ToxSci Advance Access originally published online on September 4, 2007
Toxicological Sciences 2008 101(1):140-151; doi:10.1093/toxsci/kfm226
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Identification of Genes Involved in the Toxic Response of Saccharomyces cerevisiae against Iron and Copper Overload by Parallel Analysis of Deletion Mutants
,
,
* Department of Nutritional Sciences and Toxicology, University of California Berkeley, Berkeley, California 94720
Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720
Stanford Genome Technology Center, Palo Alto, California 94304
Donnelley Centre for Cellular and Biomolecular Research, University of Toronto, Toronto, Ontario M5S3E1, Canada
1 To whom correspondence should be addressed at Department of Nutritional Sciences and Toxicology, University of California, 317 Morgan Hall, Berkeley, CA 94720. Fax: (510) 642-0535. E-mail: vulpe{at}berkeley.edu.
Received March 28, 2007; accepted August 28, 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Iron and copper are essential nutrients for life as they are required for the function of many proteins but can be toxic if present in excess. Accumulation of these metals in the human body as a consequence of overload disorders and/or high environmental exposures has detrimental effects on health. The budding yeast Saccharomyces cerevisiae is an accepted cellular model for iron and copper metabolism in humans primarily because of the high degree of conservation between pathways and proteins involved. Here we report a systematic screen using yeast deletion mutants to identify genes involved in the toxic response to growth-inhibitory concentrations of iron and copper sulfate. We aimed to understand the cellular responses to toxic concentrations of these two metals by analyzing the different subnetworks and biological processes significantly enriched with these genes. Our results indicate the presence of two different detoxification pathways for iron and copper that converge toward the vacuole. The product of several of the identified genes in these pathways form molecular complexes that are conserved in mammals and include the retromer, endosomal sorting complex required for transport (ESCRT) and AP-3 complexes, suggesting that the mechanisms involved can be extrapolated to humans. Our data also suggest a disruption in ion homeostasis and, in particular, of iron after copper exposure. Moreover, the identification of treatment-specific genes associated with biological processes such as DNA double-strand break repair for iron and tryptophan biosynthesis for copper suggests differences in the mechanisms by which these two metals are toxic at high concentrations.
Key Words: metals; iron overload; copper overload; yeast; deletion mutant.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
The transition metals iron and copper are essential trace nutrients that participate as cofactors in redox reactions and are therefore important for the function of many proteins. Several major metabolic pathways, such as pyruvate metabolism, TCA cycle, and the respiratory chain, require proteins with copper- or iron-containing prosthetic groups (De Freitas et al., 2003
). However, this ability to readily accept and donate electrons makes iron and copper toxic when present in excess. Cells have evolved several mechanisms for maintaining adequate intracellular levels of these metals to carry out biological processes, in a state between deficiency and toxicity.
In humans, genetic disorders and/or environmental exposures can increase body iron and copper contents above normal levels and induce a variety of pathologies. For example, patients with Wilson's disease have impaired copper excretion through the bile due to mutations in the ATP7B gene, resulting in the accumulation of copper in organs and eventually causing neurological problems and liver disease. Very high copper concentrations (> 1 mg/gram of dry tissue) have been found in livers of patients with this disease (Faa et al., 1995). Mutations primarily in the HFE gene cause hereditary hemochromatosis, a condition where deregulation in iron absorption produces an excess accumulation of iron in tissues and induces systemic organ dysfunction. On the other hand, frequent blood transfusions, excess dietary iron, and accidental poisoning can be sources of environmental iron overload in humans. Diet constitutes a potential route of exposure to copper in the general population, although inhalation is a more common one among miners, copper smelters, and workers in other industries that utilize copper (Dorsey and Ingerman, 2004). Regardless of the origin of the overload, high levels of these metals are detrimental to human health. Both iron and copper, together with other transition metals, can participate in the Fenton reaction to generate highly reactive hydroxyl radicals (Kehrer, 2000). Induction of oxidative stress with subsequent damage to cellular macromolecules including DNA and proteins appears to be the main mechanism underlying the toxicity of these two metals (Fraga and Oteiza, 2002; Gaetke and Chow, 2003).
There is a relatively high degree of conservation in the metabolic pathways involved in iron and copper metabolism between humans and distantly related organisms such as the baker's yeast Saccharomyces cerevisiae. Most of the human genes in these pathways have homologous genes in yeast. The fact that the metabolism of these metals in yeast parallels the one in humans makes the former a suitable model to understand iron and copper metabolism and their disorders in humans (reviewed in Askwith and Kaplan, 1998).
Near all of the open reading frames (ORFs) in yeast have been systematically deleted to create a collection of mutant strains (Giaever et al., 2002). Molecular barcodes inserted at the site of the gene deletion uniquely identify each strain and allow the assay of all mutant strains in parallel for growth under any selective condition of interest (Winzeler et al., 1999). This approach constitutes a powerful tool for studying gene function in yeast and has been used to identify genes that are essential for growth under a variety of experimental conditions (Giaever et al., 2002; Lee et al., 2005), including high concentrations of arsenic (Haugen et al., 2004).
