverage. Some of the conditions of the model are only partiallyfulfilled in a common disinfection situation, and cell-disinfectantinteractions are most certainly simplified by the restraints of themodel. However, some useful indication can be derived from themodel about MOA and the conditions that prevail during disinfec-tion and determine its outcome.
When investigating two quaternary ammonium disinfectants,Vieira and Carmona-Ribeiro used Langmuirian adsorption iso-therms of each substance to estimate the number of molecules thatadsorbed to each cell of the yeast Candida albicans and to predictthe species that desorbed (Vieira and Carmona-Ribeiro, 2006).About 6 � 109 molecules of both disinfectants were adsorbed peryeast cell. Denyer and Maillard have published a most useful com-parison of different uptake isotherms, including the Langmuirian,
460 S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467
including how they help explain the modes of action of variousbiocides in gram-negative bacteria (Denyer and Maillard, 2002).
3.3. Mode of action
The QACs are recognized as antimicrobials with broad spectra ofactivity (McDonnell and Russell, 1999; Gilbert and Al-Taae, 1985;Daoud et al., 1983). In general, QAC’s are bactericidal towards veg-etative cells of both gram-positive and gram-negative organisms,though only inhibiting outgrowth of bacterial spores, but not ac-tual germination of them. The mycobacteria are also growth-inhib-ited by the QAC’s but not killed by the molecules. The yeast and thefilamentous fungi are also inhibited, but only the yeast are killed.
In principle, a cell is damaged by a QAC because the long alkylchains permeate the membrane and disrupt its physical and bio-chemical properties, while the charged nitrogen remains at thesurface of the membrane and can disrupt the charge distributionthere. QAC also have intercellular targets and binds to DNA(Zinchenko et al., 2004). A generalized scheme of the MOA of QAC’son the cell membrane is depicted in Fig. 3. Optimal bactericidal andfungicidal alkyl chain lengths have been ascertained for the QACs,through a series of experiments in the group of Gilbert (Daoudet al., 1983; Gilbert and Al-Taae, 1985). The optimal chain lengthagainst Gram-negative bacteria was C16, against Gram-positivesC14, and against yeast and filamentous fungi C12. Ultimately, forthese QAC’s, Daoud et al. concluded that C12–14 was a practicalcompromise when considering hydrophilic and hydrophobic influ-ences on the cell membrane of various organisms (Daoud et al.,1983).
The dual functionality of the QAC molecule was investigated indetail by Gilbert and Al-Taae, and these studies were key to thesubsequent unraveling of the MOA of the QACs (Gilbert andAl-Taae, 1985). These workers followed the viability of cells after30 min. treatment with alkyltrimethylammonium bromides, vary-ing the alkyl chain length from C10 to C14. Organisms tested werePseudomonas aeruginosa, Staphylococcus aureus, and Saacharomyces
Fig. 2. Reaction of PAA with a cis-alkene double bond (redrawn from Geissman, 1968). PAits distal oxygen atom to the double bond, resulting in an epoxide. Then, one of the newmixture of the two C-hydroxy-C-acetate esters. The esters subsequently hydrolyze, resu
cerevisiae. Interestingly, the relationship between viability andchain length in each case was discontinuous, showing dispropor-tionately large increases in antimicrobial activity from one chainlength to the next in the series C10 to C14. This marked increasein effect occurred at the transition C12/C13 for P. aeruginosa, atC10/C11 for S. aureus, and at C11/C12 for S. cerevisiae. Like Daoudet al., Gilbert and Taae reasoned that this disproportionality mightreflect two phases in the adsorption of the molecule to the cell(Gilbert and Al-Taae, 1985). The first adsorption of the QAC mole-cule to the cell could be a simple Langmuirian adsorption to thecell surface by the monomer (Butt et al., 2006). See Section 4.3 be-low for details of a Langmuirian adsorption. Then, monomers areenvisioned to form dimers, as seen with other large biomoleculeswith both hydrophobic and hydrophilic regions (Phillips, 1997;Spitzer, 2011). Ultimately, the QAC dimers are envisioned to passthrough porin channels in the outer membrane with subsequentadsorption to, and disruption of, the cytoplasmic membrane(Gilbert and Al-Taae, 1985).
Since these early studies of the MOA of QAC’s, work has beencarried out using more sensitive techniques and on some of theQAC’s that are most common now in disinfectants. In 2007 Ioannouand colleagues analysed an N-alkyldimethylbenzylammoniumchloride (ADBAC) and didecyldimethylammonium chloride(DDAC), both depicted in Fig. 1 (Ioannou et al., 2007). The ADBACwas a blend of alkyl C14 (50%), C12 (40%), and C16 (10%). All studieswere carried out on S. aureus ATCC 6538. These workers employedvery low concentrations of the QACs (�1–100 lM) in order to sin-gle out prime lesions of biocide attack, be it gross perturbation ofthe outer membrane, or rearrangement of the outer membraneto allow leakage of intracellular components, or masking of vitalcell surface receptors. Higher concentrations of the biocide, itwas reasoned, might elicit multiple effects on the cell, makinganalysis of single effects impossible. The experiments of Ioannouet al. showed that the killing action of DDAC was much less tem-perature-dependent than that of ADBAC, as shown by studies at25 and 35 �C (Ioannou et al., 2007). The two biocides also showed
A is formed in an equilibrium with acetic acid and hydrogen peroxide. PAA first addsC–O bonds is subject to nucleophilic attack by an acetate anion, which produces a
lting in two diols, or glycols, where there originally was a double bond.
Fig. 3. Mechanism of action for a quaternary ammonium biocide. Reproduced with permission (Gilbert and Moore, 2005). Legend for symbols at bottom (left-to-right):phospholipids, protein, benzalkonium chloride, and hydrophilic domain. The segments (a–f) show the progressive adsorption of the quaternary headgroup to acidicphospholipids in the membrane with increasing QAC concentration. This leads to decreased fluidity of the bilayers and the creation of hydrophilic voids in the membrane.Protein function is perturbed with an eventual lysis of the cell and solubilization of phospholipids and proteins into QAC/phospholipid micelles. Inset micrograph showsvesicle formation from outer membrane caused by QAC treatment.
