|Home | About | Journals | Submit | Contact Us | Français|
We have used Saccharomyces cerevisiae to identify toxicologically important proteins and pathways involved in arsenic-induced toxicity and carcinogenicity in humans. We performed a systemic screen of the complete set of 4,733 haploid S. cerevisiae single gene deletion mutants to identify those that have decreased or increased growth, relative to wild-type, after exposure to sodium arsenite (NaAsO2). IC50 values for all mutants were determined to further validate our results. Ultimately we identified 248 mutants sensitive to arsenite and 5 mutants resistant to arsenite exposure. We analyzed the proteins corresponding to arsenite-sensitive mutants and determined that they belonged to functional categories that include protein binding, phosphate metabolism, vacuolar/lysosomal transport, protein targeting, sorting, and translocation, cell growth/morphogenesis, cell polarity and filament formation. Furthermore, these data were mapped onto a protein interactome to identify arsenite toxicity-modulating networks. These networks are associated with the cytoskeleton, ubiquitination, histone acetylation and the MAPK signaling pathway. Our studies have potential implications for understanding toxicity and carcinogenesis in arsenic-induced human conditions, such as cancer and aging.
Arsenic (As) is a ubiquitously present metalloid and a human carcinogen that is associated with skin, bladder, lung, kidney and liver cancer [1; 2; 3]. It is also implicated in vascular diseases, neurological and neurobehavioral disorders, diabetes and as a teratogen [4; 5]. Paradoxically, arsenic trioxide is currently used in the treatment of acute promyelocytic leukemia (APL) [6; 7]. Inorganic arsenic is considered the most hazardous among all the arsenic species present in the environment. Inorganic arsenic exists in the environment in two major forms, arsenite [As (III)] or arsenate [As (V)]. In general, As (III) is more acutely toxic than As (V) . Arsenic is a paradoxical non-mutagenic carcinogen, as there are arsenic-induced cancers observed in humans but there has been a lack of acceptable animal models. The mechanisms of arsenic-mediated toxicity and carcinogenesis are poorly understood, but it has been suggested that at least part of its toxicity is due to oxidative stress, which in turn causes protein denaturation, lipid damage and DNA strand breaks . Arsenic activates signal transduction pathways including AP-1, NFκB and MAPK (see review ). In addition, arsenic is involved in epigenetic mechanisms, for example, the alteration of epigenetic marks such as H3K4 tri-methylation and H3K9 di-methylation .
The availability of a complete set of single gene deletion Saccharomyces cerevisiae strains has allowed us to functionally characterize the yeast genes that respond to cellular insults at the systemic level [12; 13]. The complete yeast nucleotide sequence contains ~6,300 genes [14; 15], but only the 4,733 nonessential yeast mutants can be examined because deletion of the essential genes is lethal. Since there is a high degree of homology among the eukaryotes, S. cerevisiae can be used as a model to identify genes that might be important in arsenic-induced carcinogenesis in other eukaryotes, including humans.
Cells have developed adaptive defense systems against environmental stress, such as detoxification, repair, removal of damaged molecules. To better understand how cells respond to As (III) exposure, we screened the S. cerevisiae deletion strain set for sensitivity and resistance and identified the genes that have human homologues. In principle, genes whose deletion conferred sensitivity to arsenite would correspond to proteins involved in cellular recovery against arsenite-induced toxicity while genes whose deletion conferred resistance would correspond to proteins that arrest or reduce growth after arsenite exposure. We have analyzed the contribution of proteins that correspond to sensitive and resistant phenotypes in the framework of 12,232 protein-protein and protein-DNA interactions making up the known yeast interactome. Our results using this unbiased whole genomic approach reveal that genes whose deletion confers sensitivity to As (III) exposure correspond to proteins significantly enriched in various cellular functions, including vacuolar transport, cytoskeleton, acetylation and deacetylation processes, osmotic sensing and response, ubiquitination and proteosomal degradation, cell growth, regulation of carbon-compound and carbohydrate metabolism, protein binding, endocytosis, mitotic (M) phase, transport AT Pases, protein targeting, sorting and translocation, puine nucleotide/nucleoside/nucleobase anabolism, vacuole or lysosome function, homeostasis of protons, phosphate metabolism, stress response, budding, cell polarity and filament formation, cytoplasmic and nuclear protein degradation and MAPKKK cascade (Table 2). In contrast to 248 arsenite-toxicity sensitive proteins, only 5 arsenite-toxicity resistant proteins were identified (Mub1, Uth1, Fps1, Upf3, Ask10 and P15B12).
