Molecular mechanisms of Cr(VI) resistance in bacteria and fungi

Authors

  • Carlo Viti,

    Corresponding author
    1. Dipartimento di Scienze delle Produzioni Agroalimentari e dell'Ambiente – sezione di Microbiologia, Università degli Studi di Firenze, Florence, Italy
    • Correspondence: Carlo Viti, Piazzale delle Cascine 24, 50144 Florence, Italy.

      Tel.: +39 0553288307;

      fax: +39 0553288272;

      e-mail: carlo.viti@unifi.it

    Search for more papers by this author
  • Emmanuela Marchi,

    1. Dipartimento di Scienze delle Produzioni Agroalimentari e dell'Ambiente – sezione di Microbiologia, Università degli Studi di Firenze, Florence, Italy
    Search for more papers by this author
  • Francesca Decorosi,

    1. Dipartimento di Scienze delle Produzioni Agroalimentari e dell'Ambiente – sezione di Microbiologia, Università degli Studi di Firenze, Florence, Italy
    Search for more papers by this author
  • Luciana Giovannetti

    1. Dipartimento di Scienze delle Produzioni Agroalimentari e dell'Ambiente – sezione di Microbiologia, Università degli Studi di Firenze, Florence, Italy
    Search for more papers by this author

Abstract

Hexavalent chromium [Cr(VI)] contamination is one of the main problems of environmental protection because the Cr(VI) is a hazard to human health. The Cr(VI) form is highly toxic, mutagenic, and carcinogenic, and it spreads widely beyond the site of initial contamination because of its mobility. Cr(VI), crossing the cellular membrane via the sulfate uptake pathway, generates active intermediates Cr(V) and/or Cr(IV), free radicals, and Cr(III) as the final product. Cr(III) affects DNA replication, causes mutagenesis, and alters the structure and activity of enzymes, reacting with their carboxyl and thiol groups. To persist in Cr(VI)-contaminated environments, microorganisms must have efficient systems to neutralize the negative effects of this form of chromium. The systems involve detoxification or repair strategies such as Cr(VI) efflux pumps, Cr(VI) reduction to Cr(III), and activation of enzymes involved in the ROS detoxifying processes, repair of DNA lesions, sulfur metabolism, and iron homeostasis. This review provides an overview of the processes involved in bacterial and fungal Cr(VI) resistance that have been identified through ‘omics’ studies. A comparative analysis of the described molecular mechanisms is offered and compared with the cellular evidences obtained using classical microbiological approaches.

Introduction

Chromium, which belongs to the group VI-B transition metals of the periodic table, has an atomic number of 24, is the most abundant heavy metal, together with zinc, in the lithosphere (69 μg g−1; Li, 2000) and the 21st most abundant element in the Earth's crust (ranging from 100 to 300 μg g−1; Cervantes et al., 2001). This metal is introduced into the environment from natural sources such as volcanic eruptions, forest fires, and weathering, but the largest contribution to the deposition of chromium in the biosphere is the result of anthropogenic activities. Chromium, due to its hardness, sheen, high melting point, odorlessness, and anti-corrosiveness, is utilized in various industrial activities, including electroplating, steel, and automobile manufacturing, wood treatment, leather tanning, pigments in dyes, paints, inks, plastics, and military defense applications (Langård, 1980; James, 1996; Viti & Giovannetti, 2007).

Chromium exists in different oxidation states but its two most stable oxidation forms in the environment are the hexavalent [Cr(VI)] and trivalent [Cr(III)] forms (Bartlett, 1991; Zayed & Terry, 2003). These oxidation states have different chemical features and affect organisms in different ways. Cr(III) is conventionally considered an essential micro-nutrient in the diet of animals and humans. Nevertheless, it has been reported recently that chromium can no longer be considered an essential element because rats on a diet with low-Cr(III) suffered no adverse consequences to body composition, glucose metabolism or insulin sensitivity compared with rats on a diet with a sufficient dose of Cr(III) (Di Bona et al., 2011). On the other hand, a high dose (supra-nutritional level) of Cr(III) in the diet improved insulin sensitivity (Di Bona et al., 2011). Cr(III) complexes accumulating in the body are potentially genotoxic (Levina et al., 2003) and therefore their use in micro-nutrient or antidiabetic treatments should be reconsidered after the accurate analysis of available and/or emerging data (Levina & Lay, 2008; Di Bona et al., 2011). Cr(III) is relatively insoluble under environmental conditions (Bartlett & Kimble, 1976; Sass & Rai, 1987) and is considered less toxic than Cr(VI) because it does not pass through cell membranes (Czako-Ver et al., 1999; Francisco et al., 2002). Cr(VI) (chromate and dichromate) is soluble, very toxic, mutagenic, and carcinogenic to humans (Petrilli & De Flora, 1977; Langård, 1990; Losi et al., 1994; Gibb et al., 2000), and is involved in birth defects and in reproductive health (Kanojia et al., 1998).

Cr(VI), being a serious threat to human health and able to spread readily beyond the site of its initial generation as a water-soluble waste through aquatic systems and groundwater, is considered a severe pollutant (Kamaludeen et al., 2003; Viti & Giovannetti, 2007). Chromate, due to its chemical structural similarity with sulfate, crosses over the cellular membrane via the sulfate uptake pathway (Ramirez-Diaz et al., 2008). Once inside the cells, Cr(VI) undergoes reduction due to various enzymes and non-enzymatic activities (Ramirez-Diaz et al., 2008), resulting in the production of several active intermediates that could directly cause alterations of DNA and toxic effects (Shi & Dalal, 1990b; Viti & Giovannetti, 2007; Zhitkovich, 2011).

The intracellular reduction of Cr(VI), via cellular reductants (i.e. ascorbic acid, glutathione, and flavoenzymes), produces active intermediates Cr(V) and/or Cr(IV) (Bose et al., 1992; Suzuki et al., 1992; Stearns & Wetterhahn, 1997; Lay & Levina, 1998; Kalabegishvili et al., 2003) free radicals, with Cr(III) as the final product (Dillon et al., 1997; Ortega et al., 2005; Kanmani et al., 2012). Cr(III) affects DNA replication, causes mutagenesis, and alters enzyme structure and activity by reacting with their carboxyl and thiol groups (Bridgewater et al., 1994; Tsou et al., 1996; Cervantes et al., 2001; Sobol & Schiestl, 2012). The reduction of Cr(VI) to Cr(V) is accompanied by the reduction of molecular oxygen to peroxide, which in turn produces hydrogen peroxide. Cr(V) generates hydroxyl radicals via reaction with hydrogen peroxide (Shi & Dalal, 1990ab). Therefore, reactive oxygen species (ROS) are generated during the gradual reduction of Cr(VI) to the final Cr(III). Several authors (Kawanishi et al., 1986; Itoh et al., 1995; Luo et al., 1996) have hypothesized that oxidative damage to DNA due to intracellular reduction of Cr(VI) is responsible for the genotoxicity seen. Sumner et al. (2005) have reported that oxidative damage to proteins is a central mechanism of chromium toxicity in Saccharomyces cerevisiae. Chromium produces a wide spectrum of genomic damages such as DNA strand breaks, alkali-labile sites, DNA-protein, and DNA-DNA crosslinks, and Cr(III)-DNA adducts (Stearns & Wetterhahn, 1997; Voitkun et al., 1998; Codd et al., 2001; Sugden et al., 2001; Salnikow & Zhitkovich, 2008; Reynolds et al., 2009). Cr(VI) can also bind to cellular materials and consequently hinder their normal physiological functions (Sumner et al., 2005; Kanmani et al., 2012).

In environments polluted with Cr(VI), microorganisms must have efficient systems for neutralizing the negative effects of this form of chromium (Morais et al., 2011). Over time, numerous Cr(VI)-resistant/tolerant bacterial strains belonging to various taxa have been isolated (McLean & Beveridge, 2001; Srinath et al., 2001; Francisco et al., 2002; Camargo et al., 2003; Kamaludeen et al., 2003; Viti et al., 2003; Faisal & Hasnain, 2004; Viti & Giovannetti, 2007; Alam & Malik, 2008; Halpern et al., 2009; He et al., 2009, 2010, 2011; Sturm et al., 2011; Zhang & Li, 2011; Verma & Singh, 2013). Chromosomal and plasmid-borne chromate-determinants have also been observed (Ramirez-Diaz et al., 2008, 2011; Caballero-Flores et al., 2012). Various mechanisms of Cr(VI) resistance or detoxification have been described, such as the efflux of chromate ions from the cell cytoplasm and reduction of extracellular Cr(VI) to Cr(III) (Ramirez-Diaz et al., 2008). Numerous data are also available concerning Cr(VI) toxicity against fungi (yeasts and filamentous fungi) and their systems that help survival after unsafe exposure to chromate (Tamás et al., 2005).

Several recent reviews have summarized Cr(VI) resistance-tolerance mechanisms in microorganisms (Ramirez-Diaz et al., 2008; Poljsak et al., 2010). Studies have been performed on Cr(VI) efflux pumps (Nies et al., 2006; Branco et al., 2008), extracellular Cr(VI) reduction to Cr(III) (Priester et al., 2006; Gnanamani et al., 2010; Belchik et al., 2011; Chovanec et al., 2012), enzymes involved in the ROS detoxifying processes (Ackerley et al., 2004ab; Cheng et al., 2009), repair of DNA lesions (Llagostera et al., 1986; Hu et al., 2005; Miranda et al., 2005; Decorosi et al., 2009), and sulfur metabolism (Viti et al., 2009; Christl et al., 2012).

The knowledge of Cr(VI)-resistance systems in microorganisms traditionally derives from physiological analysis of wild-type and mutant strains stressed by the metal. The spread of the ‘omic’ approaches, especially transcriptomic and proteomic methods, has offered the possibility to carry out studies on the response of global gene and protein expression of microorganisms. The ‘omic’ approaches have shed light on cellular processes that until now were unexplained or poorly highlighted in terms of their involvement in chromate resistance (i.e. enhancement of cysteine biosynthesis, enhancement of iron transport and sequestration, derepression of some central metabolic routes, and phage lytic cycle activation), opening new fields of study to which traditional microbiological methods can contribute.

This review summarizes what is known about the molecular mechanisms that confer resistance to Cr(VI) in bacteria and fungi, with more emphasis on current findings. In particular, this overview examines the results obtained applying ‘omic’ methods, which have provided substantial information about the complexity and multiplicity of mechanisms that bacteria and fungi have to protect themselves against Cr(VI) toxicity. In addition, this review synthesizes the knowledge collected by classical approaches.