S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467 461
rather different initial uptake isotherms, when studied at concen-trations around 1 lM on 2 � 109 cells. In these experiments, thedisappearance of the QACs from solution was followed. ADBACshowed the Langmuirian uptake, i.e., an adsorption that was con-centration-dependent over a large range of concentrations, taper-ing off towards the maximum. On the other hand, DDAC showeda very strong initial uptake, representing high-affinity binding,according to these authors.
That the bacterial cytoplasmic membrane itself also is damagedby QACs was recently confirmed with benzalkonium chloride(BAC) in Bacillus cereus ATCC 14579, although the alkyl chainlength was not specified (Ceragioli et al., 2010). These workers alsocarried out transcriptome analysis of the B. cereus cells that hadbeen subjected to BAC at mild, at growth-arresting, and at lethallevels (1, 2, or 5 lg ml�1, respectively) for either 10 or 30 min.The microarrays showed increased expression of genes putativelyinvolved in fatty acid metabolism, which for these authors wouldseem to correlate with BACs ability to cause membrane damage(Ceragioli et al., 2010).
The species of QAC molecules that kill the yeast C. albicans hasalso recently been investigated by Vieira and Carmona-Ribeiro(2006) In particular, they looked at QAC aggregate size, comparingthe effects of individual molecules, bi-layer fragments, and large
vesicles. As they expressed their results, ‘‘apparently, individualcationic molecules and their small aggregates penetrate moredeeply in the dense forest of biomolecules on the cell surface thanthe large [aggregate and vesicular] molecules’’ (Vieira andCarmona-Ribeiro, 2006). Indeed, the long, hydrophobic chains arenow assumed to be the primary antimicrobial group of thesemolecules (Gilbert and Moore, 2005).
3.4. Reduced susceptibility to QAC’s
Reduced susceptibility to the QAC’s has been observed in manybacteria and investigated in genetic and molecular detail in some(Rakic-Martinez et al., 2011; Elhanafi et al., 2010; Hegstad et al.,2010). The most widespread mechanism leading to decreased sus-ceptibility to QACs is increased efflux pump activity (Poole, 2005)although other mechanisms may be involved such as alteredfatty acid composition and changes in the bacterial membrane(Guerin-Mechin et al., 2000; Ferreira et al., 2011). On the otherhand the presence of divalent cations increases susceptibility toQAC (Crismaru et al., 2011). Efflux is the extrusion of a solute froma cell (Piddock, 2006). Both eukaryotic and prokaryotic cells con-tain arrays of cytoplasmic membrane transport systems, whoseprimary functions are to maintain the cell’s homeostasis (Paulsen
462 S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467
et al., 1996b). They do this via energy-dependent export of com-pounds that are undesirable for the cell, both spent metabolitesand toxins, and via the uptake of essential nutrients (Webberet al., 2006; Paulsen et al. (1996b)). A number of recent reviewswill give the reader details of the molecular structure, functioning,and genetics of efflux pumps in both gram-negative and gram-po-sitive bacteria (Blair and Piddock, 2009; Poole, 2005, 2007; Pid-dock, 2006). Several efflux pump arrays have been implicated inthe reduced susceptibility of bacteria to many disinfectant com-pounds by pumping these compounds out of the cell once theyget in, thus ensuring a very low concentration in the cytosol. Thefunction of some of these efflux pumps is detailed in the following.
Bacterial efflux systems capable of dealing with antimicrobialsof various sorts generally fall into five classes: the ATP (adenosinetriphosphate)-binding cassette (ABC) family, the resistance–nodu-lation–division (RND) family, the major facilitator superfamily(MF), the small multidrug resistance (SMR) family, and the multi-drug and toxic compound extrustion (MATE) family (Poole, 2007).Most of the five classes can accommodate a broad spectrum of sub-strates and are therefore termed multi-drug resistance pumps. Thesubstrates can include disinfectants, antibiotics, dyes, detergents,and other toxic substances (Webber et al., 2006). Gram-positivebacteria possess disinfectant transporters of the SMR and MF fam-ilies, which are commonly plasmid-encoded, as with the qac genesthat are common in the staphylococci. Gram-negative bacteria cancarry the ABC, RND, and MATE families of efflux pumps. In mostcases these are chromosomally encoded, but in some cases alsoplasmid-encoded. There is considerable literature on the function-ing of some of these five classes of efflux systems on biocides(Poole, 2005, 2007; Piddock, 2006). Considering pumps for disin-fectants, since the first qac genes were characterized, at that timefound in S. aureus, many others have surfaced (Webber et al.,2006). Indeed, the same genes (qacA/B) have very recently been de-tected in Enterococcus faecalis from human blood (Bischoff et al.,2012).
One of the primary targets of disinfection in food and feed pro-duction is the bacterium Listeria monocytogenes. The bacterium isassociated with serious food-borne disease and has the ability topropagate under conditions that normally are used to inhibit bac-terial growth in food and feed, such as low pH, high osmolality, lowwater activity, and cold temperatures (Burall et al., 2012). For thesereasons, L. monocytogenes is the object of concerted disinfectionwith QAC’s in many food and feed industries. However, an addi-tional and confounding factor for combatting the pathogen is itspropensity for reduced susceptibility to the BACs. This is well doc-umented and seems primarily to be due to increased expression ofits efflux pumps relative to more susceptible strains (Romanovaet al., 2006; Webber et al., 2006; Rakic-Martinez et al., 2011).
In the strain L. monocytogens H7550, that was responsible for alarge outbreak in the United States in 1998–1999, Elhanafi et al.investigated the expression of the genes responsible for thisstrain’s BAC resistance (Elhanafi et al., 2010). The resistance was,indeed, the result of an efflux pump akin to one of the many knowntetracycline pumps (Poole, 2002). Although cells not subjected tothe disinfectant had a low basal, but significant level of expressionof the pump, the presence of sublethal levels of BAC (10 lg/ml)raised the expression of the pump genes. This stimulation wasnot, however, observed at higher disinfectant concentrations(e.g., 20 or 40 lg/ml). Quite notably, expression of the BAC resis-tance genes was higher at 4, 8, and 25 �C than at 37 �C. These re-sults tend to indicate that L. monocytogenes might be morecapable of survival and growth in the presence of sublethal concen-trations of BAC under cold conditions than at higher temperatures.Sublethal concentrations under cold conditions can prevail imme-diately after disinfection of food processing facilities. In anotherstudy, long-term exposure to increasing concentrations of the
QAC-containing, commercial disinfectant Triquart SUPER onlymarginally reduced susceptibility, which indicates that selectionconditions and strain variability may influence the outcome ofadaption (Kastbjerg and Gram, 2012).
