Copper as an antibacterial surface agent

Copper can easily be used in public health in its solid form in hospitals and medical environments, for example, doors, knobs, piping material or other inanimate surfaces in different facilities (Casey et al. 2010; Mikolay et al. 2010; Karpanen et al. 2012; Schmidt et al. 2012; Inkinen et al. 2017) or for medical devices (Goudarzi et al. 2017; Schmidt et al. 2017). To test its antimicrobial surface activity, researchers mostly use coupons with varying concentrations of copper.

Noyce et al. (2006) tested copper alloys (61–95% Cu) for their antibacterial activity on E. coli O157 at different temperatures (22°C and 4°C). They demonstrated an antibacterial effect in all conditions, especially at 22°C, but only the high copper alloys (95% Cu) completely killed E. coli (Noyce et al. 2006). Similarly, Wilks et al. (2005) demonstrated copper antibacterial activities on E. coli O157 at different temperatures (20°C and 4°C). They also showed antibacterial activities in all conditions, especially at 20°C and when copper concentration in alloys was superior to 85% (Wilks et al. 2005).

Additionally, the effect of copper alloy surfaces (65–100% Cu) against vegetative and sporal Clostridium difficile was studied and it was demonstrated that copper alloys (> 70% Cu) provide a significant reduction in survival of C. difficile vegetative cells and spores with a complete killing after 24–48 h (Weaver et al. 2008). The antibacterial effect of copper coupons (C19700, 99% Cu) on C. difficile was also highlighted and a complete killing of C. difficile vegetative cells within 30 min and a 99·8% reduction in the viability of C. difficile spores after 3 h was observed (Wheeldon et al. 2008).

A study investigated the antibacterial activity of copper coupons (99% Cu and 63% Cu) on clinical isolates of E. coli, Enterobacter spp., Klebsiella pneumoniae, Pseudomonas aeruginosa and Acinetobacter baumannii all being potent multidrug‐resistant Gram‐negative pathogens responsible for nosocomial infections. Copper surfaces showed antimicrobial activity in all tested strains, especially with copper coupons containing 99% Cu which had a bactericidal effect within 2, 3, 5 and 6 h for A. baumannii, Enterobacter spp., K. pneumoniae and P. aeruginosa and E. coli respectively (Souli et al. 2013).

In summary, the copper antimicrobial activity increases proportionally to its concentration, in accordance with several studies (Mehtar et al. 2008; Elguindi et al. 2009; Zhu et al. 2012; Souli et al. 2013) and pure copper coupons present higher antibacterial efficacy. These studies also highlight that the experimental temperature directly impacts the ‘contact killing’ process, in agreement with other previously published studies (Faúndez et al. 2004; Elguindi et al. 2009; Michels et al. 2009). Furthermore, it was shown that copper ions released from coupons were media dependent (Molteni et al. 2010). The authors compared four different media: 0·1 mol l⁻¹ Tris‐Cl pH 7, M17 (Terzaghi and Sandine 1975), water and 100 mmol l⁻¹ phosphate‐buffered saline pH 7 (NaP_i), and demonstrated that in Tris‐Cl and M17 media complete killing of E. hirae was obtained in 12 and 90 min, respectively, vs 6 h using the other media. They concluded that the application of bacteria to copper surfaces in Tris‐Cl buffers enhanced ‘contact killing’ through a higher release of copper ion and established a causal link between these dissolved copper concentrations and the rates of killing. In the same way, a phosphate buffer with addition of HEPES has been previously shown to increase dramatically the bactericidal effect of the Cu‐H₂O₂ mixture on E. coli (Hartemann et al. 1995).