Genes/proteins involved in central metabolism and amino acid transport/metabolism

Exposure to Cr(VI) in bacteria strongly modifies the expression of genes related to metabolism. Although microbial growth can be slightly stimulated in the presence of low concentrations of chromate, generally chromate affects the genes involved in metabolism, eliciting a reduction of the growth (Henne et al., 2009b; Zhou et al., 2012). However, energy and reducing power are needed to counteract the toxic effects of chromate exposure through extrusion, reducing activities and repair mechanisms (Ramirez-Diaz et al., 2008; Monsieurs et al., 2011). A proteomic analysis of Arthrobacter sp. strain FB24 demonstrated a large number of proteins with altered expression in the presence of chromate: 140 involved in carbohydrate transport and metabolism, 112 involved in energy production and conversion, and 176 involved in amino acid transport and metabolism (Henne et al., 2009b). The amount of enzymes involved in phosphoenolpyruvate (PEP) and pyruvate production, the components of the pyruvate dehydrogenase complex, and an aldehyde dehydrogenase decreased, resulting in a reduction of the growth of Arthrobacter sp. FB24 with exposure to Cr(VI) levels above 5 mM (Henne et al., 2009b). Similar results were obtained by studying the proteomic profiles of Shewanella oneidensis MR-1, where a dramatic time-dependent decrease of the proteins involved in energy metabolism was revealed upon exposure to 1 mM Cr(VI) (Brown et al., 2006). A reduction of growth is expected for the S. oneidensis MR-1 strain, since it is characterized by a minimum inhibitory concentration (MIC) below 2 mM for chromate, whereas the strong reduction of growth of Arthrobacter sp. FB24 upon exposure to Cr(VI) at a dose that is 1 : 40 that of the MIC [Cr(VI)-MIC of 200 mM] is not easy to explain. These results could be due to the media used. In fact, a proteomic study of Pseudomonas putida F1 performed under different growth conditions indicated that chromate toxicity was dependent on the growth medium: chromate toxicity was at least 40-old greater in minimal medium than in rich medium (Thompson et al., 2010).

The importance of energy and reducing power production for counteracting Cr(VI) exposure has been reported in several works. Genes related to carbohydrate metabolism (glyoxylate and dicarboxylate metabolism, citrate cycle, pyruvate metabolism, glycolysis, gluconeogenesis, and butanoate metabolism) showed increased expression in transcriptomics studies of Ralstonia metallidurans exposed to chromate (Monsieurs et al., 2011). Analyzing Cr(VI)-resistance in the Cr(VI)-hyperresistant Pseudomonas corrugata 28, Decorosi et al. (2009, 2011) demonstrated that two mutants, one defective in a malic enzyme family gene (ME) and the other in a pyridine nucleotide transhydrogenase, exhibited Cr(VI) sensitivity with respect to the wild-type. This feature was explained by the lower capability of the mutants to generate a reducing power to repair cellular damage and Cr(VI) reduction, generally performed by NAD(P)H-dependent Cr(VI)-reductases that are present in several bacteria (Ishibashi et al., 1990; Suzuki et al., 1992; Park et al., 2000; Decorosi et al., 2011). It is known that ME has a role in providing reducing power to repair cellular damage resulting from oxidative stress in plants, animals, and bacteria (Revilla et al., 1987; Smeets et al., 2005). A proteomic study of Pseudomonas aeruginosa, using a level of Cr(VI) that only modestly affects the viability of the strain, demonstrated that the proteins responsible for energy production, such as ATP synthase, flavoproteins, malate dehydrogenase, and fructose-1,6-bisphosphatase, were overproduced to generate ATP for the production of proteins involved in counteracting the effects of Cr(VI) (Kilic et al., 1981). Glycolytic enzymes and enzymes involved in energy production are known to be inactivated by oxidation, as it has been reported for phosphoenolpyruvate carboxykinase, pyruvate dehydrogenase complex, malate dehydrogenase, and glyceraldehyde-3-phosphate dehydrogenase in several biological systems (Fucci et al., 1983; Weber et al., 2004). A proteomic study of Arthrobacter sp. FB24 in the presence of Cr(VI) confirmed the above (Henne et al., 2009b).

In Caulobacter crescentus, a multitude of genes involved in metabolism (glutamate synthesis, phosphate starvation response, polyhydroxybutyrate, and carbon/energy utilization) and electron transport processes were observed to be up-regulated in the presence of chromate by whole genome transcriptional analysis (Hu et al., 2005). Interestingly, Hu et al. (2005) found an increased demand for glutamate, which is one of the central amino acids linking nitrogen and carbon/energy metabolism through α-keto-glutarate. In the same study the observed up-regulation of the serine biosynthesis pathway suggested an increased flow of glutamate towards α-keto-glutarate, which is not only a central molecule in metabolic pathways but also an antioxidant that plays a key role in detoxification of ROS (Mailloux et al., 2009) such as those produced by Cr(VI) reduction inside the cells. In contrast, the proteins involved in amino acid transport and metabolism were more strongly down-regulated during acute chromate challenge of Pseudomonas putida F1 when minimal medium was used (Thompson et al., 2010). This could be due to the arrest of growth rather than to a direct response to chromate stress, because the dose of chromate used was 10-fold the Cr(VI)-MIC. In Escherichia coli, approximately one-half of the genes encoding amino acid biosynthetic enzymes were down-regulated in response to growth arrest (Chang et al., 2002).

Combined transcriptomic and proteomic analysis of S. oneidensis MR-1 exposed to 0.3 mM Cr(VI) indicated that this strain modulates electron chain components, resulting in the coexistence of a subset of dehydrogenase and reductase/oxidase (Chourey et al., 2006).

Energy metabolism also plays a role in fungal responses to heavy metals. The model organism S. cerevisiae underwent a genome-wide transcriptional profiling following exposure to Cr(VI) and other metals (copper, silver, zinc, cadmium, mercury, and arsenic; Jin et al., 2008). From this analysis, a number of genes were highlighted as transcriptionally responsive to all the tested metals, with 388 up-regulated genes and 447 down-regulated genes being observed. The whole list was designated as Common Metal Responsive (CMR) genes. A number of up-regulated CMR genes are involved in carbohydrate and fatty acid metabolism, particularly in glycolysis, oxidative phosphorylation, and alcohol metabolism. These processes were previously recognized as necessary for energy production in the form of ATP subsequently used for ATP-dependent molecular chaperones and other energy-dependent stress responses (Gasch & Wemer-Washburne, 2002).

The Aspergillus sp. Ed8 and the Penicillium sp. H13 were able to decrease Cr(VI) concentration by reduction in the growth medium, without changing the total amount of chromium in the medium (Acevedo-Aguilar et al., 2006). The reduction of Cr(VI) to Cr(III) required the presence of a carbon source that is either fermentable like glucose or oxidizable like glycerol. The need for an active cellular metabolism generating reducing power was established by the fact that in the absence of a carbon source and in cell-free extracts no changes in the Cr(VI) amount in the growth medium were observed (Acevedo-Aguilar et al., 2006).

Chromate-resistance determinants (CRDs)

CRDs have been identified in Archaea, Bacteria and Eukarya (Nies et al., 1998; Flores-Alvarez et al., 2012), and consist of genes belonging to the chromate ion transport (CHR) superfamily (Ramirez-Diaz et al., 2008). Generally, CRDs include the chrA gene, which encodes a putative chromate efflux protein driven by the membrane potential (Pimentel et al., 2002).

The ChrA protein has been characterized in detail in P. aeruginosa, Cupriavidus metallidurans, and Shewanella sp. ANA3 (Cervantes et al., 1990; Nies et al., 1990; Aguilar-Barajas et al., 2008), and many putative ChrA homologs have been identified (Fig. 1). Microorganisms bearing chrA orthologs display highly variable resistance capacities to Cr(VI) concentrations ranging from 0.35 to 200 mM (Juhnke et al., 2002; Henne et al., 2009b; Monsieurs et al., 2011), indicating that the mere presence of the chrA gene cannot explain by itself the Cr(VI) resistance. Unlike other metal resistance systems, it has been reported that chrA genes provide Cr(VI) protection only in the submillimolar range. On the other hand a strong activation of the Chr efflux pump could lead to the coextrusion of sulfate, resulting in sulfur starvation conditions and, therefore, conditions not suitable for growth (Branco et al., 2008).

Figure 1.

Radial phylogenetic tree from an alignment of 513 putative ChrA chromate ion transport sequences using clustalx. The branch colors are as follows: fungi, brown; algae, olive green; Archea, orange; Proteobacteria (Alpha-, bright green; Beta-, magenta; Delta-, blue; Gamma-, purple); Cyanobacteria, turquoise; Firmicutes, yellow; Actinobacteria, red; all other Bacteria, black. Modified from Henne et al. (2009a).

In bacteria, the chrA genes can be located on plasmid or chromosomal DNA or both (Juhnke et al., 2002), and they are generally organized in operons with other chr genes (Fig. 2). In C. metallidurans, the genes involved in the resistance to heavy metals are present in a 34-kb region on the pMOL28 plasmid, where they are organized in the chrFECAB cluster, or in a chromosomal operon chr2 (chrB2, chrA2, chrF2 genes), whereas in Ochrobactrum tritici 5bv11, chr genes are part of a transposable operon organized in the chrBACF cluster (Branco et al., 2008). The genes flanking chrA could have regulatory functions or may not be directly related to chromate transport. chrB, which is a common feature of the chr operons, acts as a chromium-sensitive regulator of the chr operon; it is specific for Cr(VI) in O. tritici 5bv11 (Branco et al., 2008), whereas it is activated by Cr(III), Cr(VI) and sulfate in C. metallidurans (Peitzsch et al., 1998; Juhnke et al., 2002). No clear involvement in chromate resistance has been identified for the other genes surrounding chrA orthologs. The activity of chrC and chrF genes did not affect chromate resistance in O. tritici 5bv11 (Branco et al., 2008). Cupriavidus metallidurans strain AE126 lacking chrB1 (Juhnke et al., 2002) showed a chromate tolerance comparable to the wild-type (Juhnke et al., 2002), and only a weak reduction of Cr(VI)-MIC was observed in the same strain when chrC and chrI were deleted (Juhnke et al., 2002). In C. metallidurans, ChrC protein had similarities to SODs with active manganese or iron centers (Juhnke et al., 2002). In the same study, the deletion of chrF2 (chromosomal gene) in C. metallidurans strain AE104 (plasmid-free derived strain from C. metallidurans CH34) increased the Cr(VI)-MIC, whereas the deletion of chrA2 and chrB2 strongly decreased the Cr(VI)-MIC. In Arthrobacter sp. FB24, the deletion of chrJ, chrK or chrL resulted in a noticeable decrease in chromate resistance. Shewanella sp. strain ANA-3 has a plasmid resistance operon, chrBAC, arranged in a similar manner to the operon observed in pMOL28 of C. metallidurans CH34 (Aguilar-Barajas et al., 2008).