Although genome-wide systematic screens have been conducted in yeast to identify novel genes involved in the cellular responses to toxic levels of iron and copper, such studies have mainly focused on transcriptional profiling (Foury and Talibi, 2001; Gross et al., 2000). Functional profiling can provide additional insight as it can identify genes different from those identified by expression profiling (Birrell et al., 2002; Giaever et al., 2002). We conducted a parallel analysis of yeast homozygous diploid deletion mutants and identified a number of genes potentially involved in the response to iron and copper overload. This screen did not include essential genes as deletions of these genes result in a nonviable phenotype.
The majority of the identified genes were treatment specific, indicating different genetic requirements and cellular responses in yeast to these metals. Moreover, about half of these genes have at least one known homologous gene in human. We performed an enrichment and protein network analysis to uncover functional relationships between genes and to obtain an integrated view of the cellular responses to these conditions. In particular, biological processes and genes associated with intracellular trafficking and DNA repair that are conserved in mammals were significant in our results, suggesting that some of the mechanisms involved in the toxicity of iron and copper can be extrapolated to mammalian systems.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Functional profiling of the yeast genome.
Pool growth, genomic DNA extraction, barcode amplification, and hybridization were performed as previously described (Giaever et al., 2002
). Briefly, the complete set of 4757 homozygous diploid mutants BY4743 (MATa/MAT
his3
1/his31, ura30/ura30, leu20/leu20, lys20/+, met15 0/+) were pooled in rich media (yeast extract-peptone-dextrose, YPD) and grown at 30°C with shaking at 250–300 rpm in the presence of either 5mM iron sulfate, FeSO4.5H2O, or 10mM copper sulfate, CuSO4.7H2O. These exposure concentrations were selected based on growth inhibition experiments performed on the BY4743 wild-type strain (Supplementary Figure 9). Tolerance studies to these metals have not been conducted in this strain before. After 15 generations of growth, the surviving strains were collected and genomic DNA was extracted using the Qiagen DNeasy kit. The strain-specific barcodes contained in the DNA were amplified by PCR using a set of biotinylated primers that hybridize to universal sequences that flank the barcodes. Finally, these reactions were hybridized to Affymetrix DNA TAG3 arrays, stained, and scanned at an emission wavelength of 560 nm using an Affymetrix GeneChip scanner. For the control samples, pools were grown in YPD media as a nonselective condition and further processed in the same way as the treatments.
Outlier analysis for identification of differentially growing strains.
Hybridization intensities for the treatment and control arrays were quantified with the GCOS 1.2 software from Affymetrix. Data from each of the treatment arrays were paired with controls, and their quality was assessed by exploratory data analysis using box plots, quantile-quantile normal plot (QQNP), and scatter plots as described before (Armendariz et al., 2004).
In the Affymetryx TAG3 chips, each yeast ORF-knock out is represented by four features, two for each up-tag and down-tag barcodes, which produce four fluorescence signals after hybridization, and are replicate measurements of the growth of that particular deletion mutant. All four intensity signals associated with each individual ORF were averaged and used to calculate a log2 ratio of treatment to control for each deletion strain. The sensitive, unaffected, and resistant strains were identified by building simultaneous prediction intervals (SPIs) and by calculating a p value for each strain to statistically determine if there was a difference between the treatment and control, as previously described for the identification of differentially expressed genes in microarray experiments (Loguinov et al., 2004). Those strains that exhibited a significant change in growth, either as sensitive or resistant to the treatments, were deemed as differentially growing strains. The level of significance for a strain to be considered as such was set at p < 0.05.
Gene ontology enrichment and interactions network analysis.
Data sets were verified for enrichment for any particular biological attribute by identifying significantly enriched Gene ontology (GO) categories using the Functional Specification resource, FunSpec (http://www.funspec.med.utoronto.ca/). FunSpec uses a hypergeometric distribution to quantitatively assess functionally enriched GO categories after input of a list of genes of interest. This analysis was performed using a p value cutoff of 0.01 and without correcting for multiple comparisons. Additionally, yeast fitness data were mapped onto the extended Wi-Phi yeast interactome (Kiemer et al., 2007), consisting of > 10,000 protein-protein interactions, using Cytoscape version 2.4.1 (http://www.cytoscape.org). Significant subnetworks of activity were identified using the jActiveModules version 1.0 plug-in (Ideker et al., 2002), with default parameters. These subnetworks are connected regions in the yeast interactome that induced significant alterations in fitness under the conditions of interest. The resulting subnetworks were visualized in Cytoscape and selected for further biological interpretation based on their score and nonredundancy with other significant subnetworks.
Growth curve assay.