Furthermore, in the outbreak strain L. monocytogenes H7550,Elhanafi and co-workers saw that the pump effective on BAC is en-coded by three genes, contained within one cassette which in turnis part of a putative transposon (Elhanafi et al., 2010). This constel-lation implies that the resistance cassette could be transferred tonew hosts. Indeed, the cassette is encoded by the large 80-kbpplasmid pLM80, that is closely related to another plasmid foundin other strains of the genus (Kuenne et al., 2010). As yet, no evi-dence has been published for the conjugative ability of pLM80(Elhanafi et al., 2010).
In the hospital environment, S. aureus is an important target forcleaning and disinfection. The hospital-associated methicillin-resistant strains (HA-MRSA) are particularly hard to eradicate,and they give rise to a large fraction of the hospital-acquired infec-tions. S. aureus may express a number of efflux pumps, includingthose encoded by the plasmid-carried qac genes with varying sub-strate specificities (Paulsen et al., 1996a). Chromosomally encodedefflux pumps include NorA and MepA (Costa et al., 2013). Expres-sion of qac genes is controlled by qacR represses expression thatin the absence of substrate (Schumacher et al., 2001). Numerousreports have linked the presence of efflux genes or their expressionlevel with decreased susceptibility to QAC’s. In one study of HA-MRSA strains, a third of the isolates carried qac genes, and uponexposure these strains developed reduced susceptibility to QAC-containing disinfecting agents (Smith et al., 2008). Similarly,another study examined 61 strains of S. aureus and 177 coagu-lase-negative staphylococci isolated from hospital patients, andamong those 50% were considered to be QAC-resistant. Impor-tantly, a substantial part of these strains were antibiotic resistantand carried the blaZ and tetK genes (Sidhu et al., 2001). Expressionlevels appear to be critical when efflux pumps alter the susceptibil-ity to biocides, Huet et al. found that, if they exposed clinicalstrains of S. aureus to low concentrations of a variety of biocidesor dyes in a single step or exposed them to gradually increasingconcentrations over several days, mutants arose that overexpres-sed several different efflux pump genes with mepA predominating(Huet et al., 2008). When exposure was terminated, some of thesemutants reverted to wild-type expression levels, indicating thatefflux pump overexpression may be associated with a fitness cost.
Cationic biocides like chlorhexidine are also among the com-pounds that are effluxed by the qac-encoded efflux pumps. A re-cent study investigated coagulase-negative staphylococci thatwere isolated in the 1960s before the introduction of chlorhexidineinto the wider market (Skovgaard et al., 2013). The study showedthat qacA/B genes were absent from the old isolates, whereas theyappeared in 55% of current isolates, whether obtained from scrubnurses or from non-users of chlorhexidine hand rubs. Here, thepresence of the qac genes was not associated with altered suscep-tibility to chlorhexidine, which indicates that other compoundsmust have selected for the presence of the qac genes in currentisolates.
Investigations have also shown that gram-negative bacteria en-list efflux pumps to withstand biocides, including the QACs. Mostrecently, Whitehead et al. have used the flow cytometer to sortpopulations of Salmonella enterica serovar Typhimurium SL1344that had been subjected either to a low level biocide concentrationor to the recommended in-use concentrations (Whitehead et al.,2011). One of the biocidal products looked at contained a QAC,although the identity of the compound itself and its concentrationin the product were not given. The authors observed that low-doseexposure to the QAC product depleted the membrane potential ofmuch of the population, without disrupting the integrity of the
S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467 463
membrane, and much of this population was not capable of pro-ducing colonies on agar. These cells did over-express one of thetwo efflux pumps investigated. Upon exposure to the in-use con-centration for 5 h, the great majority of cells were dead. However,the few survivors did de-repress expression of the aforementionedefflux pump.
4. PHMB
4.1. Introduction
PHMB is a synthetic polymer, whose chief monomer is closelyrelated in structure to the bisbiguanide chlorhexidine gluconate.PHMB is catalogued under the CAS numbers both of its startingmonomers (27083-27-8) and of the resultant polymer (32289-58-0) (European Chemicals Agency, 2011). The polymer has abroad spectrum of antimicrobial activity against both gram-nega-tive and gram-positive bacteria, some fungi and protozoa (Allenet al., 2004). For many years the substance has been used as anantimicrobial in for instance cosmetics, contact lens solutions,and disinfectants and is sometimes named polihexanide (Frenzelet al., 2011; Hübner and Kramer, 2010). Two different dossierswere submitted in 2007 for the substance’s EU-wide approval asa biocidal active substance, each dossier being for use in all fivetypes of disinfectant products (Table 1) (European Commission,2013). In 2013 the European Chemicals Agency classified PHMBas ‘‘fatal if inhaled’’ [Acute Tox 2–H330 (European Parliamentand European Council, 2008)], which would discourage its use for
Fig. 4. Mechanism of action of PHMB against a bacterial cytoplasmic membrane. Repprogressive interaction of PHMB with the acidic membrane components, that leads todomains then undergo a transition to the more stable hexagonal arrangement, leading toprotein, PHMB, and hydrophilic domain.
disinfection by fumigation or fogging (European Chemicals Agency,2013b).
Concerning the basic chemistry of PHMB, biguanide groups arestrong bases, with a pK1 in the range of 10.5–11.5 and a pK2 at 2–3(Ikeda et al., 1984). This means that at physiological pH the bigua-nide groups are all mono-protonated. Therefore, being a polyelec-trolyte (i.e., polycation), PHMB has a much higher positive chargedensity in the vicinity of the molecule than do single electrolytes.This would account at least in part for strong interactions withnegatively charged molecules. A 12-mer of the polymer has amolecular mass of 2,196 g. By comparison, DDAC has a molecularmass of 326 g.