Figure 2.

Schematic representation of the main local genomic context of the chr genes analyzed by ‘omic’ approaches (Juhnke et al., 2002; Diaz-Perez et al., 2007; Aguilar-Barajas et al., 2008; Branco et al., 2008; Henne et al., 2009a). chrA, chromate ion transporter; chrB, chromium-sensitive regulator; chrC, superoxide dismutase; chrE, hypothetically involved in cleavage of some chromium-glutathione complexes; chrF, regulatory protein; chrI, regulatory protein; chrJ, putative malate:quinone reductase; chrK, YVTN beta-propeller repeat-containing protein; chrL, probable conserved lipoprotein (LppY/LpqO family); SCHR, small chromate ion transporter (chrA ortholog); tnpA, transposase; tnpR, resolvase.

The expression of the chrA gene alone in E. coli confers a high level of resistance to chromate, whereas the results relative to the expression of the entire chrBAC operon in E. coli were different if a low-copy or high-copy number vector was used for the construct preparation. The expression of the chrBAC operon on a high-copy number vector elicited no Cr(VI) resistance, whereas, using a low-copy number vector, the cells bearing the operon were 10-fold more tolerant to Cr(VI) compared both with cells bearing chrA alone and with wild-type cells (Aguilar-Barajas et al., 2008). Therefore, the effects of chrA flanking genes on chromate resistance are not unambiguous and may differ depending on the host strains. However, a description of the CHR superfamily is beyond the scope of this review and is fully treated, along with the phylogeny of chrA genes, in previous publications (Nies et al., 1998; Diaz-Perez et al., 2007; Ramirez-Diaz et al., 2008; Henne et al., 2009a).

Proteomic, transcriptomic or genomic studies in C. metallidurans, Arthrobacter, and S. oneidensis have demonstrated the effects of chromate exposure on CRDs. In C. metallidurans, a transcriptional study showed that the plasmidic chrFECAB-cluster was strongly up-regulated in the presence of chromate at a sub-lethal concentration, whereas it was not induced by Cd, Cu, Ni, Pb, Zn, Hg, Ag, Tl, Au, Sr, Cs, Mn, As, Co, and Se exposure (Monsieurs et al., 2011). In addition, the chr chromosomal cluster was up-regulated in the presence of chromate (Monsieurs et al., 2011), suggesting that these genes also participate in chromate resistance in C. metallidurans, as previously hypothesized by a transcriptional study (Juhnke et al., 2002). Shewanella oneidensis MR-1 genome analysis revealed a chromosomal putative chromate transporter gene (so0986); transcriptome profiling and mass spectrometry-based proteomics indicated that this gene was not up-regulated in either acute or chronic Cr(VI)-reducing conditions (Bencheikh-Latmani et al., 2005; Chourey et al., 2006). The response to Cr(VI) of Arthrobacter sp. FB24 has been established at the proteomic level (Henne et al., 2009b) and by real-time reverse transcriptase PCR (Henne et al., 2009a). The latter approach showed that all CRD genes were induced selectively in the presence of chromate; nevertheless, at the proteomic level the presence of proteins produced by the expression of Arth_4254 (ChrB-Cterm, chromate resistance-related protein), Arth_4248 (ChrA, chromate efflux protein), and Arth_4251 (predicted ORF carrying ChrA conserved domain) was not revealed. The authors stated that this was a consequence of the technical difficulties in detecting and quantifying proteins of low molecular weight and with membrane-spanning regions (Henne et al., 2009b).

CHR-1 protein (homologous to ChrA) is present in several ascomycetes and other fungal phyla, such as Basidiomycota and Zygomycota, whereas CHR-1 is absent in the vast majority of yeasts (Flores-Alvarez et al., 2012). To our knowledge, Neurospora crassa is the only fungus in which a CHR-1 protein has been characterized (Flores-Alvarez et al., 2012). The expression of the CHR-1 gene was induced upon Cr(VI) exposure, but the encoded protein elicited the opposite effect as the expected. Instead of chromate tolerance, CHR-1 expression resulted in an increase in Cr(VI) sensitivity, indicating that it is a chromate transporter but promotes chromate accumulation inside the cell instead of extrusion (Flores-Alvarez et al., 2012). A possibility for the role of chromate uptake for the Chr-1 protein in N. crassa may be that the transported chromate is then destined for vacuolar compartmentalization. This detoxification mechanism has been described in S. cerevisiae in relation to several toxic oxyanions (Gharieb & Gadd, 1998).

Genes/proteins involved in Cr(VI) reduction

Microbial reduction of Cr(VI) to Cr(III) can be considered a chromate detoxification mechanism and is usually not plasmid-associated (Cervantes et al., 2001). Two direct Cr(VI) reduction mechanisms have been described: (1) Cr(VI) is reduced under aerobic conditions commonly associated with soluble chromate reductases that use NADH or NADPH as cofactors (Park et al., 2000) and (2) Cr(VI) can be used as an electron acceptor in the electron transport chain under anaerobic conditions by some bacteria (Tebo & Obraztova, 1998). Cr(VI) can be also reduced indirectly by nonspecific reactions associated with redox intermediate organic compounds such as amino acids, nucleotides, sugars, vitamins, organic acids or glutathione (Myers et al., 2000; Robins et al., 2013).

The first analyzed chromate reductase was a membrane-associated enzyme from Enterobacter cloacae HO1 that transfers electrons to Cr(VI) via NADH-dependent cytochromes (Ohtake et al., 1990). Cr(VI) reductases were primarily characterized in the context of alternative substrates. These enzymes usually show a NADH:flavin oxidoreductase activity and can also act as chromate reductases. Representative examples are the reductases NfsA/NfsB from Vibrio harveyi, with a nitrofurazone nitroreductase property as a primary activity and chromate reductase as a secondary function (Kwak et al., 2003), and the ferric reductase FerB from Pseudomonas denitrificans, which uses Fe(III)-nitrilotriacetate and chromate as substrates (Mazoch et al., 2004). To discover effective nitroreductases for cancer gene therapy, a collection of bacterial oxireductases that demonstrated Cr(VI) reduction activity was assembled from different bacterial species (Robins et al., 2013). On the basis of results achieved Robins et al. (2013) suggested the use of immobilized Cr(VI)-reduction enzymes instead of living bacterial cells in the bioremediation processes to overcome the limit of the toxicity of Cr(VI) itself to the remediating bacteria. Among the 15 candidates of purified enzymes, E. coli NemA was identified as a highly efficient Cr(VI) reductase using NADH as a cofactor (Robins et al., 2013).

To date, the best studied chromate reductase is ChrR from P. putida, a soluble flavin mononucleotide-binding enzyme able to catalyze the reduction of Cr(VI) to Cr(III) (Park et al., 2000). The ChrR enzyme functions as a NADH-dependent reductase, having a broad substrate specificity and permitting the NAD(P)H dependent reduction of quinones, prodrugs, Cr(VI), and U(VI) ions (Barak et al., 2006). ChrR catalyzes a combination of one- and two-electron transfers to Cr(VI) with the formation of the unstable species Cr(V) before further reduction to Cr(III) (Fig. 3). Although a proportion of the Cr(V) intermediate is spontaneously reoxidized to generate ROS, its reduction to Cr(III) through two-electron transfer, minimizes the production of harmful radicals. Studies with purified P. putida ChrR revealed that this enzyme has a quinone reductase activity during chromate reduction. Quinols produced by quinone reduction confer tolerance to ROS (Gonzalez et al., 2005). Thus, ROS generated by ChrR activity during Cr(VI) reduction should be neutralized by quinols formed by the quinone reductase activity of the same enzyme (Ackerley et al., 2004a; Cheung & Gu, 2007; Ramirez-Diaz et al., 2008) (Fig. 3). Therefore, although ChrR activity generates ROS during Cr(VI) reduction, it reduces quinones that provide protection against ROS (Ramirez-Diaz et al., 2008).

Figure 3.

Mechanisms of enzymatic reduction of Cr(VI). ChrR of Pseudomonas putida catalyzes a combination of one- and two-electron transfers to Cr(VI) with the transient formation of Cr(V) that in the presence of several biomolecules is reduced to Cr(III), generating ROS. ChrR of Escherichia coli mediates a four-electron shuttle for the direct reduction of Cr(VI) to Cr(III) with the remaining electron transferred to oxygen. MtrC and OmcA are the terminal reductases of Cr(VI) in Shewanella oneidensisMR-1. Moreover, both in aerobiosis and in anaerobiosis, a multitude of soluble reductase (SR) or membrane-bound reductase (MR) enzymes can reduce Cr(VI) using NADH or endogenous electron reserves as electron donors. Modified from Wang & Shen (1995).

In E. coli, a ChrR (ChrR of E. coli was formerly called YieF) protein that shares sequence homology with the P. putida ChrR enzyme was observed (Barak et al., 2006). The reaction mechanism of YieF is different from that described for ChrR of P. putida and involves an obligatory four-electron reduction of chromate in which the enzyme simultaneously transfers three electrons to chromate to produce Cr(III) and one electron to molecular oxygen, generating ROS (Ackerley et al., 2004b) (Fig. 3). This reaction mechanism generates less ROS than that described for ChrR of P. putida, based on the combination of two- and one-electron reduction, thus YieF should be a more suitable enzyme for chromate detoxification than the P. putida ChrR (Ramirez-Diaz et al., 2008).

Although a large number of studies have demonstrated the role of ChrR in Cr(VI) reduction, proteomics revealed that this protein in P. putida F1, possessing a ChrR with a 100% amino acid identity to that of P. putida KT2404 (Barak et al., 2006), was down-regulated in response to acute chromate exposure in all conditions tested (Thompson et al., 2010). On the other hand, temporal genomic and proteomic studies of S. oneidensis MR-1 indicated that a NADPH-dependent FMN reductase [SO3585, incorrectly annotated as putative azoreductase (Mugerfeld et al., 2009)], sharing approximately 28% of identity with ChrR of P. putida, was significantly up-regulated in Cr(VI)-exposed cells (Brown et al., 2006; Thompson et al., 2007), especially at the highest chromate doses used (Thompson et al., 2007). The deletion of the so3585 gene was not critical for cell survival in the presence of chromate, and only an initial decrease of Cr(VI) reduction rate was observed (Mugerfeld et al., 2009) and therefore more studies are need to understand the role of so3585 in chromate resistance.