Individual deletion strains selected for phenotypic confirmation and the BY4743 wild-type strain were pregrown in YPD liquid media to mid-log phase, diluted to an optical density at 595 nm (OD595nm) of 0.0165, and seeded into different wells of a 48-well microplate. Stock solutions of either CuSO4.7H2O or FeSO4.5H2O were added to the treatment wells to the desired final concentrations with at least three replicates per strain/dose combination. Cells were then grown in a Tecan Genios spectrophotometer set to 30°C, intermittent shaking and OD595nm measurements at 15-min intervals for a period of 24 h. Raw absorbance data were averaged for all replicates, background corrected, and plotted as a function of time.
Spot assay.
Selected deletion strains together with the BY4743 wild-type strain were grown in YPD liquid media to mid-log phase. Cultures were diluted to OD595nm = 0.3 and transferred to a 96-well plate where fivefold serial dilutions were made. A volume of 5 µl from each dilution was used to spot onto Petri dishes containing either YPD agar or YPD agar with 10mM CuSO4.7H2O. Petri dishes were incubated for 2–3 days at 30°C to allow colony growth and subsequently scanned for evaluation.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Identification of Differentially Growing Strains
We treated pools of yeast homozygous diploid mutants with either 5mM FeSO4.5H2O or 10mM CuSO4.7H2O in duplicate exposures for 15 generations of cell growth. We used high doses of these metals, which lead to > 20% growth inhibition in the wild-type strain (see Materials and Methods), in order to identify all the deletion strains that exhibit a fitness alteration after exposure. For the controls, we randomly chose, from previously published data, three data sets corresponding to the same pools grown in YPD media for 15 generations (http://www.genomics.lbl.gov/YeastFitnessData/websitefiles/cel_index.html). We paired each individual treatment data set with all these three controls, resulting in a total of six treatment-control pairs for each metal, and analyzed separately as individual replicate experiments to account for the differences in strain growth due to biological and technical variability.
In the exploratory microarray data analysis (Fig. 1A and Supplementary Figure 8), inspection of the box plots showed that scale, location, number of outlying observations, and asymmetry were similar for each treatment-control pair. The average shifted histograms revealed that data from the treatment and control arrays resembled samples from univariate normal distributions, except for the presence of a lower second peak representing a subpopulation of probes in the array with low signal intensity. This was probably due to discrepancies between tag and probe sequences that affected hybridization efficiency and/or systematic experimental variability. Similarly, the QQNPs showed that the majority of points in our data conformed closely to a standard normal distribution except for the deviation previously observed at low signal intensities. The scatter plots of log2(treatment) versus log2(control) intensity values showed that the relationship between the intensities for all the treatment-control data set pairs, after normalization and log transformation, was approximately linear.
|
Since the exploratory analysis showed that the treatment-control data sets pairs for iron and copper were highly correlated with approximately symmetric distributions, we proceeded to identify the differentially growing strains. Graphical representation of the SPIs (Fig. 1B and Supplementary Figure 8) shows the data points outside the intervals as statistically significant outliers at different confidence levels, representing strains with a significant change in growth after treatment. To build the SPIs, we used robust scatter plot smoothers to account for heteroscedasticity, for example, variation in residual variance as a function of the intensity, in each data set.
We considered as outlier strains those with p < 0.05 in each individual treatment-control pair and as differentially growing strains after metal treatment those identified as outlier strains in all six treatment-control pairs associated with a specific treatment. We used this last criterion to reduce the number of false-positive results due to biological and/or technical variability. In this way, we identified a total of 74 and 200 differentially growing strains in iron and copper overload conditions, respectively. For each of these strains, we calculated a fitness score defined as the average of the log2(treatment) – log2(control) for all the six data sets specific to a treatment and then used this score to rank their corresponding ORFs (Table 1 and Supplementary Tables 4 and 5). This fitness score is an indicator of growth of the deletion mutants in the treatment relative to the control and estimates the contribution of each individual ORF in the adaptation of the cell to the conditions under study. Deletion of genes that possess an important function in the adaptation, protection, or repair following a toxic insult are likely to produce a sensitive phenotype and consequently a fitness score < 0. On the other hand, genes with fitness scores > 0 may indicate that the deletion is beneficial to the mutant for growth in the treatment condition. In addition, we used our list of genes to search for homologous ones in humans by using the NCBI BLASTP best hits from the Saccharomyces Genome Database website (ftp://genome-ftp.stanford.edu/pub/yeast/data_download/) and found that about 53% of the identified genes have at least one human homolog (Supplementary Tables 4 and 5).
|
Among the genes that induced differential growth upon deletion, were genes common and specific to the iron and copper overload treatments, with more than 80% of the gene deletions resulting in sensitivity of the deletion strain to either metal. Also, the majority of these genes had at least one reported catalytic activity, followed by a considerable number without known molecular function (Fig. 2).
|
Differences in Cellular Responses to Iron and Copper Toxicity
Both iron and copper overload identified yeast genes functioning in several GO biological processes (Tables 2 and 3). Specifically, processes associated with intracellular transport were significantly enriched with genes from these data sets. Although optimal growth in both iron and copper excess were dependent on vacuolar function–related processes, copper was also dependent on endosomal-related processes. Other significantly enriched biological processes were aromatic amino acid biosynthesis, particularly tryptophan biosynthesis, for copper and calcium ion homeostasis and response to DNA damage for iron (Supplementary Tables 6 and 7).