4.2. Mode of action
Many studies have been carried out to elucidate the mechanismof antimicrobial action of PHMB. Fig. 4 depicts the current under-standing of the interaction of PHMB with a bacterial cytoplasmicmembrane, and comparison with Fig. 3 reveals that PHMB is envi-sioned only to interact with the membrane surface, while theQAC’’s are thought to ‘‘wedge’’ into the phospholipids. In 1983Broxton et al. confirmed in E. coli that PHMB elicits leakage ofpotassium ions, 260 nm-absorbing material, and inorganic phos-phate, all by first-order kinetics (Broxton et al., 1983). These work-ers concluded that damage to the cytoplasmic membrane by PHMBis non-specific, immediate, and irreversible, and results in loss ofselective permeability.
roduced with permission (Gilbert and Moore, 2005). The frames a–d illustrate aa loss of fluidity and eventual phase separation of the individual lipids. Individualmembrane dissolution. Legend for symbols at bottom (left-to-right): phospholipids,
464 S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467
The MOA of PHMB is now quite well understood at the molec-ular level due a series of careful studies. Ikeda et al. conductedin vitro studies of the interaction of PHMB with lipids (Ikedaet al., 1984). From their results, these workers assumed that PHMBis adsorbed onto the surface of the acidic phospholipid phosphati-dylglycerol (PG) bilayer, via interaction with the polar headgroupsof the lipids with the biguanide groups (Ikeda et al., 1984). Thisprovokes disorganization of PG bilayer, which in turn leads to itshigher fluidity, lateral expansion, and a higher permeability. Thedynamics of the first-encounter between PHMB and the cell wascharacterized by Gilbert et al. (Gilbert et al., 1990b). These workersshowed that PHMB is bactericidal in gram-negatives via a ‘‘self-promoting’’ uptake of low molecular mass PHMB (n = 4), and thatthis material has a ‘‘profound synergy’’ with high molecular massPHMB (n = 35) (Gilbert et al., 1990b). Interestingly, E. coli wasinsensitive to high molecular mass PHMB alone, and it was as-sumed that those polymers could not penetrate the cell envelopein order to reach the site of sensitivity, i.e., the cell membrane(Gilbert et al., 1990a). The simultaneous presence of low molecularmass PHMB promoted uptake of the high molecular mass material.For this reason, Gilbert and others conclude that PHMB is mostbiocidally effective when present as a heterodisperse mixture ofvarying molecular masses (Gilbert et al., 1990b).
The cationic nitrogens of PHMB are also assumed to have a dis-tinctly bactericidal role in gram-negatives. These bacteria havedivalent cations that link the core lipopolysaccharide (LPS) mole-cules on the cell surface and ensure a more rigid, stable and imper-meable outer membrane (Nikaido and Vaara, 1985). If thesecations are displaced, the outer membrane is disrupted, which al-lows access of otherwise non-permeant solutes. Indeed, this isthe basis of traditional techniques of genetic transformation of bac-teria that use Ca2+ in the medium to promote the entrance of frag-ments of DNA into the cell cytosol. Gilbert et al. have shown thatPHMB does, in fact, displace the divalent cations in the LPS as aconsequence of the polymer’s adsorption to the cell (Gilbertet al., 1990a). This would also account for the efflux of solutes fromthe cell through the disrupted membrane, as noted above (Broxtonet al., 1983; Moore et al., 2008).
Of practical relevance, Gilbert and co-workers also demon-strated that the adsorption of PHMB to the cell surface is signifi-cantly reduced in the presence of only 40 mM MgCl2 (Gilbertet al., 1990a). We presume this is because the added cations eitherreplace the displaced divalent cations or compete with the PHMBfor the potential anionic adsorption sites. These presumptions cor-relate with those made for disinfectant cloths that have beenimpregnated with PHMB. Such cloths, or towelettes or wipes, aremore and more common as a vehicle for delivering disinfectantsto surfaces via wiping with non-woven textiles. PHMB is one ofthe active substances commonly incorporated into the cloths(Lonza Group Ltd., 2012). However, PHMB, as a cation, is believedto readily adsorb to weakly anionic cellulose moieties in non-woventextiles (Blackburn et al., 2006). Therefore, it has been suggested thatthis adsorption can be reduced by the addition of cations such as Na+
to the water phase in the textile (Lonza Group Ltd., 2012).
4.3. Reduced susceptibility to PHMB
In 2005 Gilbert and Moore asserted that they saw no evidencethat susceptibility to PHMB was affected by the induction or hy-per-expression of multi-drug efflux pumps (Gilbert and Moore,2005). Neither, according to these workers, were there any reportsof acquired resistance. Whatever small changes in MIC that werereported (e.g., elevation of MIC by a factor of 2) correlated withalterations in envelope lipid composition and cation binding. Itwas noted that the absence of greatly reduced susceptibility mostprobably could be attributed to PHMB interacting literally only
superficially with the lipid bilayer, where it alters the bilayer’s flu-idity by cation displacement and changes in headgroup bridging. Inthis connection Gilbert and Moore note that the multi-drug effluxpumps that protect cells from the QAC’s are not efficacious towardsPHMB presumably because PHMB do not become solubilized with-in the membrane core (Gilbert and Moore, 2005). Since theseobservations were made by Gilbert and Moore, to our knowledgeno reports on reduced susceptibility towards PHMB have beenpublished. However, concerted genetic experiments would be use-ful that could confirm that use of PHMB as a disinfectant does notselect for resistance to the QAC’s.