Microbial respiration with Cr(VI) as the terminal electron acceptor has never been rigorously shown (Richter et al., 2012). Nevertheless, the global transcriptomic analysis of S. oneidensis MR-1, treated with 100 μM Cr(VI) as the sole electron acceptor, revealed the up-regulation of genes encoding MtrA, MtrB, MtrC, and OmcA (Bencheikh-Latmani et al., 2005), which are involved in the dissimilatory extracellular reduction of solid ferric iron [Fe(III)] (hydr)oxides, uranium [U(VI)] and technetium [Tc(VII)] (Belchik et al., 2011). The cytochromes MtrC and OmcA of S. oneidensis MR-1 were deeply characterized to understand their role in Cr(VI) reduction. The data obtained supported the idea that MtrC and OmcA are the terminal reductases of Cr(VI) in S. oneidensis MR-1 (Belchik et al., 2011).

Chromate reduction has been also associated with biosorption. Fein et al. (2002) showed nonmetabolic reduction of Cr(VI) to Cr(III) by bacterial surfaces under nonutrient conditions as probable results of the oxidation of organic molecules within the cell wall that serve as electron donors for Cr(VI) reduction to Cr(III). Nancharaiah et al. (2010), studying Cr(VI) reduction by aerobically grown granular bacterial biofilms, found that there was not reduction of Cr(VI) to Cr(III) under nonutrient conditions, whereas they efficiently reduced Cr(VI) from minimal media in the presence of acetate. In both studies, X-ray absorption near-edge structure (XANES) spectroscopy and extended X-ray adsorption fine structure (EXAFS) were used, demonstrating that these approaches produce useful information about the speciation and association of the Cr immobilized on microbial biomass.

Very little is known about the mechanisms mediating Cr(VI) reduction in fungi. However, fungi have the ability to reduce Cr(VI), and many studies have been performed to exploit this capability for environmental bioremediation. Filamentous fungi, such as Aspergillus sp., Penicillium sp., and Trichoderma inhamatum, reduce Cr(VI) to Cr(III) by exploiting the reducing power generated by carbon metabolism as mechanism of Cr(VI) detoxification (Acevedo-Aguilar et al., 2006; Morales-Barrera & Cristiani-Urbina, 2008). Paecilomyces lilacinus has demonstrated the ability to both biotransform Cr(VI) and accumulate it in the biomass, exerting the maximum reduction activity during the log phase of growth, when cellular metabolic activity is maximized, and maximum accumulation during the stationary phase (Sharma & Adholeya, 2011). Aspergillus niger strains have been described as coping with chromium mainly via the biosorption of the metal into the cells, rather than via the use of reducing activity (Sandana Mala et al., 2006). The Ed8 strain of Aspergillus tubingensis, included in the A. niger species complex, demonstrated the ability to decrease Cr(VI) concentration in the medium via a reduction mechanism stimulated by carboxylic acids and metal-chelating agents (Coreno-Alonso et al., 2009).

Extracellular reduction of Cr(VI) to Cr(III) was observed during the growth of Candida utilis by mechanisms independent from the intensity of culture growth or initial chromium concentration (Muter et al., 2001). On the basis of their results Muter et al. (2001) hypothesized that Cr(VI) reduction in C. utilis could be partly dependent on pH changes of broth during the exponential phase or on exo-enzymatic activities during stationary phase. Candida maltosa, isolated from tanning liquors from a leather factory and characterized by a high tolerance level of chromate in comparison with the yeast laboratory strains C. albicans, S. cerevisiae, and Yarrowia lipolytica, demonstrated the ability to reduce Cr(VI) both in the presence of viable intact cells and in cell-free extracts (Ramirez-Ramirez et al., 2004). This ability was related to NADH-dependent chromate reductase activity associated with soluble proteins and, to a lesser extent, with the membrane fraction (Ramirez-Ramirez et al., 2004). Recently, the reduction of Cr(VI) to Cr(III) through an enzymatic mechanism has been observed in Pichia. Both in intact cells and in cell-free extracts of P. jadinii M9 and P. anomala M10 strains chromate was reduced, suggesting the presence of a chromate reductase activity possibly associated with the cytosolic or membrane proteins (Martorell et al., 2012). A non-enzymatic mechanism of Cr(VI) reduction has been described for A. niger (Park et al., 2005).

The interplay between biosorption and reduction mechanisms, which could also shed light on the role of pH can be summarized as shown in Fig. 4. Low pH values contribute to Cr(VI) detoxification by two actions, providing the protons needed for reduction reaction and increasing the protonation level of the adsorbent surface, which is thus more positively charged and more attractive for negative Cr(VI) ions (Park et al., 2005; Martorell et al., 2012).

Figure 4.

Mechanisms of extracellular chromate reduction in fungi. (a) Mechanism I: 1 chromate binding by anionic absorption, 2 reduction by adjacent electron donor, and 3 release by electronic repulsion. (b) Mechanism II: 1 reduction by contacting with electron donor and 2 release by electronic repulsion. (c) Mechanism III: reduction by a chromate reductase enzymatic activity and release by electronic repulsion. Modified from Park et al. (2005).

Genes/proteins involved in DNA metabolism

The mutagenic effect of chromate in bacteria is well established (Stearns & Wetterhahn, 1997; Voitkun et al., 1998; Sugden et al., 2001; Salnikow & Zhitkovich, 2008). After being internalized, Cr(VI) is reduced by cellular reductants (i.e. cysteine, glutathione, and ascorbate) and several reductases to Cr(V), Cr(IV) and finally to the stable form Cr(III) (De Flora & Wetterhahn, 1989; Bose et al., 1992; Suzuki et al., 1992; Dillon et al., 1997; Stearns & Wetterhahn, 1997; Lay & Levina, 1998; Myers et al., 2000; Kalabegishvili et al., 2003; Ortega et al., 2005; Johannes et al., 2008). Cr(V) and Cr(IV) are reactive species that, as soon as formed, can react with DNA, inducing damage. Furthermore, it has been shown that Cr(V) can be oxidized to Cr(VI) by molecular oxygen in a cyclic process that induces the formation of a significant amount of ROS (Ackerley et al., 2004a). Cr(III) causes DNA damage and negatively affects DNA topology by inhibiting topoisomerase DNA relaxation activity (Plaper et al., 2002). These processes explain DNA alterations observed in cells exposed to chromate such as single-strand breaks, DNA–DNA interstrand links, DNA–protein cross-links, Cr-DNA adducts, nucleotide oxidation and abasic sites (Stearns & Wetterhahn, 1997; Voitkun et al., 1998; Sugden et al., 2001; Salnikow & Zhitkovich, 2008).

The SOS DNA repair system, initially described in E. coli (Radman, 1974), is triggered by the treatment of bacteria with DNA-damaging agents. The SOS system, which induces the arrest of DNA replication and cell division, involves more than 40 independent SOS genes, most of which encode proteins engaged in the protection, repair, replication, mutagenesis, and metabolism of DNA, and it is regulated by the transcriptional repressor LexA and the coprotease RecA, which aids the autocatalytic self-cleavage of LexA (Janion, 2008). Llagostera et al. (1986) evaluated the induction of the SOS system by chromium compounds in E. coli, as measured by beta-galactosidase activity, using lacZ gene fusions under the control region of different SOS genes. They demonstrated that Cr(VI) induces the SOS system. This finding is supported and enriched by the transcriptomic and proteomic studies, which have shown that the activation of the SOS system is involved in the immediate response of bacteria to chromate. In C. crescentus, lexA and recA were up-regulated 30 min after chromate or dichromate was added to the cultures (Hu et al., 2005). Similarly, in S. oneidensis MR-1, the up-regulation of lexA, recA, and additional genes involved in SOS response (recX, recN, dinP and umuD) was detected 5, 30, 60, and 90 min after the addition of chromate to the cultures (Brown et al., 2006). However, only two of these genes (dinP and recN) displayed a corresponding increase in the abundance levels of their expressed protein at both 45- and 90-min time points following chromate treatment. In C. metallidurans CH34 exposed to chromate for 30 min, the lexA gene was observed to be up-regulated, but the fold change did not reach the chosen cut-off of 2 (Monsieurs et al., 2011).

In cultures of P. aeruginosa, S. oneidensis, and Arthrobacter sp. FB24 adapted to grow in the presence of chromate, the up-regulation of genes involved in the SOS system was not observed. Although it cannot be excluded that this apparent lack of stimulation of the SOS genes might be due to the intrinsic limitations of transcriptomic and proteomic analysis, it is known that the induction of the SOS response proceeds for a limited period of time after the treatment of bacteria with DNA-damaging agents and then abruptly ceases (Janion, 2008).

Proteomic and transcriptomic studies of S. oneidensis and Arthrobacter sp. FB24 adapted to grow in the presence of chromate indicate that a large number of proteins up-regulated in Cr(VI)-stressed cultures are involved in DNA recombination, replication and repair (Chourey et al., 2006; Henne et al., 2009b), suggesting that bacteria growing under chromate stress need to make a special effort to maintain DNA integrity. Both S. oneidensis MR-1 and Arthrobacter sp. FB24 respond to long-term exposure to chromate via the up-regulation of DNA helicases, which catalyze the unwinding of the stable duplex DNA and play a pivotal role in all aspects of DNA metabolism and repair (Tuteja & Tuteja, 2004). In S. oneidensis MR-1 the up-regulation of three genes encoding helicases (SO0368, UvrD, and HrpA) was observed at the transcript level in cultures exposed to Cr(VI) for 24 h compared with untreated cultures. Similarly, in Arthrobacter sp. FB24, two UvrD/REP helicase domain proteins (Arth_2756 and Arth_2757) exhibited an approximately 20-fold increase in cultures adapted to grow in the presence of 5 and 20 mM Cr(VI) compared with untreated cultures. The involvement of helicases in chromate resistance has been demonstrated in Pseudomonas and O. tritici 5bv11 through the use of mutant strains. The inactivation of recG induces a severe decrease of Cr(VI)-MIC in P. aeruginosa and P. corrugata 28 (Miranda et al., 2005; Decorosi et al., 2009). The Q152 mutant of O. tritici 5bv11, harboring an interrupted ruvB helicase-encoding gene, was significantly more sensitive than the wild-type strain to chromate (Morais et al., 2011). All the above-mentioned studies suggest that DNA helicases are crucial to repair the DNA damage induced by chromate.