|
|
The protein network analysis uncovered significant subnetworks of activity within the yeast interactome that were essential for optimal growth after exposure to toxic doses of iron and copper (Fig. 3) and that were in agreement with the GO enrichment findings. Within these active subnetworks, we found protein complexes that may have a relation with detoxification under these conditions. High-scoring subnetworks in iron overload included proteins required for growth at high calcium concentrations (Fig. 3A) and Golgi-to-vacuole transport (Fig. 3B). In the latter case, the AP-3 complex appeared to mediate an important function as deletion of two of its subunits, Aps3p and Apl5p, induced sensitivity to iron. In the case of copper, high-scoring subnetworks contained proteins primarily involved in the endosome-to-Golgi retrograde transport (Fig. 3C). In particular, deletions of any of the members of the retrograde complex consisting of Pep8p, Vps29p, Vps5p, Vps35p, and Vps17p induced sensitivity to copper.
|
Genetic Requirements for Resistance to Iron and Copper Overload in Yeast
Deletion of a gene that has an important function in the adaptation to a selective condition may induce growth sensitivity in the mutant strain, given that there is no functional redundancy between the deleted gene and other genes in the genome. In our data sets, such genes have a fitness score < 0, indicating that the genes are required for resistance against iron and copper overload. We used growth curves to confirm individual strain growth under these two conditions as this was a simple and sensitive method. In agreement with the array data, the yeast strains containing deletions in CCC1, CSG2, HOR7, PHO80, RAD57, and VAM7 showed an expected sensitivity when exposed to lower concentrations of FeSO4.5H2O (Fig. 4), whereas strains with deletions in ACE2, CSG2, CUP2, FTR1, PHO80, PMP1, SUR1, TOP1, VPS3, and ZRT2 did the same for CuSO4.7H2O (Fig. 5).
|
|
Analysis of the growth curves for individual mutants grown in toxic conditions revealed different types of responses. The sensitivity of the deletion strains to the metal treatments was observed as an increase in the length of the lag phase, a decrease in the growth rate, or a lower plateau phase, when compared to the same strains grown in rich media (Fig. 5). A sharply reduced culture density at the plateau phase, such as with pho80
, is an indication that the deleted gene plays an essential role in the toxic response and that its absence cannot be compensated by other mechanisms, leading to premature cell death. On the other hand, an extended lag phase or reduced growth rate, such as with csg2 and ace2, suggests that while the mutants have an increased sensitivity to the toxic metal, it is not lethal and can be compensated by other cellular mechanisms.
An opposite case to the sensitive mutants above was exhibited by aft1, which was found to be resistant to FeSO4.5H2O (Fig. 4). Aft1p induces the expression of genes involved in the cellular uptake of iron in response to low iron conditions and its absence in the deletion strain may induce iron deficiency and suboptimal growth, a defect that is compensated when the mutants are grown in iron-replete media.
As indicated by the enrichment of GO categories, genes encoding products associated with intracellular trafficking and vacuolar function are highly required for growth under toxic concentrations of iron and copper. In the case of copper, we identified a total of 55 genes associated with known vacuolar sorting mutants (Bonangelino et al., 2002), corresponding to 27.5% of the identified genes. In a similar way, iron identified genes in these pathways but to a lesser extent (14 genes or 18.9%).
Some other identified genes with fitness scores < 0 that were indicative of the stress and damage induced by these metals were RAD57, MMS1, YKU70, TOP1, SGS1 (DNA repair), ISC1, SUR1, CSG2, CCC1, HOR7, SRO7 (cation homeostasis), DSK2 and YGK3 (protein degradation) for iron and NTG2, RAD55, TOP1 (DNA repair), VID27 (protein degradation), SUR1, CSG2, MID1, ISC1, NHX1, SPF1, FTR1, ZRT2, HOR7 (cation homeostasis), PIN4 and OCA1 (DNA damage-induced cell cycle arrest) for copper.
Tryptophan Requirement in Yeast for Growth under Copper Overload
We found four genes in the tryptophan biosynthetic pathway with low fitness scores in the copper data set, namely TRP1, TRP2, TRP4, and ARO2, and further confirmed their requirement for optimal growth (Fig. 6). We looked at growth sensitivity of the corresponding deletion strains when exposed to copper by using both spot and growth curve assays and found a higher sensitivity of the latter to detect differences in growth. In the growth curve assay, OD595nm readings are taken throughout a period of 24 h to measure cell growth, as opposed to the spot assay where the ability to growth is an endpoint measurement. For strains that have a slight or intermediate fitness defect under a selective condition, as is the case of the tryptophan auxotrophic strains grown in copper overload, an endpoint measurement leaves enough time for the inhibited cells to grow to saturation and decreases any marked differences from the controls.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
We have screened the collection of yeast homozygous diploid deletion mutants for strains that exhibited a significant alteration in growth after exposure to toxic concentrations of iron and copper sulfate and identified genes involved in the response to these toxic, yet essential, metals. In order to obtain a global perspective of the specific processes involved in the response against iron- and copper-induced toxicity, we identified biological processes significantly enriched with genes from our data sets, as well as significant protein subnetworks, and found that the cellular responses of yeast include mainly protective, homeostatic, and repair mechanisms.