All yeast strains were grown in YPD (1% yeast extract, 2% peptone, 2% dextrose, 2% agar for plates) supplemented with 200 µg/ml G418. Sodium arsenite was purchased from sigma (St. Louis, MO). The complete set of 4,733 non-essential haploid S. cerevisiae single gene deletion mutants were obtained and described as before [17; 18].
High-throughput genomic screening was performed using the complete set of 4,733 mutants, as described before [17; 18]. Briefly, 96-well master plates containing individual deletion strains were resuspended with 60 µl bursts of forced air from a Hydra liquid handling apparatus(Robbins Scientific, Sunnyvale, CA), and then 1 µl samples were spotted on YPD agar plates containing 0, 0.75, and 1 mM sodium arsenite. Inoculated plates were incubated for 60 h at 30 °C and the resulting plates were imaged using an AlphaImager (Alpha Innotech Corporation, San Leandro, CA). The mutants were scored as sensitive or resistant compared with the non-treated and the wild type strain (BY4741). The experiments were done in triplicate.
5 µl of log-phase yeast culture was transferred into 195 µl YPD medium containing sodium arsenite in 96 well plates. The concentrations of sodium arsenite were 0, 0.375, 0.75, 1, 1.25 and 2.5 mM for sensitive strains and 0, 1.5, 2, 2.5, 3.75 and 5 mM for resistant strains. The cultures were incubated at 30 °C for 20 h, and cell density was determined by measuring the absorbance at 590 nm by Perkin Elmer HTS 7000 Bio Assay Reader. The concentration responsible for half-maximal inhibition of growth (IC50) was calculated using GraphPad Prism 5 program.
The deletion mutants that showed arsenic-sensitive phenotypes were categorized based on the biological functions using the program FunSpec (Functional Specification). The categories were downloaded from the MIPS Database and the GO Database. The p-values, calculated using the hypergeometric distribution, represent the probabilities that the intersection of a given list with any given functional category occurs by chance. Note that many genes are contained in many categories, especially in the MIPS database (which are hierarchical) and that this can create biases.
The deletion mutants were analyzed using the Cytoscope software for protein interaction networks as described . S. cerevisiae protein–protein interaction information were obtained from the Database of Interacting Proteins . In all we compiled 14,493 interactions between 5,433 proteins. The interactome is an extensive framework that can be used to identify protein networks activated by stress but it is a non saturated structure with regard to molecular interactions. None the less it provides a framework to analyze and associate discrete data points. Protein-protein interaction information was imported into Cytoscape for network visualization and subnetwork filtering. Subnetwork filtering was performed by tab selection of identified arsenite-toxicity modulation proteins and their associated protein-protein interactions. Analysis of the filtered interactome was set to identify sub networks => 4 nodes. It should be noted that interactome filtering does not use statistical validation to assign p-values to sub-networks. Instead, the filtering step identifies all connected As-toxicity modulating proteins in the interactome to provide a global view of how different functional activities are potentially coordinated.
Human homologues to the identified yeast genes were determined by BLAST using the tBLASTn program, which is available online from the National Center for Biotechnology Information (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi) . Note only the top scoring human homologue for each gene was used.
We screened a library of 4,733 S. cerevisiae gene deletion strains in triplicate to determine which proteins influence resistance or sensitivity after exposure to As (III). Strains from saturated cultures grown in 96-well format were robotically spotted onto agar plates with and without As (III) in the agar. Following 60 h incubation, plates were recorded by digital imaging of colony growth. By visual inspection of the imaged colonies, the strains were scored for sensitivity or resistance. The strains that had a significant decrease in the growth of colonies relative to wild type BY4741 were scored as sensitive, and the strains that displayed increased colony growth relative to wild-type were scored as resistant. Typical colony growth images were shown in Figure 1. In order to identify the possible pathways of arsenic response in humans, only the genes that have human homologues were listed (Table 1).