In S. oneidensis adapted to chromate, the DNA topoisomerase III (TopB) has been observed to be up-regulated (Chourey et al., 2006). This protein was also overexpressed in culture of S. oneidensis challenged with chromate for 90 min (Brown et al., 2006), suggesting that TopB plays a key role in chromate resistance both in the immediate and in the long-term response. DNA topoisomerases regulate the number of topological links between two DNA strands (i.e. change the number of superhelical turns) by catalyzing transient single- or double-strand breaks, crossing the strands through one another, then resealing the breaks to maintain the chromosome in a topological state commensurate with the particular replicative and transcriptional needs (Wang, 2002). The up-regulation of TopB in cells challenged with chromate likely reflects the need to deal with the alterations in DNA topology induced by chromate (Brown et al., 2006).

A proteomic analysis of P. aeruginosa did not reveal any differential production of enzymes involved in DNA metabolism in Cr(VI)-treated vs. untreated cultures (Kilic et al., 2010). A similar result was obtained with the proteomic analysis of P. putida F1 challenged with Cr(VI) for 75 min in LB medium (Thompson et al., 2010). Although these findings suggest that Cr(VI) exposure does not modify DNA metabolism in these two Pseudomonas strains, several factors may explain this unexpected result. The limitations intrinsic to proteomic analysis might leave some proteins experimentally undetected. Moreover, some enzymes involved in DNA metabolism playing a key role in chromate resistance might be of constitutive production and, as a consequence, would not be revealed as differentially expressed by transcriptomic and/or proteomic analysis. Lastly, the chromate concentration used in the work might not be sufficiently toxic to induce DNA damage. This hypothesis is supported by the finding that in P. putida F1, challenged with Cr(VI) in minimal medium (instead of in the complex LB medium), an ATPase involved in DNA repair-like protein (Pput 2963), was overexpressed compared with untreated cultures, suggesting that DNA damage occurs (Thompson et al., 2010). Minimal medium enhances the toxic effect of Cr(VI) compared with complex medium as proven by the great difference in the Cr(VI)-MICs of the strain evaluated in the two media (4–8 mM in LB medium and 0.1–0.3 mM in M9 medium; Thompson et al., 2010).

DNA metabolism processes have also been shown to play a central role in chromate responses in the yeast S. cerevisiae. For this model organism, an important available tool, which has been exploited to study yeast response to chemical stressors, is a heterozygous deletion strain collection, which allows the simultaneous testing of a large number of heterozygous yeast mutants in competition in the presence of a selected condition (Giaever et al., 1999). The heterozygous deletants are characterized by the reduction of the copy number of any given gene (from two copies to one) and thus it is also possible to analyze mutations in essential genes in competitive growth experiments. In these experiments, two possible phenotypes arise from data analysis: haploproficiency (increased growth rate when hemizygous) and haploinsufficiency (reduced growth rate when hemizygous). Using this tool, Holland et al. (2007) investigated over 6000 heterozygous deletants of S. cerevisiae in competitive growth in the presence of Cr(VI) to deepen the knowledge on molecular mechanisms that underline the toxicity of the metal. Two phenotypes were found. The former was characterized by a decrease of growth in the presence of Cr(VI) (tendency toward haploinsufficiency), the latter showed an improved relative growth (tendency toward haploproficiency) in the presence of Cr(VI). Overall, 115 strains exhibited a Cr(VI)-specific growth defect (indicating gene functions that normally protect against Cr), whereas the relative competitiveness of 203 strains was enhanced by Cr(VI) (indicating functions through which metal toxicity could be mediated). Among enriched gene ontology (GO) categories associated with haploinsufficiency, Holland et al. (2007) observed the establishment and/or maintenance of chromatin architecture, DNA packaging, and nucleotide excision repair, underlying the fact that these processes normally protect cells from Cr(VI)-induced damage. In the same work, the first enriched GO category found among the haploproficient genes was the SWI/SNF (SWItch/Sucrose NonFermenting) ATP-dependent chromatin remodeling complex (Carlson et al., 1981; Neigeborn & Carlson, 1984; Nasmyth & Shore, 1987), suggesting that chromatin state plays a role in mediating the toxic effect of chromate. It was observed that chromatin structure influences both gene expression and DNA accessibility not only for transcription and replication but also for repair systems (Ataian & Krebs, 2006; Osley et al., 2007; Czaja et al., 2012). Oxidative stress responses and DNA repair mechanisms are widely shared among eukaryotic cells from yeast to human (Sigler et al., 1999; Resnick & Cox, 2000; Marti et al., 2002; Poljsak et al., 2010). Cheng et al. (1998) demonstrated that in both yeast and mammalian cells, Cr(VI) induced similar patterns of DNA lesions (deletions, rearrangements, base substitutions). Schnekenburger et al. (2007) observed that in humans, Cr(VI) induces the formation of cross-links between histone deacetylase 1-DNA methyltransferase 1 (HDAC1-DNMT1) complexes and chromatin at the level of gene promoters, decreasing RNA polymerase recruitment. Studies on yeast strains bearing mutations in the genes involved in DNA repair indicated that all main repair pathways are involved in coping with chromate genotoxicity. The major form of Cr-DNA adducts in eukaryotic cells are Cr(III)-mediated DNA cross-links with glutathione, cysteine or ascorbate (ternary adducts). Many of the induced lesions are specific substrates for the nucleotide excision repair system (NER; O'Brien et al., 2002; Reynolds et al., 2004) and the mismatch repair system (MMR; Reynolds et al., 2009) (Fig. 5). It was observed that Msh2p–Msh6p dimers act as sensors for Cr-induced DNA damage, having a high affinity for the above-mentioned ternary adducts. A peculiar aspect of MMR coping with Cr-DNA crosslinks is the induction of double strand breaks (DSB) by endonucleases recruited by Msh3p at crosslink loci (Reynolds et al., 2009). Concerning the repair of DSB, mutants in genes involved both in homologous recombination (HR - Δrad52) and in non-homologous end joining (NHEJ - Δyku70, Δrad50, Δdnl4, Δmre11, Δxrs2) were more sensitive to chromate than the corresponding wild-type strains (Santoyo & Strathern, 2008; Zecevic et al., 2009). Furthermore, S. cerevisiae double mutants in HR and NHEJ (Δyku70/Δrad52, rad50/Δrad52, Δdnl4/Δrad52, Δmre11/Δrad52, Δxrs2/Δrad52) were more sensitive than any of the single mutants, indicating that both repair pathways are involved in the repair of Cr(VI)-induced lesions (Santoyo & Strathern, 2008).

Figure 5.

A model describing mismatch repair system-dependent (MMR) genotoxicity of Cr(VI) in eukaryotic cells (Marti et al., 2002; Reynolds et al., 2009). MMR coping with ternary adducts induced by Cr with DNA and intracellular reductants: image_n/fmr12051-gra-0001.png ascorbate, glutathione, or cysteine; Msh2/Msh6, Msh2/Msh3 (MutS complexes), responsible for damage recognition and binding; Mlh1/Pms1 (MutL complex that binds MutS/DNA complex, activating the downstream repair proteins); DSB, double strand breaks.

Genes/proteins involved in sulfur metabolism

Transcriptomic and proteomic studies have shown that a common bacterial mechanism to deal with chromate is the regulation of sulfate transport (Hu et al., 2005; Brown et al., 2006; Thompson et al., 2007, 2010; Henne et al., 2009b; Monsieurs et al., 2011). Chromate is chemically analogous to sulfate and in a variety of bacterial species, it enters the cells by means of sulfate ABC transporter (Cervantes et al., 2001; Aguilar-Barajas et al., 2011). The components of the sulfate ABC transporter (Sbp or CysP the periplasmic sulfate and thiosufate-binding proteins; CysT and CysW, the two inner-membrane transport proteins; CysA, the membrane-associated ATP-binding protein) are usually arranged in operons (Aguilar-Barajas et al., 2011).

In transcriptomic and proteomic studies of S. oneidensis MR-1, P. putida F1, C. metallidurans CH34, and Arthrobacter sp. FB24, the sulfate ABC transporter was up-regulated after chromate exposure (Brown et al., 2006; Thompson et al., 2007, 2010; Henne et al., 2009b; Monsieurs et al., 2011). A similar regulation of the sulfate ABC transporter was observed in Cr(VI)-stressed cultures of P. corrugata 28, a hyper-resistant-Cr(VI) strain (Viti et al., 2007). In this bacterium, sbp is located upstream cysTWA, but it forms an independent operon with oscA, a gene encoding a small protein involved in the utilization of organosulfur compounds (Viti et al., 2009). Both the transcriptional units oscA-sbp and cysTWA were overexpressed 5, 15, and 30 min after the addition of chromate to the cultures (Viti et al., 2009). The observed overexpression of the sulfate transporter after chromate exposure suggests that chromate induces sulfur starvation in cells. Two main factors have been evoked to explain why chromate induces sulfur limitation: (1) competition between sulfate and chromate for the transport reduces the bacterial capability to uptake sulfate, and (2) the oxidative stress induced by Cr(VI) decreases sulfur availability in cells (Brown et al., 2006). Proteomic analysis of Arthrobacter FB24 and P. putida F1 exposed to acute chromate stress indicates that these strains up-regulate proteins involved in the uptake and utilization of organic sulfur sources such as aliphatic sulfonates and cysteine (Henne et al., 2009b; Thompson et al., 2010). These results are consistent with the finding that a Cr(VI)-sensitive mutant of P. corrugata 28 lost the ability to grow on organic sulfur sources (i.e. taurine, methanesulfonic acid, buthanesulfonic acid), a phenotype depending on the activity of the above-mentioned oscA gene (Viti et al., 2009). The uptake and utilization of organic sulfur sources may be a mechanism used by bacteria to survive sulfur starvation induced by Cr(VI). The combined effect induced by sulfur compounds and different concentrations of sulfate in the medium on the Cr(VI)-MIC of P. corrugata 28 was evaluated by Decorosi (2010). Similar to what has been observed in other bacteria (Ohtake et al., 1987; Nies & Silver, 1989), the Cr(VI)-MIC of P. corrugata 28 is dependent on the concentration of sulfate, suggesting the competition of sulfate and chromate for uptake. The addition of cysteine or glutathione to the medium induced a strong increase of the Cr(VI)-MIC of P. corrugata 28 and made it independent of the concentration of sulfate. The addition of methionine to the medium increased the Cr(VI)-MIC that remained dependent on the sulfate concentration. Decorosi (2010) hypothesized that when cysteine or glutathione are available in the medium, the bacterium turns off the sulfate transport, blocking at the same time the uptake of chromate, and activates the transport of these sulfur compounds. In the presence of methionine, sulfate transport may be only partially reduced (Decorosi, 2010) (Fig. 6). Transcriptional studies aimed at evaluating the regulation of the genes involved in the transport of sulfur compounds are required to verify this hypothesis.