The majority of the genes that we identified have been previously associated with vacuolar sorting–defective phenotypes (Bonangelino et al., 2002). In yeast, free metal ions are sequestered into vacuoles (Bode et al., 1995), preventing damage to cellular constituents, and is probably the main reason why deletion strains with impaired vacuolar function showed sensitivity in our results. This provides evidence that vacuoles constitute a common mechanism of protection against iron and copper. However, the subnetwork scoring and GO enrichment analysis indicated the presence of two different detoxification pathways that target iron and copper separately to the vacuole (Fig. 7). Indeed, free intracellular iron is transported directly into vacuoles, whereas copper is transported into late endosomes (Li et al., 2001; Yuan et al., 1997), which eventually form vacuoles. We found that deletion of the vacuolar iron transporter CCC1 resulted in sensitivity to iron and was supportive of a role for the vacuole and Ccc1p in iron detoxification. Since vacuoles can also be formed from the Golgi via an alternative endosome-independent pathway, genes associated with endosomal function were not essential for growth in iron overload. On the other hand, copper resistance required several components functioning in the endosomal pathway, providing evidence that copper is detoxified to the vacuole via endosomes in yeast.
|
Intracellular protein trafficking is a conserved process among eukaryotic organisms, including humans. For example, the vacuoles and endosomes in yeast are analogous to the lysosomes and late endosomes/multivesicular bodies found in mammals, respectively. Likewise, the AP-3, endosomal sorting complex required for transport (ESCRT)-II, ESCRTIII, and retromer complexes that we identified as essential for growth in iron or copper (Fig. 7) share homology with human complexes. The mammalian AP-3 is ubiquitously expressed in all cell types and involved in the sorting of endosomal proteins to the lysosome. It is essential for the proper function of organelles such as melanosomes and its disruption is linked to the Hermansky-Pudlak syndrome type 2 in humans (Stinchcombe et al., 2004
). The mammalian ESCRT complexes are involved in the trafficking of ubiquitinated proteins from endosomes to lysosomes (Williams and Urbe, 2007). Similarly, the putative human retromer complex is thought to participate in retrograde transport in the same manner as its yeast counterpart (Haft et al., 2000). To the best of our knowledge, the relationship between these complexes to iron and copper detoxification in mammals is unknown. Evidence from mammalian systems have shown that excess copper is sequestered into late endosomes by the Wilson's disease protein ATP7B, transported into lysosomes, and excreted into the bile by exocytosis (Harada et al., 2000; Schaefer et al., 1999). Under copper excess conditions, the mammalian ESCRT and retromer complexes may somehow be required for proper function of endosomes and therefore essential in copper detoxification.
A number of strains with deletions in genes associated with DNA double-strand break repair showed sensitivity to 5mM FeSO4.5H2O but not to 10mM CuSO4.7H2O, suggesting that either iron is more potent than copper in producing this type of damage or that copper does not produce it. This is consistent with previous reports that iron, but not copper, contributes to the formation of single-strand breaks induced by hydrogen peroxide (Barbouti et al., 2001). These differences in producing DNA damage can be due to additional mechanisms of defense against copper, such as metallothioneins. In fact, deletion of CUP2 encoding a transcription factor that induces the expression of the metallothionein genes resulted in a very sensitive phenotype to copper (Fig. 5). Among the genes in our data sets significantly required for growth in iron were TOP1 and SGS1, homologous to the human topoisomerase TOP1MT and DNA helicase RECQL, respectively, and which products interact with each other to maintain genome stability. Mutations in RECQL have been associated with the Bloom and Werner syndromes in humans. Other required genes were ESC2, MMS1, and YKU70 (homologous to XRCC6), all of which have been found to genetically interact with SGS1 (Pan et al., 2004; Yamana et al., 2005), suggesting that they may act in conjunction to repair damaged DNA.
Ion homeostasis was a biological process that we found significantly enriched for both iron and copper overload and that included CSG2 and SUR1, which products are required for growth under high calcium concentrations. In the case of iron, we also identified these proteins as part of a significant subnetwork (Fig. 3A). The requirement for these genes could be an indication of a disturbance in calcium homeostasis in exposed cells. Indeed, calcium levels are increased in cells undergoing oxidative stress and correlate with DNA damage (Nicotera et al., 1988). Oxidative stress can cause an influx of extracellular calcium and induce the opening of gated channels in the endoplasmic reticulum to release stored calcium, resulting in increased intracellular levels that can be detrimental to the cell (Ermak and Davies, 2002).