Copper as an antibacterial particle agent

Copper is also recognized as a potent antibacterial particle agent too. Its use in particle form (especially nanoparticles) shows very interesting properties (Halbus et al. 2017). For example, a high antibacterial activity of chitosan‐copper nanoparticles on many bacterial strains including MRSA, Bacillus subtilis, P. aeruginosa and Salmonella choleraesuis has been shown (Usman et al. 2013). Giannousi et al. (2014) also demonstrated the antibacterial activity of copper‐based nanoparticles and showed that copper nanoparticles (Cu‐NPs), as well as cuprous oxide (Cu₂O) induced plasmid DNA degradation in a dose‐dependent manner in Gram‐positive and Gram‐negative strains. Moreover, they showed that the concentration of released ions was below the level of inhibiting bacterial growth (minimum inhibitory concentration). Based on these results, the authors concluded that the concentration of released ions was less important than the nanoparticle size for the antibacterial activity. In addition, and as for copper surface, it appears that the outer membrane would be the front line of bacterial defence against copper ions (Speer et al. 2013). Therefore, the size of Cu‐NPs turned out to be a major contributing factor for copper antimicrobial activity. Other authors synthesized CuO nanoparticles (CuO‐NPs) which were purified and dried into different CuO‐NPs sizes (Thekkae Padil and Černík 2013). Small CuO‐NPs (4·8 ± 1·6 nm) were found to have significantly better antibacterial activity than larger particles (7·8 ± 2·3 nm). Azam et al. (2012) also confirmed these results and Applerot et al. (2012) highlighted that the smaller particles have a better antimicrobial activity due to their greater capacity to penetrate cells (Applerot et al. 2012; Azam et al. 2012). However, their nano size is also responsible for their cyto‐ and genotoxicity and must be taken into account when considering their use (Karlsson et al. 2009; Midander et al. 2009; Wang et al. 2012).

Many other recent studies have confirmed the antimicrobial activity of copper in nanoparticle forms and their potential usefulness against infections (Pramanik et al. 2012; Pinto et al. 2013; Shankar et al. 2014; Wei et al. 2014; Bogdanović et al. 2015; Kruk et al. 2015; Gutiérrez et al. 2017; Maqbool et al. 2017; Qadri et al. 2017). However, their rapid oxidation upon air exposure limits their antimicrobial use in aerobic conditions (Usman et al. 2013).

Copper as an antifungal surface agent

Although less research has been carried out on antifungal activity of copper it is unanimously accepted that an activity similar to that described for bacterial species exists in fungus.

Quaranta et al. (2011) investigated the mechanism of ‘contact killing’ in Candida albicans and yeast Saccharomyces cerevisiae cells in contact with copper coupons (C11000 99·9% Cu and C75200 62% Cu) (Quaranta et al. 2011). The modification of copper homeostasis caused four to six times faster killing of C. albicans lacking copper‐ATPase export and S. cerevisiae deficient for copper uptake transporters, both involved in the intracellular regulation of copper, than of wild‐type cells due to an intracellular accumulation of copper. The authors showed that the first damages were localized on membranes, similar to the bacterial mechanism of ‘contact killing’. In addition, mutation detection assays showed the complete absence of DNA damage. The Live/Dead staining assay confirmed this hypothesis by showing rapid and extensive cytoplasmic membrane damage after yeast exposure to copper surfaces.

However, in C. albicans strains, which express higher levels of the CRP1 P1‐type ATPase copper transporter gene, the resistance against copper is greater by regulating the intracellular uptake of copper. ALS1 and ALS3, a cluster of genes encoding cell surface‐associated glycoproteins, could regulate CRP1 and de facto suggest a different resistance mechanism from ‘contact killing’ (Zheng et al. 2016).

Copper as an antifungal particle agent

As in bacteria, the Cu‐NPs have a potent antifungal activity. Ghasemian et al. (2012) tested antifungal activity against filamentous fungi such as Alternaria alternata, Aspergillus flavus, Fusarium solani and Penicillium chrysogenum. They synthesized 8 nm Cu‐NPs and showed an important antifungal activity of these particles with minimal inhibitory concentrations (MIC) ranging from 40 to 80 mg l⁻¹ (Ghasemian et al. 2012) and that particle size is an important factor in the antimicrobial activity of copper. Other studies have reported the antifungal activity of Cu‐NPs against Candida species (Usman et al. 2013; Kruk et al. 2015).

Copper as an antiviral surface agent

Copper also has the ability to destroy viruses, such as influenza viruses, noroviruses or human immunodeficiency virus (HIV).