Figure 6.

Schematic model of the transport of sulfate, chromate, and organic sulfur sources in Pseudomonas corrugata 28 depending on the sulfate concentration, and presence/absence of organic sulfur sources in the growth medium (Decorosi, 2010). (a) High sulfate (large rectangle) – no organic sulfur sources; (b) low sulfate (small rectangle) – no organic sulfur sources; (c) high sulfate – cysteine; (d) low sulfate - cysteine; (e) high sulfate - methionine; (f) low sulfate - methionine. Chromate concentration is constant. The background color of the bacterial cell indicates the intracellular Cr(VI) concentration reached in the different medium compositions: image_n/fmr12051-gra-0002.png null; image_n/fmr12051-gra-0003.png, image_n/fmr12051-gra-0004.png intermediate; image_n/fmr12051-gra-0005.png high. The arrow size indicates the relative flux of each compound: image_n/fmr12051-gra-0006.png sulfate; image_n/fmr12051-gra-0007.png chromate; image_n/fmr12051-gra-0008.png cysteine; image_n/fmr12051-gra-0009.png methionine. The symbol image_n/fmr12051-gra-0010.png indicates blocking of the sulfate transporter.

Proteomic and/or transcriptomic studies of E. coli K-12, P. putida F1, S. oneidensis MR-1, C. metallidurans CH34, and Arthrobacter sp. FB24 have demonstrated that these bacteria respond to chromate stress by up-regulating the genes/proteins involved in cysteine biosynthesis (Fig. 7). ATP sulfurylase, which catalyzes the activation of sulfate to adenosine 5′-phosphosulfate (APS), was up-regulated at the gene or protein level in all the above-mentioned bacteria after chromate exposure (Ackerley et al., 2006; Brown et al., 2006; Chourey et al., 2006; Henne et al., 2009b; Thompson et al., 2010; Monsieurs et al., 2011). Furthermore, under acute chromate stress the APS kinase-encoding gene was up-regulated in cultures of S. oneidensis (Brown et al., 2006), and the APS reductase-encoding gene was up-regulated both in C. metallidurans CH34 and S. oneidensis MR-1 (Brown et al., 2006; Monsieurs et al., 2011). The 3′-phosphoadenosine-5′-phosphosulfate (PAPS) reductase has been observed to be induced in Arthrobacter sp. FB24 under chromate stress (Henne et al., 2009b), and sulfite reductase-encoding genes were observed to be up-regulated in S. oneidensis MR-1 under chronic or acute chromate stress (Brown et al., 2006; Chourey et al., 2006). The O-acetylserine-thiol-lyase is up-regulated in chromate-stressed E. coli K-12 (Ackerley et al., 2006) and S. oneidensis MR-1 (Thompson et al., 2007). In Arthrobacter sp. FB24, both O-acetylserine-thiol-lyase and serine acetyl transferase are up-regulated in chromate-challenged cultures (Henne et al., 2009b).

Figure 7.

Regulation of the sulfur metabolic pathway in bacteria stressed by Cr(VI). Genes/proteins regulated in following Cr(VI)-stressed bacteria studied through transcriptomic and proteomic approaches: Escherichia coli K-12 (Ackerley et al., 2006), Pseudomonas putida F1 (Thompson et al., 2010), Shewanella oneidensisMR-1 (Brown et al., 2006; Chourey et al., 2006; Thompson et al., 2007), Cupriavidus metalliduransCH34 (Monsieurs et al., 2011), Arthrobacter sp. FB24 (Henne et al., 2009b) and Caulobacter crescentusCB15N (Hu et al., 2005). The up-regulated and down-regulated genes/proteins are reported in red and green, respectively. The protein-encoding genes regulated in C. crescentus are marked with * because of the peculiarity of its metabolic response. CysND, ATP sulfurylase; CysC, adenosine 5′-phosphosulfate (APS) kinase; CysH, APS reductase; CysIJ, sulfite reductase; CysK, O-acetylserine (OAS)-thiol-lyase; CysE, serine acetyl transferase; GshA, glutamate-cysteine ligase; GshB, glutathione synthase; Gst, glutathione transferase; Cth, cystathionine gamma-lyase; MetC, cystathionine beta-lyase; CysO, serine sulfhydrase; MetE, cobalamin-independent methionine synthase; MetH, cobalamin-dependent methionine synthase; MetK, S-adenosylmethionine synthetase; MT, methyl-transferase; AhcY, adenosylhomocysteinase.

Glutathione is a key molecule in oxidative stress condition that helps to maintain cellular sulfhydryl groups in their reduced form (Jozefczak et al., 2012; Presnell et al., 2013). Glutathione and glutathione-dependent enzymes ensure correct folding, synthesis, and degradation of enzymes (Vuilleumier, 1997). Glutathione has emerged as a post-translational regulator of protein function under conditions of oxidative stress, by the direct modification of proteins via glutathionylation (Masip et al., 2006). Glutathione synthetase, the second enzyme of the glutathione biosynthesis pathway, and glutathione S-transferase are up-regulated in P. aeruginosa grown in medium supplemented with chromate (Kilic et al. 2010) (Fig. 7). One glutathione S-transferase-encoding gene was also induced in C. crescentus after chromate exposure, supporting the importance of glutathione in chromate resistance (Hu et al., 2005). In E. coli K-12, a significant depletion of glutathione and other free thiols, major bacterial antioxidants (Fahey, 2013), was observed after chromate exposure (Ackerley et al., 2006). Glutathione levels began to be replenished some hours later, coincidently with the reduction of oxidative stress and the induction of CysN and CysK, which are required for cysteine biosynthesis. Since glutathione is synthesized from cysteine (Copley & Dhillon, 2002), the up-regulation of cysteine biosynthesis after chromate exposure, shown by the above-mentioned transcriptomic and proteomic studies, may be required to produce enough substrate to sustain glutathione production (Thompson et al., 2010).

The transcriptional profile of sulfur metabolism in C. crescentus exposed to acute Cr(VI) stress is noticeably different from that observed in the other bacteria: the sulfate ABC transporter is down-regulated in Cr(VI)-stressed cultures of C. crescentus and up-regulated in the other bacteria studied (Hu et al., 2005). These authors suggested that the down-regulation of the sulfate transport could reduce the efflux of chromate into the cells. Dichromate exposure induces the down-regulation of the same transporter. Under the experimental conditions (pH 6–7), the equilibrium between dichromate and chromate was such that a significant amount of chromate was present in solution, and thus the response of bacterial cells to dichromate overlapped that of chromate (Hu et al., 2005). In addition, two enzymes involved in cysteine biosynthesis (ATP-sulfurylase and PAPS reductase) were down-regulated after Cr(VI) exposure, in C. crescentus. At the same time, the strong down-regulation of genes encoding proteins involved in methionine metabolism was observed (Hu et al., 2005). Genes encoding both enzymes catalyzing the biosynthesis of methionine from homocysteine (cobalamin-dependent methionine synthase I and cobalamin-independent methionine synthase II) and two enzymes (S-adenosylmethionine synthetase and adenosylhomocysteinase) involved in the methyl-activated cycle were down-regulated about 10- to 30-fold. The activated methyl cycle produces S-adenosylmethionine (SAM), an important methyl donor in biological systems (Weissbach & Brot, 1991; Parveen & Cornell, 2011) that is also involved in the biosynthesis of autoinducers, the quorum sensing chemical signals. The approximately sixfold up-regulation of glutathione S-transferase (CC2311) after exposure to Cr(VI) suggests that in C. crescentus glutathione is a key molecule for counteracting the stress induced by the metal. However, the Cr(VI)-induced down-regulation of genes involved in the biosynthesis of cysteine, the glutathione precursor, cannot immediately be explained. The observed depression of methionine synthesis could increase the availability of homocysteine that might be converted into cysteine through the trans-sulfuration pathway (Vermeij & Kertesz, 1999) (Fig. 7).

The genes involved in sulfur metabolism and GSH biosynthesis are part of the common fungal response to heavy metals and metalloids, being present in the induced CMR gene cluster and required for arsenite, cadmium, chromium, and copper tolerance (Thorsen et al., 2007; Jin et al., 2008). The fungi sulfur transporters Sul1p/Sul2p are the main proteins involved in high affinity chromate uptake by the cells, as identified in a screening of yeast strains resistant to chromate (Cherest et al., 1997); however, Pereira et al. (2008) have hypothesized that protein(s) other than Sul1p and Sul2p may be involved in Cr(VI) transport, such as other permease-transporting compounds structurally similar to chromate (i.e. molybdate or phosphate).

The use of the heterozygous deletion collection of S. cerevisiae indicated that chromate toxicity is normally mediated by processes associated with sulfur utilization, assimilation, and metabolism (Holland et al., 2007). Furthermore, the genes encoding proteins playing a role in the sulfur pathway and transport of sulfur-containing compounds (SUL1, SUL2, JLP1, DAL5, YCT1, BDS1, MMP1, CYS3, MET3, MET32, SEO1, MUP1, MET25, HGT1, MET2, CYS4, MET14, MHT1) were transcriptionally up-regulated upon treatment with toxic chromate concentrations in low sulfate medium and displayed a good correlation with corresponding protein levels (Pereira et al., 2008). In addition, Jin et al. (2008) observed that the GO categories associated with sulfur amino acid transport and biosynthesis were enriched in both transcriptionally up-regulated genes and genes required for growth (haploinsufficiency; CYS3, CYS4, SAM1) in the presence of Cr(VI). Nevertheless, intracellular Cr(VI) inhibits sulfate uptake and, as a consequence, induces sulfur starvation and reduces the sulfur metabolite pool (Thompson et al., 2007; Pereira et al., 2008). It has been shown that cells carry out a reprogramming of sulfur metabolism to redirect sulfur toward molecules promoting resistance mechanisms such as GSH (Fauchon et al., 2002; Thorsen et al., 2007). The so-called sulfur-sparing response, first highlighted as part of the cadmium response (Fauchon et al., 2002), consists of the repression of some sulfur-rich enzymes concomitant with the induction of sulfur-poor isoenzymes, which consequently increases the sulfur availability for GSH synthesis. The sulfur-sparing response and the up-regulation of the genes involved in sulfur metabolism in chromate-treated cells might be a consequence of sulfur depletion rather than a need to redirect sulfur toward incorporation in GSH, whose levels decrease in Cr(VI)-exposed cells (Wysocki & Tamas, 2010). Interestingly, both the sulfur-sparing response and GSH synthesis depend on the transactivator Met4p, whose regulon is formed of nearly 45 genes; the expression of the core genes requires Met31p or Met32p (Lee et al., 2010). MET32 has been observed among the up-regulated genes in microarray S. cerevisiae experiments (Pereira et al., 2008).