In yeast, the iron permease FTR1 is upregulated after exposure to high copper (Gross et al., 2000), indicating a cellular requirement for iron, as this gene is expressed under low iron conditions. In agreement with this result, we found that ftr1 was highly sensitive to copper. In Hep-G2 cells, copper overload decreases iron levels (Arredondo et al., 2004). An explanation for these observations is that, as cells attempt to detoxify copper by upregulating protection mechanisms, less copper is made available for high-affinity iron uptake, resulting in iron deficiency. In mammalian cells, under steady-state copper concentrations, ATP7B is located in the trans-Golgi network, transporting copper ions from the cytoplasm which are incorporated into ceruloplasmin (ferroxidase). However, under copper excess conditions, ATP7B translocates to endosomes to mediate biliary copper detoxification (Cater et al., 2006; Schaefer et al., 1999). This translocation is reversible upon return to normal copper concentrations. However, under chronic exposures, ATP7B could be permanently localized in endosomes, affecting copper loading onto ceruloplasmin in the trans-Golgi. Taken together, evidence suggests that copper overload may have an indirect effect on intracellular iron pools by inducing iron deficiency in yeast and that similar mechanisms operate in higher eukaryotes.
The fact that several genes in the tryptophan biosynthetic pathway were essential for optimal growth in copper overload suggests an important unknown role of tryptophan in yeast. In humans, tryptophan is an essential amino acid. The role of tryptophan in the response to copper-induced toxicity in yeast may be via the antioxidant properties of its metabolites produced during degradation in the kynurenine pathway (Christen et al., 1990) or its radical-scavenging activity, as superoxide radical is used as a cofactor to cleave the pyrrole ring in tryptophan (Hayaishi et al., 1977). Alternatively, tryptophan may be required as a critical residue in certain proteins involved in the defense against copper toxicity. At toxic doses, copper affects membrane integrity and increases permeability to ions, resulting in an increase in the activity of the plasma membrane H+-ATPase Pma1p (Fernandes and Sa-Correia, 2001). This pump is regulated by Pmp1p, a low molecular weight proteolipid that contains in its C-terminal domain single Phe38, Trp28, and Tyr25 residues that appear to act together to anchor the protein in the membrane and thus essential for protein function (Mousson et al., 2001). These three amino acids are synthesized from chorismate, a product of Aro2p. We found that both aro2 and pmp1 were sensitive to copper. Interestingly, both PMA1 and PMP1 have been found to be translationally upregulated after treatment with butanol (Smirnova et al., 2005), a chemical that also induces proton leakage across the plasma membrane. Indeed, the copper-sensitive strains trp1, trp2, trp4, aro2, and pmp1 showed sensitivity to 0.5% butanol (data not shown). Another protein with a critical tryptophan residue is the copper-zinc superoxide dismutase (Yamakura et al., 2001), which constitutes an important antioxidant mechanism. We were not able to evaluate the phenotype of sod1 as this mutant exhibits growth defect under aerobic conditions.
We have analyzed the growth patterns of several thousand yeast mutants to identify genes required for optimal growth under iron and copper overload and to gain understanding of their effects at the cellular level. Ultimately, this information could be applied in the study of the metabolism of these two metals in humans, particularly under overload conditions, as many of the genes that we identified have a human homolog. We showed that several of these human genes participate in conserved biological processes with yeast and therefore could play a role in sensitivity and/or resistance to iron and copper toxicity in people. Lastly, the parallel assay of deletion mutants provides a simple and rapid way to systematically screen for genes associated with basic cellular responses to toxic insults that can be exploited in toxicology.
|
SUPPLEMENTARY DATA |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Supplementary figures 8 and 9 and tables 4 and 5 are available online at http://toxsci.oxfordjournals.org/.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONSUPPLEMENTARY DATAREFERENCES |
|---|
Armendariz AD, Gonzalez M, Loguinov AV, Vulpe CD. Gene expression profiling in chronic copper overload reveals upregulation of Prnp and App. Physiol. Genomics (2004) 101(1):140–151.
Arredondo M, Cambiazo V, Tapia L, Gonzalez-Aguero M, Nunez MT, Uauy R, Gonzalez M. Copper overload affects copper and iron metabolism in Hep-G2 cells. Am. J. Physiol. Gastrointest. Liver Physiol. (2004) 287(1):G27–G32.
Askwith C, Kaplan J. Iron and copper transport in yeast and its relevance to human disease. Trends Biochem. Sci. (1998) 23(4):135–138.[CrossRef][Web of Science][Medline]
Barbouti A, Doulias PT, Zhu BZ, Frei B, Galaris D. Intracellular iron, but not copper, plays a critical role in hydrogen peroxide-induced DNA damage. Free Radic. Biol. Med. (2001) 31(4):490–498.[CrossRef][Web of Science][Medline]
Birrell GW, Brown JA, Wu HI, Giaever G, Chu AM, Davis RW, Brown JM. Transcriptional response of Saccharomyces cerevisiae to DNA-damaging agents does not identify the genes that protect against these agents. Proc. Natl. Acad. Sci. U.S.A. (2002) 99(13):8778–8783.