Warnes and Keevil (2013) showed a rapid inactivation of murine norovirus (MNV‐1) on copper alloy dry surfaces (65% to 99·9% Cu) at room temperature (Warnes and Keevil 2013). Similar to ‘contact killing’ of bacterial and fungal species, they demonstrated that Cu(II) and especially Cu(I) were the primary effectors of MNV‐1 inactivation. Quenching superoxide and hydroxyl radicals did not provide better protection suggesting ROS (reactive oxygen species) had little effect on norovirus inactivation in this situation. Finally, they showed that the viral genome is targeted by copper, especially a gene encoding VPg (viral‐genome‐protein‐linked, a viral protein essential for viral infectivity), by gene copy number reduction. Warnes et al. (2015) confirmed these results and reported that MNV‐1 was inactivated more rapidly on alloys containing 79–89% copper than on alloys containing 70% copper (Warnes et al. 2015). These results confirm that copper concentration affects the antiviral efficiency of copper.

In addition, the inactivation of monkeypox and vaccinia viruses have also been demonstrated on copper surfaces (Bleichert et al. 2014).

Copper as an antiviral particle agent

Noyce et al. (2007) tested the antiviral property of copper against influenza A virus (Noyce et al. 2007). They inoculated 2·10⁶ virus particles onto a copper surface at 22°C and 50–60% relative humidity. They demonstrated that only 500 viral particles were active after an incubation of 6 h on copper surface while up to 500·000 viral particles still remained infectious after 24 h of incubation on stainless steel.

Moreover, Fujimori et al. (2012) showed that Cu‐NPs (nano‐sized 160 nm copper iodide CuI) exert antiviral activity on influenza A virus by degradation of viral proteins (Fujimori et al. 2012). They also demonstrated that antiviral activity was closely related to the generation of hydroxyl radicals derived from Cu(I).

Other studies have looked at the usefulness of these antiviral properties for public health. Borkow et al. (2010) have demonstrated that oxide copper impregnation in respiratory protective face masks confers biocidal properties against human and avian influenza A virus without altering their physical barrier properties (Borkow et al. 2010). Tested in simulated breathing conditions (28·3 l min⁻¹), no infectious viral particle was recovered from the copper oxide containing masks after 30 min, unlike with the control masks. In the same way, Borkow et al. (2008) demonstrated the effectiveness of copper during HIV‐1 passage through filters containing copper oxide powder or copper oxide‐impregnated fibres (Borkow et al. 2008). HIV‐1 is inactivated when exposed to copper oxide in a dose‐dependent manner, without neither cytotoxicity nor strain specificity. It is also assumed that the mechanism of HIV‐1 inactivation targets the viral genome, and more particularly the HIV viral protease, which is essential for the replication of the virus. Karlström and Levine (1991) showed that an approximate stoichiometric concentration of Cu(II) ions could inhibit this viral protease in a rapid and irreversible manner (Karlström and Levine 1991).

More recently, Hang et al. (2015) showed antiviral activity of Cu‐NPs against hepatitis C virus (HCV) and that Cu‐NPs inhibit HCV infection by targeting the binding of infectious HCV particles to hepatic cells and the virus entry into the cells (Hang et al. 2015).

Membrane damage

Many studies suggest that ‘contact killing’ is initiated by the dissolved copper ions released from the copper surfaces by the culture medium and causing cell alterations (Grass et al. 2011). Espírito Santo et al. (2012) studied ‘contact killing’ on Staphylococcus haemolyticus using copper (C11000 99·9% Cu) and stainless steel coupons (Espírito Santo et al. 2012). They showed a great intracellular accumulation of copper inducing cell death by membrane damage. In addition, cells exposed to copper did not present more DNA mutation than those exposed to steel. These data suggest that copper is not genotoxic and does not kill by DNA damage but rather by damaging the cell membrane. Many other studies on Escherichia coli have also shown that copper exposure rapidly leads to membrane alterations before DNA degradation (Espírito Santo et al. 2011; Warnes et al. 2012). Hong et al. 2012 studied more specifically the mechanism of membrane degradation in E. coli exposed to different copper alloy surfaces (70–99·9% Cu) or in medium containing CuSO₄. They showed that ‘contact killing’ of E. coli is triggered by nonenzymatic oxidative damage of membrane phospholipids, resulting in the loss of membrane integrity and cell death.

All these studies concluded that the antibacterial effect of copper was related to its ability to release copper ions and their damaging effect on cell membrane. In addition, numerous recent studies on new copper‐based material confirm this theory (Zhang et al. 2013; Liu et al. 2014; Mathews et al. 2015; Emam et al. 2017).