5. Discussion
In time, data on MOA and propensity for resistance will be partof the dossier to be approved for every disinfectant in the EU. Theprescribed conditions of use will vary substantially for each indi-vidual disinfectant product, and such conditions will greatly influ-ence the risk for development of resistance. As a consequence, eachparty wishing approval of a disinfectant will have to formulate itsown discourse and argumentation about the development of resis-tance taking into account the conditions of use. For instance, itmight be foreseen that after disinfection certain uses could leavesmall surface areas behind with residual, sub-inhibitory biocideconcentrations. These residues might permit selection of cloneswith reduced susceptibility to the disinfectant substance(s) inquestion. Such a surface area could be a kitchen counter in thehome that has been quickly rinsed with tap water after disinfec-tion, or it might be a hand that was not completely rinsed underthe tap after disinfection with a QAC. For example low, subinhibi-tory concentrations of QACs are known experimentally to induceexpression of the qac-encoded efflux pumps (Smith et al., 2008).Such conditions may reduce susceptibility simply due to increasedefflux activity, while at the same time providing selection formutations that increase the resistance level. Likewise, wipes asvehicles for applying disinfectants might mean that a wiped sur-face receives a much lower concentration of active substance thanput into the wipe by the manufacturer. Research is in order toinvestigate interactions between disinfectant active substancesand various types of wipe fibers.
The three groups of disinfectant actives in this review areamong the most commonly used in the food and feed industriesand in household and institutional products. PAA has an MOA asa very active, non-specific oxidizer, reacting with electron-richcenters such as reduced sulfur and C–C double bonds. See Table 2.This means that potential targets for oxidation by PAA abound inthe microbial cell, and that the cell is seriously damaged after con-tact with the PAA molecule. One only has to recall that one of thetwo fatty acids on every phosphoglyceride in every cell membraneis usually unsaturated, or that part of the active center of someessential metabolic enzymes is the sulfhydryl group of cysteine(Lehninger, 1970). It follows that a cell would have little chanceof developing mechanisms that endow it with reduced susceptibil-ity to PAA. Indeed, for PAA no reports on resistance or induction ofcross-resistance to antibiotics were found in the literature. Ofcourse, if a cell is encapsulated in a biofilm, including its own thicklayer of extracellular polymers, the chances of PAA reaching thecell surface are less. This may be considered as one of several rea-sons in the food and feed industries for prescribing thorough clean-ing before disinfection (Bridier et al., 2011). Needless to say, thestrong oxidating capacity of PAA demands safety precautions ofthe user.
Contrary to observations about PAA, numerous reports havebeen published for the QAC’s on their induction of resistance,including by induction of efflux pumps. Genetic determinants for
S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467 465
efflux pumps can be encoded by mobile elements, such as conjuga-tive transposons and plasmids, and are thus in principle transfer-able between bacteria. Even though the initial transfer frequencymay be low, subsequent selection of transconjugants by the pres-ence of QAC’s can mean enrichment of clones with reduced suscep-tibility to the QAC’s (Gillings et al., 2009). Therefore, an increase inthe use of QAC’s can worsen problems with resistance to disinfec-tants as well as cross-resistance and co-resistance to antibiotics.Gillings et al. have recently demonstrated the general occurrencein nature of transferable gene cassettes that encode efflux pumpsthat expel both QAC’s and antibiotics (Gillings et al., 2009).Although susceptibility testing was not performed this couldpotentially make the cell co-resistant to the compounds. Theseworkers argue that these double-function cassettes probablyevolved as genetic constructs over time since the first use of QAC’sas hospital disinfectants in the 1930s. This observation is ominousand lends strong support to regulatory efforts to reduce use of QACdisinfectants.
The polymer PHMB molecules, with molecular masses upwardsof 2000 g, exert their effect on the cytoplasmic membrane in a lit-erally superficial manner, in which no chemical reaction is in-volved but only a physical ‘‘bridging’’ between phospholipidheadgroups. It would be difficult to envision how a cell couldmobilize a mechanism to oppose such ‘‘attack’’ by the polymer,and indeed no mechanism for reduced susceptibility has beenreported.
PHMB is already wide-spread as an antimicrobial for varioususes, and not only as a disinfectant. If the conclusion of the EUevaluation of PHMB permits its continued use as a disinfectant,use volumes can be expected to rise. However, the polymer is notreadily biodegradable in the environment, and it is toxic to aquaticorganisms at levels of 10 lg L�1, which is quite low (EuropeanChemicals Agency, 2011). In addition, the polymers’ toxicity tohumans upon inhalation calls for very careful deployment whendisinfecting with PHMB; aerosols must be avoided. Thus, extendeduse of the substance in the future cannot be encouraged.
The data in this review on the three groups of disinfectant ac-tives are only a minor fraction of what exists on MOA and develop-ment of resistance. Even on the basis of these data, it would appearthat basic knowledge of MOA allows qualified predictions to bemade of possible development of resistance, or reduced susceptibil-ity. Indeed, this is one of the reasons that EU law requires knowledgeof MOA for approval of disinfectant active substances and of disin-fectant products (European Council & European Parliament, 1998;European Parliament & European Council, 2012).
6. Conclusions
The EU laws on the approval and marketing of biocides requiresubstantial documentation for approval of disinfectants. The docu-mentation is to ensure that risks during use of the disinfectant donot outweigh the benefits. The documentation that is required in-cludes description of MOA and description of tendency to promoteresistance, or reduced susceptibility. Both descriptions must beaccompanied by scientific documentation and soundargumentation.
With this review we wish to support the use of knowledge ofMOA to make qualified predictions about propensity for resistance.For instance, the mode of action of PAA is a so pervasive oxidationthat a strain of microorganism could hardly modify all exposed C–Cdouble bonds and reduced sulfurs to decrease susceptibility toPAA. For many of the QAC’s, on the other hand, the literatureabounds with studies on microbial resistance to them. Mechanismsfor resistance, molecular biology, genetics, and even possible trans-fer to new host cells have been elucidated for several QAC’s in the
bacteria. Therefore, use of these compounds in general should bereduced over time. For PHMB, no reports were found in the litera-ture of resistance in microorganisms. However, the high toxicity ofthe polymer upon inhalation dictates great caution during deploy-ment to avoid formation of aerosols.
Conflict of interest statement
The authors declare that there are no conflicts of interest.
Acknowledgments
This study was funded by public Grants from the Danish Coun-cil for Strategic Research and the Danish Agency for Science, Tech-nology and Innovation. These agencies had no involvement in theconduct of this study or in the submission of this review.
References
Allen, M.J., Morby, A.P., White, G.F., 2004. Cooperativity in the binding of thecationic biocide polyhexamethylene biguanide to nucleic acids. Biochem.Biophys. Res. Commun. 318, 397–404.
