As revealed by this study, which interrogated all nonessential genes of S. cerevisiae
for their role in modulating metal toxicity, 16 functional subnetworks (comprised of 207 genes, at least half of which had never been implicated in metal tolerance previously) negatively influence cadmium and/or nickel tolerance when disrupted. Core genes influencing cadmium and nickel tolerance were mapped to nine subnetworks, a subset of which (for example, V-ATPase, vacuole fusion, and the ERG pathway) cause enhanced co-sensitivity to mercury, arsenite, cobalt, zinc and iron, thus pointing to the occurrence of multimetal defense systems in yeast. Seven of these subnetworks were expanded to include additional mutations specifically associated with cadmium and/or nickel sensitivity, along with six additional subnetworks causing metal-specific (especially cadmium-specific) sensitivity (Table ). Only one of the latter subnetworks (with a bearing on the ALP branch of the Golgi-to-vacuole traffic pathway) was found to be specifically involved in nickel tolerance, as opposed to five subnetworks causing cadmium-specific sensitivity when disrupted. Thus, cadmium is not only more toxic than nickel, but it also has a broader spectrum of cellular processes that directly or indirectly contribute to its detoxification. Most prominent among these processes are those related to vesicular protein traffic (including the endocytotic pathway and a different branch of the Golgi-to-vacuole traffic), antioxidant defense, and DNA repair. The latter, in particular, further strengthens the causal relationship between cadmium genotoxicity and DNA repair [6
]. In fact, although cadmium and nickel have both been recognized as human carcinogens [2
], mutagenic activity appears to be a distinguishing feature of cadmium [1
]. Nickel, instead, is a weak mutagen with a marked nuclear tropism, whose carcinogenicity is thought to primarily rely on unprogrammed chromatin modification [5
]. It is interesting to note in this regard that the nuclear pore complex is one of the few core subnetworks enriched in nickel-specific sensitive mutants. Also interesting is that three out of eight mutants specifically resistant to nickel (but unrelated to vesicular traffic; Additional data file 7) are deleted in genes encoding distinct chromatin modification enzymes (HDA1
, and SPT7
) and one is deleted in a Ran homolog of the Ras GTPase family (MOG1
) that is involved in protein traffic through the nuclear pore.
Many metal toxicity-modulating pathways are related to metal damage prevention or repair, whereas others appear to play a more general (and indirect) role in promoting cell survival/recovery under stress conditions. Especially noteworthy among the latter are mRNA decay and nucleocytoplasmic transport, two processes that to our knowledge have not previously been implicated in metal tolerance and that might explain the variety of putative target genes previously identified as cadmium stress responsive [11
]. Their identification among metal sensitivity-conferring mutations suggests that not only the clearance of damaged (or unwanted) proteins by the proteasome and transcriptional regulation, but also mRNA turnover and relocalization are important for translational/metabolic reprogramming under conditions of metal stress. Interestingly, coordinate downregulation of iron-related proteins mediated by mRNA degradation under iron starvation conditions [95
] as well as mRNA mistranslation after chromium exposure [96
] have recently been described in yeast. The fact that structurally diverse yet functionally related gene products cause metal sensitivity when disrupted provides strong evidence that the cellular processes represented in specific subnetworks play an important role in preventing or repairing metal-induced cell damage. However, this does not exclude the possibility that a subset of the mutant strains that we have identified as metal sensitive are due to chemical-genetic synthetic lethality resulting from direct attack (and inactivation) of a functionally related protein target by the metal. Mutations associated with this kind of metal-induced lethality are likely to be enriched in the genes we classified as 'solitary'.
Among the unrelated stressors we examined, alkaline pH emerged as the most closely related to cadmium/nickel stress. This genomic phenotyping resemblance was traced back to iron deficiency, which - albeit with different mechanisms - is caused by both cadmium and nickel and appears to be a fairly general effect of metal toxicity (Figure ). Broad-range transporters were identified as the most proximal effectors of iron deficiency-related and other kinds of altered metal tolerance. The latter include Tat1, which is a low-affinity Trp/His transporter negatively regulated by Nrg1 [66
], which emerged from this study as one of the downstream effectors of the multimetal resistance caused by disruption of the Rim101 pathway. Toxic metal internalization (or abnormal intracellular mobilization) thus appears to be one of the most general and detrimental effects caused by transporter promiscuity, a trait that has probably evolved as a way to deal with multiple nutritional deficiencies under nutrient limiting (but toxic metal-free) conditions. This provides novel mechanistic support to the notion that nutrient limitation (especially iron and copper, but also amino acids and vitamins) may aggravate metal toxicity in malnourished human populations. Another outcome of this study was the identification of 24 uncharacterized ORFs that are involved in metal tolerance, which lend themselves as novel candidate genes that are worthy of further investigation.
Systematic comparison of the cellular toxicity signatures of cadmium and nickel with those of five additional metals revealed significant overlap between their chemical and cellular toxicity properties. However, it also uncovered an unexpected degree of metal specificity, especially regarding mutations that cause resistance to nickel but sensitivity to most other metals. The hot spots for such mutations were mapped to the ESCRT and the retromer complexes, thus pointing to the ability of these pathways to discriminate between otherwise similar metals and to the potential use of selected toxic metals (for example, cadmium and nickel) as chemical probes of intracellular traffic functionality.