The role of GSH in response to chromate in fungi is controversial. In Schizosaccharomyces pombe, it was found that decreased intracellular GSH levels were associated with the Cr(VI)-sensitive phenotype of the chr-51S cells, indicating that GSH might effectively act against chromate by scavenging hydroxyl radicals (Pesti et al., 2002). Subsequent experiments in S. pombe revealed that a mutant with reduced GSH reductase activity was characterized by chromate tolerance as a consequence of a decrease of hydroxyl radical generation in Cr(VI)-treated cells. When the mutant was transformed with a vector carrying a glutathione reductase, it recovered the parental susceptibility level, losing the tolerant phenotype. This finding underlines the importance of the GSH reductase-NADPH system in Cr(VI) as a major one-electron donor for reduction reactions (Gazdag et al., 2003; Koosz et al., 2008). On the other hand, in S. cerevisiae GSH does not appear to be involved in the tolerance to chromium and other metals, although it could play a role in Cr(VI) uptake (Gharieb & Gadd, 2004).

The importance of sulfur metabolism in chromate response has also been observed in other yeast species. Some strains of Candida sp. passed from a Cr(VI)-tolerant to a Cr(VI)-sensitive phenotype depending on the supplementation of sulfur-containing compounds such as methionine, cysteine, djenkolic acid, and sulfate itself. When these strains grew with S-amino acids as sulfur sources they were more resistant to chromate than when grew in the presence of sulfate (Pepi & Baldi, 1992). The inhibition of sulfate uptake by djenkolic acid elicited chromate hypersensitivity in Rhodosporidium sp. This strain having an inefficient sulfate transport system was always resistant to Cr(VI) and assimilated sulfur in the form of S-amino acids (Pepi & Baldi, 1992). These authors showed that Cr(VI)-sensitive yeasts required larger amounts of S-amino acids, especially methionine, to tolerate Cr(VI) toxicity.

Also in filamentous fungi chromate tolerance resulted in association with sulfate transport. Aspergillus nidulans mutants in the sB gene encoding for sulfate permease were resistant to toxic analogs selenate and chromate in the presence of methionine (Arst, 1968); another sulfate transporter has been identified as a suppressor of a sulfate transport-negative mutant, named astA (alternative sulfate transporter), belonging to a poorly characterized family of allantoate permeases (Piłsyk et al., 2007). Selection of chromate-resistant mutants in Neurospora crassa led to the identification of two sulfate transporter encoding genes, CYS-13 and CYS-14 (Marzluf, 1970).

Genes/proteins involved in iron metabolism

Transcriptomic and proteomic studies of chromate bacterial response have revealed that iron homeostasis is one of the most important metabolic routes ruled by Cr(VI).

Iron, a transition metal that participates in many major biological processes, exists in the environment in one of the two readily interconvertible redox states: the reduced Fe(II) form and oxidized Fe(III) form, which is predominant and extremely insoluble. Microorganisms are able to scavenge Fe(III) efficiently from their surroundings via the secretion of high-affinity extracellular ferric chelators called siderophores. In many Gram-negative bacteria, uptake of ferric-siderophore complexes occurs in the periplasmic space through specific outer membrane (OM) receptors in a process driven by the inner membrane potential mediated by the energy-transducing TonB-ExbB-ExbD system (Andrews et al., 2003; Noinaj et al., 2010). Periplasmic binding proteins shuttle ferric siderophores from the OM receptors to the ABC transporters, which in turn deliver the ferric-siderophores to the cytosol, where the complexes are dissociated (Andrews et al., 2003; Braun & Hantke, 2011; Miethke, 2013). In Gram-positive bacteria, the iron uptake system is generally simpler than that observed in Gram-negative bacteria, and consists only of an ABC permease (Andrews et al., 2003; Klebba et al., 2012). Bacteria are also able to uptake iron complexes (transferrin, lactoferrin, heme, hemoglobin) by dedicated transporters instead of Fe (III) (Rouault, 2004; Haley & Skaar, 2012). Once inside the cell, iron is deposited in storage proteins (ferritins, heme-containing bactoferritins, Dps proteins) to provide a source of iron that can be drawn upon when external supplies are limited, and to maintain a low intracellular level of the metal, which can produce oxidative stress in aerobic conditions through the Fenton reaction (Andrews, 1998; Touati, 2000).

In a proteomic study of P. putida F1 exposed to acute Cr(VI) challenge, a TonB-dependent hemoglobin/transferrin/lactoferrin family receptor was up-regulated both in LB medium and M9 minimal medium. The study demonstrated that iron uptake is part of the ‘core’ response of P. putida F1 to Cr(VI) (Thompson et al., 2010). However, the bacterial need for iron appears to be definitely higher in LB medium than in minimal medium. In fact, whereas the proteins involved in iron homeostasis did not appear among the highest up-regulated proteins in Cr(VI)-shocked minimal medium-grown cells, the most frequently represented categories in the highest up-regulated proteome dataset of the LB-grown chromate shocked cells was related to iron homeostasis (i.e. TonB-dependent siderophore receptors, heme oxygenase, TonB transport protein ExbD; Thompson et al., 2010). In Cr(VI)-stressed cultures of P. putida grown in LB medium, five amino acid adenylation domain-harboring proteins were observed to be up-regulated. Since these proteins are likely involved in siderophore biosynthesis (Quadri et al., 1999; Quadri, 2000), Thompson et al. (2010) suggested that their up-regulation might also be connected to iron homeostasis.

Caulobacter crescentus exposed to chromate in minimal medium, significantly up-regulated one gene involved in iron homeostasis (Hu et al., 2005), a TonB-dependent OM receptor that is also responsive to other metals (cadmium and uranium). It is unlikely that the up-regulated TonB receptor interacts with the TonB protein (whose gene is suppressed under heavy metal stress), and thus it may not have a role in iron uptake but may instead bind to the metal and communicate extracellular environmental information (Hu et al., 2005).

Dynamic and dosage-dependent studies of the transcriptomic and proteomic response of S. oneidensis MR-1 to acute Cr(VI) challenge in LB medium have demonstrated a metabolic response similar to that observed in Cr(VI)-shocked P. putida F1 grown in the same complex medium. The majority of the iron uptake and sequestration genes/proteins were up-regulated, and some had the highest induction level (in some cases >100-fold) observed among the genes/proteins induced by chromate. A TonB1-ExbB1-ExbD1 complex, a ferric alcaligin siderophore receptor, a ferric vibriobactin receptor, a probable heme transport complex, a hemin ABC transporter, and a siderophore biosynthesis protein were substantially induced at gene and/or protein level following chromate exposure (Brown et al., 2006; Thompson et al., 2007). A study of a S. oneidensis Cr(VI)-sensitive mutant, harboring a deletion of the so2426 gene and deficient in siderophore production, demonstrated that the activation of many genes involved in iron uptake/storage and in the Cr(VI) acute response depended on the activity of the so2426 gene (Chourey et al., 2008), which is annotated as a DNA-binding response regulator.

Prolonged exposure of S. oneidensis to Cr(VI) elicits a very different molecular response from that observed in acute Cr(VI)-exposure studies. Iron sequestration and transport/binding genes were repressed or induced at low levels in response to 24-h Cr(VI)-exposure (Chourey et al., 2006), suggesting that S. oneidensis exposed to Cr(VI) has an immediate need to recruit iron that is reduced with time. Similarly, cultures of Arthrobacter sp. FB24 grown in LB medium amended with chromate up-regulated only two ferritin proteins (Dps family), but no proteins involved in iron uptake (Henne et al., 2009b). These authors suggested that the up-regulation of ferritin might lead to the sequestration of intracellular iron, which can enhance the oxidative stress induced by chromate.

Transcriptomic and proteomic studies of Cr(VI)-stressed bacteria suggested that a strong up-regulation of systems for iron uptake occurs particularly in strains grown in a complex medium after acute chromate challenge (Brown et al., 2006; Thompson et al., 2007), indicating that strains exposed to chromate and grown in a complex medium may have to cope with iron starvation.

Wang & Newton (1969), investigating an E. coli B/r mutant (strain B/rlt) sensitive to Cr(III) with a deletion in tpr-tonB, found that the mutant presented Cr(III) sensitivity when the medium was supplemented with an intermediate level of iron, but grew similarly to the wild-type when the medium was supplemented with a high excess of iron. On the basis of these results, the authors suggested that Cr(III) interferes with the uptake of iron from the environment. Current knowledge further supports these findings. The mutant B/rlt is likely not able to use the cytosolic membrane potential to take up the siderophore-iron complex. Wang & Newton (1969) isolated a second mutant, Chr2, which is sensitive to Cr(III) and has an impaired iron uptake. This mutant lacks the complete system to synthesize 2,3-dihydroxybenzoylserine (DHBS), a strong iron chelator that is currently known to be at the basis of the structure of the siderophore enterochelin (Young et al., 1971). As observed for B/rlt, the mutant Chr2 also presented reversion of Cr(III) sensitivity to the wild-type phenotype when iron was added at a high level to the medium. On the basis of these findings, the same authors suggested that Cr(III) appears to compete with ferric ion for entry into cells.

In S. cerevisiae, some pathways involved in the Cr(VI) response were associated with transport and metal ions homeostasis, resulting in both transcriptional up-regulation (Jin et al., 2008) and over-representation as functional categories displaying haploinsufficient or haploproficient phenotypes (Holland et al., 2007). In particular, the genes involved in transition metal and siderophore-iron transport processes (ARN1, FET3, ATX1, CCS1, FTR1, CCC2, ARN2, SIT1) were identified as important in the Cr(VI) response. Jin et al. (2008) hypothesized that the increased expression of these genes could contribute to tolerance by removing the excess of metals, also suggesting that exposure to elevated levels of one metal could interfere with the sensing and homeostasis balance of the others (Jin et al., 2008). Nevertheless, experiments performed with S. cerevisiae demonstrated that, following pretreatment with 0.1 mM Fe(III) ions at the start of the cultivation, there was cross-protection against subsequent exposure to sub-lethal doses of 2.5 mM Cr(III) (Fujs et al., 2007), suggesting an interplay between iron and chromium ions at the level of the cell wall or cell membrane. Pesti et al. (2002) proposed that Cr(III) cations preferentially localize in negatively charged regions of cell membrane, tending to create bridges between the phospholipid molecules and side-chains of the negatively charged amino acid residues. Fe(III) pre-treatment could alter the cell wall or membrane due to the chemical similarity between Fe(III) and Cr(III), resulting in decreased chromium accumulation in Fe(III)-pretreated cells (Fujs et al., 2007).