Allison, J. D. & Allison, T. L. (2005). Partition coefficients for metals in surface water,soil, and waste (Rep. No. EPA/600/R-05/074). U.S. Environmental ProtectionAgency: Washington, D.C.
Bischoff, M., Bauer, J., Preikschat, P., Schwaiger, K., Molle, G., Holzel, C., 2012. Firstdetection of the antiseptic resistance gene qacA/B in Enterococcus faecalis.Microb. Drug Resist. 18, 7–12.
Blackburn, R.S., Harvey, A., Kettle, L.L., Payne, J.D., Russell, S.J., 2006. Sorption ofpoly(hexamethylenebiguanide) on cellulose: mechanism of binding andmolecular recognition. Langmuir 22, 5636–5644.
Blair, J.M., Piddock, L.J., 2009. Structure, function and inhibition of RND effluxpumps in Gram-negative bacteria: an update. Curr. Opin. Microbiol. 12, 512–519.
Bridier, A., Briandet, R., Thomas, V., Dubois-Brissonnet, F., 2011. Resistance ofbacterial biofilms to disinfectants: a review. Biofouling 27, 1017–1032.
Broxton, P., Woodcock, P.M., Gilbert, P., 1983. A study of the antibacterial activity ofsome polyhexamethylene biguanides towards Escherichia coli ATCC 8739. J.Appl. Bacteriol. 54, 345–353.
Burall, L.S., Laksanalamai, P., Datta, A.R., 2012. Listeria monocytogenes mutants withaltered growth phenotypes at refrigeration temperature and high saltconcentrations. Appl. Environ. Microbiol. 78, 1265–1272.
Butt, H. J., Graf, K., & Kappl, M. (2006). Physics and chemistry of interfaces.Ceragioli, M., Mols, M., Moezelaar, R., Ghelardi, E., Senesi, S., Abee, T., 2010.
Comparative transcriptomic and phenotypic analysis of the responses ofBacillus cereus to various disinfectant treatments. Appl. Environ. Microbiol.76, 3352–3360.
Costa, S.S., Viveiros, M., Amaral, L., Couto, I., 2013. Multidrug efflux pumps inStaphylococcus aureus: an update. Open Microbiol. J. 7, 59–71.
Crismaru, M., Asri, L.A., Loontjens, T.J., Krom, B.P., de Vries, J., van der Mei, H.C.,Busscher, H.J., 2011. Survival of adhering staphylococci during exposure to aquaternary ammonium compound evaluated by using atomic force microscopyimaging. Antimicrob. Agents Chemother. 55, 5010–5017.
Daoud, N.N., Dickinson, N.A., Gilbert, P., 1983. Antimicrobial activity and physico-chemical properties of some alkyldimethylbenzylammonium chlorides.Microbios 37, 73–85.
Davies, J., Davies, D., 2010. Origins and evolution of antibiotic resistance. Microbiol.Mol. Biol. Rev. 74, 417–433.
Denyer, S.P., Maillard, J.Y., 2002. Cellular impermeability and uptake of biocides andantibiotics in gram-negative bacteria. Symp. Ser. Soc. Appl. Microbiol., 35S–45S.
Dröge, W., 2002. Free radicals in the physiological control of cell function. Physiol.Rev. 82, 47–95.
Elhanafi, D., Dutta, V., Kathariou, S., 2010. Genetic characterization of plasmid-associated benzalkonium chloride resistance determinants in a Listeriamonocytogenes strain from the 1998–1999 outbreak. Appl. Environ.Microbiol. 76, 8231–8238.
Erichsen, A.C., Moehlenberg, F., Closter, R.M., & Sandberg, J. (2010). Models fortransport and fate of carbon, nutrients and radionuclides in the aquaticecosystem at Öregrundsgrepen Stockholm: Swedish Nuclear Fuel and WasteManagement Co.
European Chemicals Agency (2011). Background Document to the Opinionproposing harmonised classification and labelling at Community level ofPolyhexamethylene biguanide or Poly(hexamethylene) biguanidehydrochloride or PHMB. ECHA/RAC/CLH-O-0000001973-68-01/A1.
European Chemicals Agency (2013a). HPV-LPV Chemicals Information System.http://esis.jrc.ec.europa.eu/index.php?PGM=hpv Accessed on June 14, 2013.Database: HPVCs (High Production Volume Chemicals) and LPVCs (LowProduction Volume Chemicals).
466 S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467
European Chemicals Agency (2013b). Proposal for Harmonised Classification andLabelling. CLH Report for PHMB–CAS 27083-27-8 or 32289-58-0 Helsinki:European Chemicals Agency.
European Commission (2001a). Communication from the Commission on aCommunity strategy against antimicrobial resistance (Rep. No. COM(2001)333 Final). Brussels.
European Commission (2001b). Proposal for a Council Recommendation on theprudent use of antimicrobial agents in human medicine (Rep. No. COM(2001)333 final). Brussels.
European Commission (2006). Clarification on substances notified under the BKCand DDAC generic headings in Annex II of Commission Regulation (EC) No.2032/2003. (Rep. No. ENV B.3/PC D(2006)). Brussels.
European Commission (2008a). Assessment of other unacceptable effects. InTechnical Notes for Guidance on Product Evaluation. European Commission:Brussels (Chapter 6).
European Commission (2008b). Efficacy assessment. In Technical Notes forGuidance on Product Evaluation. Brussels (Chapter 7). pp. 59–76.
European Commission (2008c). Technical Notes for Guidance on DataRequirements. Brussels.
European Commission (2009). Assessment of the Antibiotic Resistance Effects ofBiocides. SCENIHR (Scientific Committee on Emerging and Newly IdentifiedHealth Risks).
European Commission (2013). List of participants/applicants whose dossiers arebeing examined under the Review Programme for existing active substancesused in biocidal products. http://ec.europa.eu/environment/biocides. Accessedon July 3, 2013. EU Commission website.
European Council (1993). Regulation (EEC) No 793/93 of 23 March 1993 on theevalation and control of the risks of existing substances. Off. J. Europ. Union, L84/1, 1–75.