To identify genes whose deletion renders yeast most sensitive or resistant to arsenite toxicity, we determined the IC50 of the indentified mutants as well as the wild type strain. The degree of sensitivity or resistance of each gene was ordered based on the IC50 value (Table 1). The IC50 of the wild type strain for sodium arsenite is 4.47 mM. While most mutants are sensitive to As (III), a few of them (Mub1, Uth1, Fps1, Upf3, Ask10 and P15B12) are resistant to As (III) when compared to the IC50 of the wild type (Table 1).
Several studies have used S. cerevisiae as a tool to identify the molecules and cellular pathways linking arsenic induced toxicity and carcinogenicity. Nucleic acid metabolism, oxidative phosphorylation, protein synthesis and vacuolar acidification were involved in either arsenite sensitivity or resistance as determined by screening single gene knockout strains of S. cerevisiae in mitochondrial biogenesis and function . The strains whose deletion confers sensitivity to arsenic trioxide were found to be significantly enriched in the biological processes of osmoregulation, stress-related transcription regulation, cytoskeletal assembly and maintenance, signal transduction, DNA repair, oxidative stress, glutathione synthesis, secretory pathways and vacuole function, and general defense mechanisms . Here we analyzed the effect of arsenite on yeast single-gene deletion mutants. Our studies with sodium arsenite found many genes in common with those of studies done by others using arsenic trioxide, and yet additional genes whose deletion leads to sensitivity or resistance exclusively to arsenite were also identified in this study.
The strain most sensitive to sodium arsenite exposure, with an IC50 of 0.24 mM, lacks Pfd1 (Table 1). Pfd1 is subunit 1 of prefoldin, involved in the biogenesis of actin and of alpha- and gamma-tubulin, which are, in turn, important for cytoskeleton stability. A strain that lacks Gim4, which is prefoldin subunit 2 and a component of the Gim protein complex that promotes formation of functional alpha- and gamma-tubulin , was the second most sensitive strain to sodium arsenite exposure in the category of cytoskeleton assembly and maintenance. Thirteen other mutants that were sensitive to As (III), including Hsl7, Bem1, Ste50, Rvs161, Rpn4, Rvs167, Pac10, Sac1, Yke2, Tub3, Cla4, Svl3 and Nip100, correspond to cytoskeleton or structural proteins (Table 2). Cytoskeleton formation is important in establishing cell shape, providing mechanical strength, regulation of cell motility, chromosome separation in mitosis and meiosis, and intracellular transport of vesicles and protein complexes. Microtubules are one of the components of the cytoskeleton and are polymers of α- and β-tubulin dimers. Sodium arsenite directly interacts with the sulfhydryl-containing cysteine residues of tubulin, disrupting tubulin organization and microtubule assembly, and is proposed to induce aneuploidy in arsenite-treated human lymphocytes . Thus, the requirement of Pdf1 and Gim4 in synthesizing tubulins, as well as other proteins that maintain cell structure, should be important in protecting cells from arsenite-induced damage to the cytoskeleton.
Histone acetylation is associated with activation of gene expression and it also seems to be affected by arsenic. We have found that eleven of the S. cerevisiae strains sensitive to arsenite lack proteins involved in the acetylation or deacetylation process (Table 2), and these include Sgf29, Mak31, Ada2, Sgf73, Gcn5, Ard1, Hda1, Pho23, Hfi1, Mak3 and Hda3. Ada2, Gcn5 and Sgf29 are part of Spt-Ada-Gcn5 acetyltransferase(SAGA) complex which contains more than 20 subunits . The IC50 of Ada2, Gcn5 and Sgf29 deletion mutants are 0.375, 0.5 and 0.53, respectively. SAGA preferentially acetylates multiple lysine residues on the N-terminal tails of histone H3 and H2B , including acetylation of K9, K14, K18 and K23 of H3 . The component protein Gcn5 (general control nonderepressible 5) has histone acetyltransferase activity  and Ada2 potentiates Gcn5 acetyltransferase activity . SAGA regulates transcription of approximately 10% of the genome, most of which are upregulated in response to environmental stresses, including heat, oxidation, acidity, DNA damage, carbon or nitrogen starvation, and excess unfolded proteins .