Other Cr(VI) resistance mechanisms

Additional mechanisms involved in Cr(VI) resistance and detoxification could involve the production of exopolysaccharides (EPS) (Kilic & Donmez, 2008; Harish et al., 2012). EPS composition varies from microorganism to microorganism. Pseudomonas aeruginosa produce alginate, a polymer of uronic, mannuronic, and guluronic acids, as a main polysaccharide in biofilms (Jain & Ohman, 1998). In a proteomic study MucD, which regulates the transcription of alginate biosynthesis genes in P. aeruginosa (Damron & Yu, 2011), was observed to be up-regulated in the presence of chromate (Kilic et al., 2010). Several genes involved in alginate metabolism were identified as differentially expressed by transcriptomic cDNA-microarray experiments (C. Viti, E. Marchi, F. Decorosi, L. Giovannetti, unpublished data) in a completely sequenced Pseudomonas alcaliphila 34 strain (Santopolo et al., 2013): the outer membrane protein AlgE, the alginate biosynthesis protein AlgX, and AlgJ, which is required for alginate O-acetylation. These findings suggest that the overproduction of EPS is a strategy against Cr(VI) stress in P. alcaliphila.

Several stress factors cause prophage induction in bacteria. In S. oneidensis MR-1, long-term Cr(VI) exposure leads to a strong induction of numerous prophage-related genes, suggesting that Cr(VI) exposure may induce the lytic cycle of lysogenic bacteriophages (Chourey et al., 2006), as previously reported for UV radiation (Qiu et al., 2005). More studies are needed to determine whether the exposure to Cr(VI) or the accumulation of Cr(III) inside the cells induces the lytic cycle; nevertheless, Chourey et al. (2006) concluded that prophage activation is a major contributor to the toxic effects of prolonged exposure to Cr(VI). Similar results were observed for Nitrosospira multiformis 25196. Cr(VI) caused a significant induction of the lytic cycle even at concentrations recommended as minimum inhibitory values, demonstrating that cell lysis due to prophage induction could strongly influence ammonia oxidation efficiency in an anoxic–aerobic activated sludge system (Choi et al., 2010).

From the analysis of a heterozygous deletant collection of S. cerevisiae, other classes of genes have emerged as being involved in the Cr(VI) response. The most significantly enriched category among haploinsufficiency-inducing genes in the presence of Cr(VI) was associated with actin/cytoskeleton organization, in particular with the cortical patches recognized as the sites of endocytosis (Holland et al., 2007). The authors first observed that actin-mediated endocytosis limited Cr(VI) accumulation inside the cell, and subsequent experiments confirmed that Cr(VI) exposure induced a marked stimulation of endocytosis, whereas null mutants Δsac6, Δend3, Δchc1 (initial internalization steps of endocytosis), and Δdoa4 (ubiquitin isopeptidase) exhibited increased Cr(VI) sensitivity associated with increased Cr(VI) intracellular accumulation (Holland & Avery, 2009). On the basis of these results it was hypothesized that the role of endocytosis is to mediate the internalization of a Cr(VI) transporter dependent on protein ubiquitination, and the failure of transporter internalization would elicit sensitivity to Cr(VI) because of its continuous entrance into the cells. This mechanism was previously observed for high-affinity transporters for zinc (Gitan et al., 1998) and iron (Felice et al., 2005).

Based on the genome-wide experiments on S. cerevisiae, a role for the vacuolar compartment was recognized as part of chromate tolerance. In addition to the first observation that vacuolar-defective mutants present increased chromium sensitivity (Gharieb & Gadd, 1998), the genes involved in vacuole biogenesis and organization were observed to be essential for viability at low Cr(VI) concentrations in experiments performed with the yeast deletome, a homozygous diploid gene deletion library. CUP5, PEP3, PPA1, VMA10/YHR039C-B, VMA13, VMA2, VMA21, VMA22, VMA4, VMA6, VMA7, VMA8, VMA9, VPH2, VPS16, and VPS34 were the genes involved in vacuolar organization and biogenesis whose deletion was detrimental for S. cerevisiae viability in the presence of chromate (Jin et al., 2008). The involvement of the vacuolar compartment was also underlined by Holland & Avery (2009), who observed that intracellular redistribution of a specific dye to the vacuolar membrane was more rapid in Cr(VI)-treated than Cr(VI)-untreated cells, indicating that Cr(VI) treatment increases membrane trafficking.

Jin et al. (2008) observed that knock-out mutants in genes involved trehalose biosynthesis were sensitive to Cr(VI). Trehalose is recognized as a stress protectant disaccharide in yeast (Eleutherio et al., 1995; Gancedo & Flores, 2004) and other living organisms, exerting its action by preventing protein denaturation, genetic material damage, and membrane disruption during stress (Yancey, 2005). It has been also observed that accumulated solutes such as trehalose or glycerol can contribute to the phenomenon of stress cross-protection (Lewis et al., 1995). Frederick et al. (2013) found that also in bacteria, trehalose might contribute to a more robust growth in the presence of Cr(VI) stress and to an increase in the reduction rate of extracellular chromate.

Other mechanisms to cope with chromate, in addition to those identified in S. cerevisiae, have been identified in fungi. It was observed that flavinogenic yeasts, which are able to increase the production of vitamin B2, and riboflavin (RF), as a consequence of iron deprivation (Pichia guilliermondii, Candida rhagii, Candida species, Candida famata, Debaryomyces klöckeri, Debaryomyces hansenii, Schwanniomyces occidentalis), when treated with sub-lethal concentrations of Cr(VI) are characterized by a strong increase in RF production, whereas non-flavinogenic yeasts did not show this response (Fedorovych et al., 2001). The intense flavinogenesis in response to Cr(VI) could represent a resistance mechanism eliciting increased cell survival, in accordance with the previous observation that RF supplementation reduces Cr(VI) toxicity in animal tissues (Appenroth et al., 1996) and that the addition of exogenous RF increases yeast cell viability in the presence of sub-lethal concentrations of chromate (Ksheminska et al., 2003). Interestingly, the flavinogenesis induced by Cr(VI) treatment is higher than that induced by iron deprivation (Fedorovych et al., 2001). However, the general mechanism of RF overproduction in yeast is common for both Fe-deficiency and Cr(VI)-exposure conditions, putatively involving the derepression of the enzymes of flavinogenesis pathway reactions, namely GTP-cyclohydrolase and RF-synthase (Fedorovych et al., 2001; Ksheminska et al., 2003).

Recent studies with Aspergillus have revealed an important role of fungal morphology in chromium tolerance (Das et al., 2008; Mishra & Malik, 2012). By atomic force (AFM) and transmission electron (TEM) microscopy, Das et al. (2008) evaluated changes of the cell surface properties along with the morphology of the fungal mycelia in Aspergillus versicolor exposed to chromate. These authors concluded that the cell wall components of the mycelia possess the major binding sites for chromate and their action prevents chromate transport into the cytoplasm, avoiding the contact of chromate with intracellular organelles. Moreover, the mechanisms hypothesized suggest that the reduction of chromate to Cr(III) by wall components, which are able to bind chromate ions electrostatically, leads to the formation of chromium multilayered structure on the cell wall (Fig. 8).

Figure 8.

Mechanisms of chromium multilayer formation in Aspergillus versicolor: (1) initial binding of chromate ions on the cell wall; (2) reduction to chromic hydroxide; (3) chromic hydroxide electrostatically binds chromate ions, resulting in multilayer formation. Modified from Das et al. (2008).

Cr(III) ions are very effective perturbants of the outer region of the membrane. It has also been observed in C. albicans that specific or nonspecific interactions of Cr(VI) ions with membrane components could contribute to the fluidity change of the membrane induced by Cr(VI) treatment, emphasizing that the effect of Cr(VI) on cell viability depends on the lipid composition of the membrane (Belagyi et al., 1999).

Outlook

We have examined and learned from studies which use classical and/or ‘omic’ approaches concerning how microorganisms counteract Cr(VI) toxicity (Fig. 9). Nevertheless, the knowledge is largely theoretical and we still lack a molecular understanding of many aspects of Cr(VI)–microorganism interactions.

Figure 9.

Interactions of Cr(VI) with the most important eukaryotic (a) and prokaryotic (b) model microorganisms.

Generally, both prokaryotic and eukaryotic microorganisms respond to Cr(VI) challenge by combining cellular networks acting at several levels, such as the reducing power generated by basal energy metabolism, iron and sulfur acquisition and homeostasis, protein oxidative stress protection, DNA repair, efflux pumps, detoxification enzymes. Microorganisms bearing chrA-encoding efflux pump orthologs demonstrate highly variable Cr(VI) resistance, indicating that the simple presence of the chrA gene cannot itself explain the chromate resistance in these strains. Reduction mechanisms of Cr(VI) to Cr(III) are widely present in bacterial and fungal species, but the biochemical properties of only a restricted number of Cr(VI) reductases have been elucidated. Several authors have indicated that the reduction of chromate is a secondary role for Cr(VI) reductases based on evolutionary features.

From genome-wide experiments on yeast, a new perspective has arisen, shifting the focus of Cr(VI) toxicity from a historically established mechanism based on DNA damage to a novel protein-centered response. On the other hand, this seems a good opportunity to re-examine and deepen our understanding of all studies that examine the mechanisms underlying the significant interplay that Fe and Cr ions exert on cellular homeostasis.

Global analysis of available studies indicates that careful attention should be paid to ‘omic’ data analysis because genes that are transcriptionally up-regulated are not always identified as differentially represented at the protein level. This incoherence between the transcriptomic and proteomic data might be explained by the incompleteness of proteomic datasets, due to the limited capability of the current proteomic technology to detect the whole cellular proteome, leaving experimentally undetected some proteins that are of low molecular weight, poorly represented or located in cellular districts that make their extraction difficult. Furthermore, it should not be forgotten that the cellular proteome at any given time reflects not only protein synthesis activity but also post-translational regulation, protein degradation and turn-over.

In conclusion, the existing ‘omic’ studies underline that Cr(VI) exposure appears to be associated with both general stress and specific Cr(VI) responses. Nevertheless, we believe that the use of mutants, which permits the validation at the cellular level of the results obtained with ‘omic’ approaches, will result in a better understanding of the complexity of cell responses to Cr(VI) toxicity.

Ancillary