Oxidative stress and DNA degradation

Many studies attribute the antibacterial activity of copper to its capacity to release ions that cause oxidative stress by production of ROS in aerobic conditions (Espírito Santo et al. 2008, 2011). Following membrane degradation, copper‐released ions can penetrate into the cell. The increase in the intrinsic amount of copper rapidly causes an important oxidative stress demonstrated by redox cycling between the different forms of copper: native Cu, Cu(I) and Cu(II) (Espírito Santo et al. 2011). Under aerobic conditions, redox potential enables copper to produce hydroxyl radicals according to the Haber–Weiss and Fenton reactions (Liochev and Fridovich 2002). The release of ROS leads to damage of lipids, proteins and nucleic acids, and eventually to the destruction of all genetic material (Dalecki et al. 2017).

Other studies tend to demonstrate that the membrane is the first target to undergo damage, allowing copper ion penetration into the cell, followed by oxidative stress and DNA degradation (Espírito Santo et al. 2012; Tian et al. 2012; Cui et al. 2014; San et al. 2015). All these data are consistent with the review of Grass et al. (2011) and Prado et al. (2012) explaining the same sequence of events for the ‘contact killing’ mechanism (Grass et al. 2011; Prado et al. 2012).

Despite these mechanisms, the exact molecular mechanism of ‘contact killing’ remains controversial. Weaver et al. (2010) investigated the “contact killing” mechanism of copper surface on methicillin‐resistant S. aureus (MRSA) (Weaver et al. 2010). They inoculated MRSA onto copper coupons (C11000 99·9% Cu) and showed a rapid killing of MSRA by compromising cellular respiration and inducing DNA damage. However, little effect on cell membrane integrity was found. The same results were obtained by other authors with vancomycin‐resistant E. faecalis (VRE) and Enterococcus faecium, under similar experimental conditions (Warnes et al. 2010; Warnes and Keevil 2011). They established that enterococci exposed to a copper surface for at least 2 h showed no detectable viable cell using cell culture and plate counts. They demonstrated an inhibition of cell respiration and cell death and therefore suggested that membranes could not be damaged after cell death. Hence, they concluded that membrane damage could not be assumed to be a universal and unique ‘contact killing’ starting point. They proposed an alternative model where copper antimicrobial efficacy is linked to the release of copper ionic species that trigger ROS production and lead to cell respiration arrest and DNA breakdown at the first stages of cell death. Recently, the same authors confirmed this hypothesis on MRSA and methicillin‐sensitive S. aureus strains and confirmed the role of ROS in ‘contact killing’ but not via Fenton chemistry (Warnes and Keevil 2016). In addition, Mathews et al. (2015) suspected that ROS action was probably not the primary cause of cell death but an accompanying effect (Mathews et al. 2015).

Interestingly, another primary mechanism for copper toxicity in bacteria has been shown by non‐Fenton destruction of iron–sulphur cluster enzymes in E. coli. Toxicity of copper ions is due to the displacement of iron atoms from iron–sulphur clusters and a copper binding to the corresponding sulphur atoms (Macomber and Imlay 2009). The rapid inactivation of iron–sulphur cluster enzymes by copper results in the damage of the central catabolic and biosynthetic pathways. Therefore, copper ions may compete with other metal ions for important protein binding sites.

Copper resistance in Gram‐negative bacteria

Gram‐negative bacteria activate copper‐ATPase export and chaperone gene transcription via an ‘activator‐type’ (CueR‐like and CusRS‐like) when cytoplasmic and periplasmic concentrations of copper increase. The CueR‐like are one‐component regulators that respond to cytoplasmic copper concentration while the CusRS‐like are two‐component regulators that respond to periplasmic copper concentration. In addition, many Gram‐negative species synthesize multicopper oxidases (MCO), by the same ‘activator‐type’, to allow the transformation of Cu(I) into the less toxic Cu(II) form. Likewise, Gram‐negative species synthesize resistance nodulation cell division (RND)‐type multicomponent copper efflux systems that export copper to the periplasmic space when a high concentration of copper is detected (Rademacher and Masepohl 2012). When copper concentration increases, the CusRS‐like two‐component system detects the high periplasmic copper amounts, triggers transcriptional activation of cusRS gene, which subsequently leads to the export of copper to the extracellular compartment. In parallel, the one‐component CueR‐like system recognizes high cytoplasmic copper concentrations and triggers transcriptional activation of copper resistance genes, with or without operon organization, for example, MCO, chaperone and copper‐ATPase export (Figure 2).