Bode HP, Dumschat M, Garotti S, Fuhrmann GF. Iron sequestration by the yeast vacuole. A study with vacuolar mutants of Saccharomyces cerevisiae. Eur. J. Biochem. (1995) 228(2):337. (Abstract).[Web of Science][Medline]
Bonangelino CJ, Chavez EM, Bonifacino JS. Genomic screen for vacuolar protein sorting genes in Saccharomyces cerevisiae. Mol. Biol. Cell (2002) 13(7):2486–2501.
Cater MA, La Fontaine S, Shield K, Deal Y, Mercer JF. ATP7B mediates vesicular sequestration of copper: Insight into biliary copper excretion. Gastroenterology (2006) 130(2):493–506.[CrossRef]
Christen S, Peterhans E, Stocker R. Antioxidant activities of some tryptophan metabolites: Possible implication for inflammatory diseases. Proc. Natl. Acad. Sci. U.S.A. (1990) 87(7):2506–2510.
De Freitas J, Wintz H, Kim JH, Poynton H, Fox T, Vulpe C. Yeast, a model organism for iron and copper metabolism studies. Biometals (2003) 16(1):185–197.[CrossRef][Web of Science][Medline]
Dorsey A, Ingerman L. Potential for human exposure. In: In Toxicological Profile for Copper (2004) 1st ed. Agency for Toxic Substances and Disease Registry. 121–191. Available at http://www.atsdr.cdc.gov/toxprofiles/tp132.pdf.
Ermak G, Davies KJ. Calcium and oxidative stress: From cell signaling to cell death. Mol. Immunol. (2002) 38(10):713–721.[CrossRef][Web of Science][Medline]
Faa G, Nurchi V, Demelia L, Ambu R, Parodo G, Congiu T, Sciot R, Van Eyken P, Silvagni R, Crisponi G. Uneven hepatic copper distribution in Wilson's disease. J. Hepatol. (1995) 22(3):303–308.[CrossRef][Web of Science][Medline]
Fernandes AR, Sa-Correia I. The activity of plasma membrane H(+)-ATPase is strongly stimulated during Saccharomyces cerevisiae adaptation to growth under high copper stress, accompanying intracellular acidification. Yeast (2001) 18(6):511–521.[CrossRef][Web of Science][Medline]
Foury F, Talibi D. Mitochondrial control of iron homeostasis. A genome wide analysis of gene expression in a yeast frataxin-deficient strain. J. Biol. Chem. (2001) 276(11):7762–7768.
Fraga CG, Oteiza PI. Iron toxicity and antioxidant nutrients. Toxicology (2002) 180:23–32.[CrossRef][Web of Science][Medline]
Gaetke LM, Chow CK. Copper toxicity, oxidative stress, and antioxidant nutrients. Toxicology (2003) 189:147–163.[CrossRef][Web of Science][Medline]
Giaever G, Chu AM, Ni L, Connelly C, Riles L, Veronneau S, Dow S, Lucau-Danila A, Anderson K, Andre B, et al. Functional profiling of the Saccharomyces cerevisiae genome. Nature (2002) 418(6896):387–391.[CrossRef][Medline]
Gross C, Kelleher M, Iyer VR, Brown PO, Winge DR. Identification of the copper regulon in Saccharomyces cerevisiae by DNA microarrays. J. Biol. Chem. (2000) 275(41):32310–32316.
Haft CR, de la Luz Sierra M, Bafford R, Lesniak MA, Barr VA, Taylor SI. Human orthologs of yeast vacuolar protein sorting proteins Vps26, 29, and 35: Assembly into multimeric complexes. Mol. Biol. Cell. (2000) 11(12):4105–4. 016.
Harada M, Sakisaka S, Terada K, Kimura R, Kawaguchi T, Koga H, Taniguchi E, Sasatomi K, Miura N, Suganuma T, et al. Role of ATP7B in biliary copper excretion in a human hepatoma cell line and normal rat hepatocytes. Gastroenterology (2000) 118(5):921–928.[CrossRef][Web of Science][Medline]
Haugen AC, Kelley R, Collins JB, Tucker CJ, Deng C, Afshari CA, Brown JM, Ideker T, Van Houten B. Integrating phenotypic and expression profiles to map arsenic-response networks. Genome Biol. (2004) 5(12):R95.[CrossRef][Medline]
Hayaishi O, Hirata F, Ohnishi T, Henry JP, Rosenthal I, Katoh A. Indoleamine 2,3-dioxygenase: Incorporation of 18O2—and 18O2 into the reaction products. Biol. Chem. (1977) 252(10):3548–3550.