European Council & European Parliament (1998). Directive 98/8/EC of the EuropeanParliament and of the Council of 16 February 1998 concerning the placing ofbiocidal products on the market. Off.J.Europ.Union, L123, 1-63.
European Parliament & European Council (2008). Regulation (EC) No 1272/2008 ofthe European Parliament and of the Council of 16 December 2008 onclassification, labelling and packaging of substances and mixtures, amendingand repealing Directives 67/548/EEC and 1999/45/EC, and amending Regulation(EC) No 1907/2006. Off.J.Europ.Union, L 353, 1-1355.
European Parliament & European Council (2012). Regulation (EU) No 528/2012 ofthe European Parliament and of the Council of 22 May 2012 concerning themaking available on the market and use of biocidal products. Off.J.Europ.Union, L167, 1-123.
Ferreira, C., Pereira, A.M., Pereira, M.C., Melo, L.F., Simões, M., 2011. Physiologicalchanges induced by the quaternary ammonium compoundbenzyldimethyldodecylammonium chloride on Pseudomonas fluorescens. J.Antimicrob. Chemother. 66, 1036–1043.
Finnegan, M., Linley, E., Denyer, S.P., McDonnell, G., Simons, C., Maillard, J.Y., 2010.Mode of action of hydrogen peroxide and other oxidizing agents: differencesbetween liquid and gas forms. J. Antimicrob. Chemother. 65, 2108–2115.
Frenzel, E., Schmidt, S., Niederweis, M., Steinhauer, K., 2011. Importance of porinsfor biocide efficacy against Mycobacterium smegmatis. Appl. Environ. Microbiol.77, 3068–3073.
Geissman, T.A., 1968. Principles of Organic Chemistry, third ed. W.H. Freeman andCompany, London.
Giguère, P.A., Olmos, A.W., 1956. Sur le spectre ultraviolet de l’acide peracétique etl’hydrolyse des peracétates. Can. J. Chem. 34, 689–691.
Gilbert, P., Al-Taae, A., 1985. Antimicrobial activity of somealkyltrimethylammonium bromides. Lett. Appl. Microbiol. 1, 101–104.
Gilbert, P., Moore, L.E., 2005. Cationic antiseptics: diversity of action under acommon epithet. J. Appl. Microbiol. 99, 703–715.
Gilbert, P., Pemberton, D., Wilkinson, D.E., 1990a. Barrier properties of the gram-negative cell envelope towards high molecular weight polyhexamethylenebiguanides. J. Appl. Bacteriol. 69, 585–592.
Gilbert, P., Pemberton, D., Wilkinson, D.E., 1990b. Synergism withinpolyhexamethylene biguanide biocide formulations. J. Appl. Bacteriol. 69,593–598.
Gillings, M.R., Xuejun, D., Hardwick, S.A., Holley, M.P., Stokes, H.W., 2009. Genecassettes encoding resistance to quaternary ammonium compounds: a role inthe origin of clinical class 1 integrons? ISME J. 3, 209–215.
Guerin-Mechin, L., Dubois-Brissonnet, F., Heyd, B., Leveau, J.Y., 2000. Quaternaryammonium compound stresses induce specific variations in fatty acidcomposition of Pseudomonas aeruginosa. Int. J. Food Microbiol. 55 (1–3), 157–159.
Hegstad, K., Langsrud, S., Lunestad, B.T., Scheie, A.A., Sunde, M., Yazdankhah, S.P.,2010. Does the wide use of quaternary ammonium compounds enhance theselection and spread of antimicrobial resistance and thus threaten our health?Microb. Drug Resist. 16, 91–104.
Hendrickson, J.B., Cram, D.J., Hammond, G.S., 1970. Organic Chemistry, third ed.McGraw-Hill Book Company, London.
Hübner, N.-O., Kramer, A., 2010. Review on the efficacy, safety and clinicalapplications of polihexanide, a modern wound antiseptic. Skin Pharmacol.Physiol. 23 (Suppl.), 17–27.
Huet, A.A., Raygada, J.L., Mendiratta, K., Seo, S.M., Kaatz, G.W., 2008. Multidrugefflux pump overexpression in Staphylococcus aureus after single and multiplein vitro exposures to biocides and dyes. Microbiology 154, 3144–3153.
Ikeda, T., Ledwith, A., Bamford, C.H., Hann, R.A., 1984. Interaction of a polymericbiguanide biocide with phospholipid membranes. Biochim. Biophys. Acta 769,57–66.
Ioannou, C.J., Hanlon, G.W., Denyer, S.P., 2007. Action of disinfectant quaternaryammonium compounds against Staphylococcus aureus. Antimicrob. AgentsChemother. 51, 296–306.
Kastbjerg, V.G., Gram, L., 2012. Industrial disinfectants do not select for resistance inListeria monocytogenes following long term exposure. Int. J. Food Microbiol. 160,11–15.
Kitis, M., 2004. Disinfection of wastewater with peracetic acid: a review. Environ.Int. 30, 47–55.
Kuenne, C., Voget, S., Pischimarov, J., Oehm, S., Goesmann, A., Daniel, R., et al., 2010.Comparative analysis of plasmids in the genus Listeria. PLoS One 5.
Lehninger, A.L., 1970. Biochemistry. Worth Publishers Inc, New York.Linley, E., Denyer, S.P., McDonnell, G., Simons, C., Maillard, J.Y., 2012. Use of
hydrogen peroxide as a biocide: new consideration of its mechanisms ofbiocidal action. J. Antimicrob. Chemother. 67, 1589–1596.
Lonza Group Ltd. (2012). Vantocil(TM) TG Antimicrobial. Antimicrobial for Wet Wipes.http://www.innovadex.eu/Cleaners/Detail/987/74456/VANTOCIL-TG-ANTIMICROBIAL.
Maillard, J.Y., 2002. Bacterial target sites for biocide action. J. Appl. Microbiol. 92(Suppl.), 16S–27S.
Martin, D.J., Denyer, S.P., McDonnell, G., Maillard, J.Y., 2008. Resistance and cross-resistance to oxidising agents of bacterial isolates from endoscope washerdisinfectors. J. Hosp. Infect. 69, 377–383.