Osman et al. (2013) demonstrated how these mechanisms are taking place in Salmonella enterica sv. Typhimurium (Osman et al. 2013). When copper cytoplasmic concentration increases, copper is transported from the cytoplasm to the periplasm by the CopA and GolT—two copper‐ATPases export—and transferred to SodCII—a superoxide dismutase—and CueO, a MCO, with a periplasmic copper chaperone CueP. It has been shown that CueP expression is controlled by a CueR‐like ‘activator‐type’ which can detect an increase in cytoplasmic concentration of copper using the two‐component system CpxR/CpxA when stress is detected at the bacterial membrane level (Pezza et al. 2016).

Recently, a CopM Cu‐binding protein was found both in the periplasm and in the extracellular space of the cyanobacterium Synechocystis. CopM could bind copper in the extracellular compartment, preventing copper accumulation in the periplasm, but its role outside the cell is not clearly known (Giner‐Lamia et al. 2015, 2016).

These observations raise many questions regarding the exact mechanism of copper homeostasis and underline the need to carry out further investigations (Rademacher and Masepohl 2012).

Copper resistance in Gram‐positive bacteria

Contrary to Gram‐negative bacteria, Gram‐positive bacteria repress copper‐ATPase export and chaperone gene transcription by a ‘repressor‐type’ (CopY‐like and CsoR‐like) under copper deficiency conditions (Rademacher and Masepohl 2012). When the amount of copper increases, the CopY‐like and CsoR‐like repressors are released from the DNA to allow expression of the downstream genes, enabling transcription of copper‐ATPase and chaperone for export and sequestration of free copper ion respectively (Figure 3).

Cobine et al. (1999) demonstrated, in E. hirae, that copper chaperones CopZ deliver Cu(I) to the Zn(II)CopY complex resulting in displacement of Zn(II) (Cobine et al. 1999). This direct interaction leads to the release of the CopY repressor from the promoter and allow the transcription of copYZAB operon when copper is in excess. However, direct copper transfer between CopZ and CopA is strongly suspected via specific interactions but further studies are needed to confirm this hypothesis (Multhaup et al. 2001). A similar hypothesis was made for CopB (Solioz and Stoyanov 2003).

Beside these transcriptional regulators, a study in E. hirae has demonstrated that higher copper concentrations (>0·5 mmol l⁻¹) down‐regulate CopZ by inducing its proteolysis by a copper‐dependent protease (Lu and Solioz 2001; Solioz 2002). This study reports a post‐translational mechanism responsible for the control of the intracellular amount of CopZ. Authors concluded that this copper‐stimulated proteolysis of CopZ could play an important role in E. hirae copper homeostasis. Further studies are needed to confirm whether the post‐translational proteolysis of copper regulators occurs in other bacterial species and is a regulatory mechanism of CopZ‐like proteins and possibly also copper‐ATPases.

Other aspects of bacterial copper resistance

It is generally recognized that Gram‐positive species express ‘repressor types’ while Gram‐negative bacteria express ‘activator types’ ones. However, these mechanisms are not systematically used. For example, the Gram‐positive bacillus, Corynebacterium glutamicum, expresses the two‐component system CopRS, a copper‐ATPase CopB and the extracellular MCO CopO following exposure to an increase in copper concentration. Furthermore, C. glutamicum also has a copper‐ATPase ctpV gene and a repressor csoR gene, both repressed by CsoR during copper deficiency (Schelder et al. 2011). Thus, C. glutamicum possesses both copper resistance mechanisms, via activation and repression. Similarly, Thermus thermophilus, a Gram‐negative bacterium, possesses a copZ‐csoR‐copA operon that is repressed by CsoR during copper deficiency (Sakamoto et al. 2010).