Ideker T, Ozier O, Schwikowski B, Siegel AF. Discovering regulatory and signaling circuits in molecular interaction networks. Bioinformatics (2002) 18(Suppl. 1):S233–S2240.[Abstract]
Kehrer JP. The Haber-Weiss reaction and mechanisms of toxicity. Toxicology (2000) 149:43–50.[CrossRef][Web of Science][Medline]
Kiemer L, Costa S, Ueffing M, Cesareni G. WI-PHI: A weighted yeast interactome enriched for direct physical interactions. Proteomics (2007) 7(6):932–943.[CrossRef][Web of Science][Medline]
Lee W, St Onge RP, Proctor M, Flaherty P, Jordan MI, Arkin AP, Davis RW, Nislow C, Giaever G. Genome-wide requirements for resistance to functionally distinct DNA damaging agents. PLoS Genet. (2005) 1(2):e24.[Medline]
Li L, Chen OS, McVey Ward D, Kaplan J. CCC1 is a transporter that mediates vacuolar iron storage in yeast. J. Biol. Chem. (2001) 276(31):29515–29519.
Loguinov AV, Mian IS, Vulpe CD. Exploratory differential gene expression analysis in microarray experiments with no or limited replication. Genome Biol. (2004) 5(3):R18.[CrossRef][Medline]
Mousson F, Beswick V, Coic YM, Baleux F, Huynh-Dinh T, Sanson A, Neumann JM. Concerted influence of key amino acids on the lipid binding properties of a single-spanning membrane protein: NMR and mutational analysis. Biochemistry (2001) 40:9993–10000.[CrossRef][Medline]
Nicotera P, McConkey D, Svensson SA, Bellomo G, Orrenius S. Correlation between cytosolic Ca2+ concentration and cytotoxicity in hepatocytes exposed to oxidative stress. Toxicology (1988) 52:55–63.[CrossRef][Web of Science][Medline]
Pan X, Yuan DS, Xiang D, Wang X, Sookhai-Mahadeo S, Bader JS, Hieter P, Spencer F, Boeke JD. A robust toolkit for functional profiling of the yeast genome. Mol. Cell (2004) 16(3):487–496.[CrossRef][Web of Science][Medline]
Schaefer M, Hopkins RG, Failla ML, Gitlin JD. Hepatocyte-specific localization and copper-dependent trafficking of the Wilson's disease protein in the liver. Am. J. Physiol. (1999) 276(3 Pt 1):G639–G646.[Web of Science][Medline]
Smirnova JB, Selley JN, Sanchez-Cabo F, Carroll K, Eddy AA, McCarthy JE, Hubbard SJ, Pavitt GD, Grant CM, Ashe MP. Global gene expression profiling reveals widespread yet distinctive translational responses to different eukaryotic translation initiation factor 2B-targeting stress pathways. Mol. Cell. Biol. (2005) 25(21):9340–9349.
Stinchcombe J, Bossi G, Griffiths GM. Linking albinism and immunity: The secrets of secretory lysosomes. Science (2004) 305(5680):55–59.
Williams RL, Urbe S. The emerging shape of the ESCRT machinery. Nat. Rev. Mol. Cell Biol. (2007) 8(5):355–368.[CrossRef][Web of Science][Medline]
Winzeler EA, Shoemaker DD, Astromoff A, Liang H, Anderson K, Andre B, Bangham R, Benito R, Boeke JD, Bussey H, et al. Functional characterization of the S. cerevisiae genome by gene deletion and parallel analysis. Science (1999) 285(5429):901–906.
Yamakura F, Matsumoto T, Fujimura T, Taka H, Murayama K, Imai T, Uchida K. Modification of a single tryptophan residue in human Cu,Zn-superoxide dismutase by peroxynitrite in the presence of bicarbonate. Biochim. Biophys. Acta (2001) 1548(1):38–46.[CrossRef][Medline]
Yamana Y, Maeda T, Ohba H, Usui T, Ogawa HI, Kusano K. Regulation of homologous integration in yeast by the DNA repair proteins Ku70 and RecQ. Mol. Genet. Genomics (2005) 273(2):167–176.[CrossRef][Web of Science][Medline]
Yuan DS, Dancis A, Klausner RD. Restriction of copper export in Saccharomyces cerevisiae to a late Golgi or post-Golgi compartment in the secretory pathway. J. Biol. Chem. (1997) 272(41):25787–25793.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  T. C. Sideri, S. A. Willetts, and S. V. Avery Methionine sulphoxide reductases protect iron-sulphur clusters from oxidative inactivation in yeast Microbiology, February 1, 2009; 155(2): 612 - 623. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. J. Kennedy, A. A. Vashisht, K.-L. Hoe, D.-U. Kim, H.-O. Park, J. Hayles, and P. Russell A Genome-Wide Screen of Genes Involved in Cadmium Tolerance in Schizosaccharomyces pombe Toxicol. Sci., November 1, 2008; 106(1): 124 - 139. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. Gonzalez, A. Reyes-Jara, M. Suazo, W. J Jo, and C. Vulpe Expression of copper-related genes in response to copper load Am. J. Clinical Nutrition, September 1, 2008; 88(3): 830S - 834S. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||