Moore, L.E., Ledder, R.G., Gilbert, P., McBain, A.J., 2008. In vitro study of the effect ofcationic biocides on bacterial population dynamics and susceptibility. Appl.Environ. Microbiol. 74, 4825–4834.
Nikaido, H., Vaara, M., 1985. Molecular basis of bacterial outer membranepermeability. Microbiol. Rev. 49, 1–32.
Orlando, J.J., Tyndall, G.S., 2003. Gas phase UV absorption spectra for peracetic acid,and for acetic acid monomers and dimers. J. Photochem. Photobiol. A 157, 161–166.
Paulsen, I.T., Brown, M.H., Littlejohn, T.G., Mitchell, B.A., Skurray, R.A., 1996a.Multidrug resistance proteins QacA and QacB from Staphylococcus aureus:membrane topology and identification of residues involved in substratespecificity. Proc. Natl. Acad. Sci. USA 93, 3630–3635.
Poole, K., 2002. Mechanisms of bacterial biocide and antibiotic resistance. Symp.Ser. Soc. Appl. Microbiol. 31, 55S–64S.
Poole, K., 2005. Efflux-mediated antimicrobial resistance. J. Antimicrob. Chemother.56, 20–51.
Poole, K., 2007. Efflux pumps as antimicrobial resistance mechanisms. Ann. Med. 39,162–176.
Rajala-Mustonen, R.L., Toivola, P.S., Heinonen-Tanski, H., 1997. Effects of peraceticacid and UV irradiation on the inactivation of coliphages in wastewater. WaterSci. Technol. 35, 237–241.
Rakic-Martinez, M., Drevets, D.A., Dutta, V., Katic, V., Kathariou, S., 2011. Listeriamonocytogenes strains selected on ciprofloxacin or the disinfectantbenzalkonium chloride exhibit reduced susceptibility to ciprofloxacin,gentamicin, benzalkonium chloride, and other toxic compounds. Appl.Environ. Microbiol. 77, 8714–8721.
Rokhina, E.V., Makarova, K., Golovina, E.A., Van, A.H., Virkutyte, J., 2010. Free radicalreaction pathway, thermochemistry of peracetic acid homolysis, and itsapplication for phenol degradation: spectroscopic study and quantumchemistry calculations. Environ. Sci. Technol. 44, 6815–6821.
Romanova, N.A., Wolffs, P.F., Brovko, L.Y., Griffiths, M.W., 2006. Role of efflux pumpsin adaptation and resistance of Listeria monocytogenes to benzalkoniumchloride. Appl. Environ. Microbiol. 72, 3498–3503.
Russell, A.D., 2003a. Biocide use and antibiotic resistance: the relevance oflaboratory findings to clinical and environmental situations. Lancet Infect. Dis.3, 794–803.
Russell, A.D., 2003b. Similarities and differences in the responses of microorganismsto biocides. J. Antimicrob. Chemother. 52, 750–763.
Rutala, W.A., Weber, D.J., HICPAC, 2008. Guideline for Disinfection and Sterilization inHealthcare Facilities, 2008. Centers for Disease Control and Prevention, Atlanta,GA, USA.
Schlegel, H.G., 1981. Allgemeine Mikrobiologie, fifth ed. Georg Thieme Verlag,Stuttgart.
Sidhu, M.S., Heir, E., Sorum, H., Holck, A., 2001. Genetic linkage between resistanceto quaternary ammonium compounds and beta-lactam antibiotics in food-related Staphylococcus spp. Microb. Drug Resist. 7, 363–371.
S. Wessels, H. Ingmer / Regulatory Toxicology and Pharmacology 67 (2013) 456–467 467
Skovgaard, S., Larsen, M.H., Nielsen, L.N., Skov, R.L., Wong, C., Westh, H., et al., 2013.Recently introduced qacA/B genes Staphylococcus epidermidis do not increasechlorhexidine MIC/MBC. J. Antimicrob. Chemother. 68, 2226–2233.
Smith, K., Gemmell, C.G., Hunter, I.S., 2008. The association between biocidetolerance and the presence or absence of qac genes among hospital-acquiredand community-acquired MRSA isolates. J. Antimicrob. Chemother. 61, 78–84.
Spitzer, J., 2011. From water and ions to crowded biomacromolecules: in vivostructuring of a prokaryotic cell. Microbiol. Mol. Biol. Rev. 75, 491–506.
Stepan Company (2012b). Product Bulletin BTC 818 SERIES - www.stepan.com -Accessed Oct. 25, 2012.
Stepan Company (2012c). Product Bulletin STEPANQUAT 50 NF and 65 NF -www.stepan.com - Accessed Oct. 25, 2012.
Vieira, D.B., Carmona-Ribeiro, A.M., 2006. Cationic lipids and surfactants asantifungal agents: mode of action. J. Antimicrob. Chemother. 58, 760–767.
Villarini, M., Moretti, M., Dominici, L., Fatigoni, C., Dorr, A.J., Elia, A.C., et al., 2011. Aprotocol for the evaluation of genotoxicity in bile of carp (Cyprinus carpio)exposed to lake water treated with different disinfectants. Chemosphere 84,1521–1526.
Webber, M.A., Woodward, M.J., Piddock, L.J.V., 2006. Disinfectant resistance inbacteria. In: Aarestrup, F.M. (Ed.), Antimicrobial Resistance in Bacteria ofAnimal Origin. ASM Press, Washington D.C., pp. 115–125.
White, D.G., McDermott, P.F., 2001. Biocides, drug resistance and microbialevolution. Curr. Opin. Microbiol. 4, 313–317.
Whitehead, R.N., Overton, T.W., Kemp, C.L., Webber, M.A., 2011. Exposure ofSalmonella enterica serovar Typhimurium to high level biocide challenge canselect multidrug resistant mutants in a single step. PLoS One 6, e22833.
Zinchenko, A.A., Sergeyev, V.G., Yamabe, K., Murata, S., Yoshikawa, K., 2004. DNAcompaction by divalent cations: structural specificity revealed by thepotentiality of designed quaternary diammonium salts. Chembiochem 5, 360–368.