Several other control systems of copper homeostasis have been reported. Mycobacterium tuberculosis has a ctpV gene, repressed by CsoR, responsible for the expression of copper‐ATPase membrane exporter (CtpV) (Liu et al. 2007; Ward et al. 2010; Marcus et al. 2016). However, Wolschendorf et al. (2011) showed another M. tuberculosis defence system against copper: the loss of the mctB gene, encoding the copper transport protein B (MctB), causes considerable intracellular accumulation of copper (Wolschendorf et al. 2011). Similarly, another defence pathway has been shown in M. tuberculosis involving the copper repressor RicR regulon encoding for (i) MymT, a copper‐protective metallothionein causing copper sequestration (Mierek‐Adamska et al. 2017), (ii) LpqS, a putative lipoprotein able to export copper and (iii) Rv2963, a putative permease aiding the export of copper (Festa and Thiele 2011; Shi et al. 2014). Finally, another study has shown that MmcO, a mycobacterial MCO, was an important mechanism of copper resistance in the same bacterium (Rowland and Niederweis 2013).

Another way of copper sequestration has been demonstrated in E. coli. Phenolate siderophores, such as yersiniabactin, sequester Cu(II) outside the cell and prevent its reduction to the more toxic Cu(I) form (Chaturvedi et al. 2012). In addition, copper‐yersiniabactin complexes can catalyse superoxide dismutase, and thus reduce the concentration of superoxide (Chaturvedi et al. 2014). As a result, Cu(I) concentration is reduced while Cu(II) amount increases. Therefore, copper sequestration by yersiniabactin can promote bacterial survival (Chaturvedi and Henderson 2014).

Resistance in E. hirae and more generally enterococcus genus has also been studied. Studies showed that the tcrYAZB operon conferred copper resistance in E. faecium, E. faecalis and other enterococcal species (Hasman 2005; Hasman et al. 2006). This operon is genetically organized as the copYZAB operon found in E. hirae and presents the same mode of repressor control. However, few studies have reported the intra‐ and interspecies transferability potential of these genes, located on a plasmid, and their role in acquiring copper resistance (Hasman and Aarestrup 2002; Amachawadi et al. 2013, 2015).

Copper resistance in yeasts

In yeasts, the three most common regulator mechanisms of copper homeostasis are: (i) reducing the copper influx pump; (ii) exporting copper in the extracellular environment with copper export‐ATPases and (iii) expression of metallothioneins to chelate free copper ions (Ooi et al. 1996; Georgatsou et al. 1997; Labbé et al. 1997; Liu and Thiele 1997; Yamaguchi‐Iwai et al. 1997; Riggle and Kumamoto 2000; Weissman et al. 2000). However, the ‘contact killing’ mechanism in yeasts is not fully understood.

In C. albicans and under copper‐limiting conditions, the CTR1 and FRE7 genes, coding respectively for the copper import protein Ctr1p and two proteins with cupric reductase activity Fre7p and Fre10p, are expressed (Marvin et al. 2004; Jeeves et al. 2011; Mackie et al. 2016). On the contrary, under copper‐replete conditions their expression is repressed (Woodacre et al. 2008). Other genes were identified and linked to copper homeostasis in C. albicans. CRP1 encodes copper‐ATPase export pumps and CUP1 and CRD2 encode copper metallothioneins. A deletion of one of these genes decreases the copper resistance of C. albicans (Riggle and Kumamoto 2000; Weissman et al. 2000; Mackie et al. 2016). The literature reports similar mechanisms for S. cerevisiae (Thiele 1988; Buchman et al. 1989; Szczypka and Thiele 1989) and Aspergillus nidulans (Antsotegi‐Uskola et al. 2017).

Other studies of copper homeostasis in fungi have shown the same mechanisms and proteins involved. Ding et al. (2013) showed that Cryptococcus neoformans neutralizes copper to promote infection by encoding copper importer (Ctr1), copper‐detoxifying metallothionein (Cmt) and a transporter implicated in copper sequestration (Ding et al. 2013) similarly, Zhang et al. (2016) showed that the Ctr4 transporter is involved in copper homeostasis (Zhang et al. 2016).

Wiemann et al. (2017) reported that copper transporter proteins Ctr1 and copper‐ATPase export pumps play a major role in copper resistance of A. fumigatus (Wiemann et al. 2017) and Dietl et al. (2016) recently showed that histidine is involved in copper detoxification in A. fumigatus. They demonstrated that deletion of the gene encoding imidazoleglycerol‐phosphate dehydratase (HisB) reduces resistance of A. fumigatus to copper (Dietl et al. 2016).