Six of the S. cerevisiae strains sensitive to As (III) were missing genes whose corresponding proteins were involved in osmoregulation. These included Ste50, Doa4, PbsS2, Hog1, Nst1 and Ssk2. In yeast, cells respond to osmotic stress through a high-osmolarity glycerol (HOG1) pathway to maintain optimal cell volume and viability . In humans, the mitogen-activated protein kinase (MAPK) super-family consists of three major sets of kinases: the extracellular-receptor kinases (ERKs), the c-Jun N-terminal kinases/stress-activated protein kinases (JNK/SAPK), and the p38 MAPK. Hog1 is homologous to the p38 MAPK  and it activates its targets, including several transcription factors, which in turn activate genes devoted to osmoadaptation [33; 34; 35]. Osmotic stress activates Hog1 through the MAPKKK Ssk2 and the MAPKK Pbs2. Notably, HOG1 is the second most sensitive mutant with an IC50 of 0.31 mM (Table 1). Strains lacking Pbs2 and Ssk2 are also very sensitive to As (III), with IC50 of 0.5 mM and 0.55 mM, respectively (Table 1). In mammalian cells, the MAPK p38 pathway is activated by As (III) . Similarly, tolerance of fission yeast Schizosaccharomyces pombe to As (III) involves the MAPK Spc1, a homologue of mammalian p38 MAPK .
Vacuoles function to compartmentalize materials that may be harmful to cells. Glutathione-conjugated arsenic can be sequestered by Ycf1 in the vacuole, which contributes to cellular tolerance of arsenic . The proteins whose deletions confer sensitivity to As(III) in the category of vacuolar transport are Yps8, Stp22, Fen1, Cup5, Vps25, Vps24, Vps51, Snf7, Srn2, Pep3, Vps36, Vma6, Mvp1, Vps21, Vma4, Vts1, Snf8, Vps16 and Bro1(Table 2).
Removal of damaged molecules is a defense mechanism that maintains cellular and genetic integrity in response to environmental insults. Proteins are generally degraded by the ubiquitin (Ub)-mediated protein degradation pathway. Ub is conjugated to proteins by ubiquitin ligases. This tagging process leads to their recognition by the 26S proteasome, and ubiquitinated proteins are targeted to the 26S proteasome for degradation. Eight of the sensitive mutants lack genes whose corresponding proteins are involved in ubiquitination and deubiquitination, including Doa4, Ubc8, Ubp3, Rtt101, Grr1, Ubi4, Bre5 and Ubp2, and ten of the sensitive mutants lack proteins involved in proteosomal degradation, namely Rpn4, Ubc8, Bst1, Doc1, Pre9, Rpl40A, Rtt101, Grr1, Doa1 and Bro1(Table 2).
In the category of ubiquitination and proteosomal degradation, Grr1 displays a significant sensitivity to As (III) with an IC50 of 0.35 mM (Table 1). Similarly, Grr1 null yeast cells exhibit an elongated sausage-shape, and are sensitive to osmotic stress caused by ethylene glycol . Grr1 is an F-box protein and is part of the SCF ubiquitin ligase complexes . SCF consists of four proteins, Skp1, Cdc53/cullin, Rbx1/Roc1 and an F-box protein. The F-box protein functions as a substrate adaptor and mediates substrate specificity. Although it’s known that Grr1 is involved in glucose repression and that it targets the G1 cyclins Cln1 and Cln2 for degradation [39; 41; 42], the mechanism of As (III) induced-sensitivity of Grr1 mutants is unclear.
For a complete understanding of the toxicity induced by arsenite, it is important to study the function and regulation of uptake or secretory pathways. As(III) is transported into the cells through the aquaglyceroporin Fps1 . Deletion of Fps1 decreases As (III) influx into the cell and allows glycerol accumulation when cells are treated with As (III). Fps1 deletion mutants are resistant to As (III) toxicity (Table 2). Interestingly, the activity of Fps1 is modulated by Hog1 . Hog1 inactivates Fps1 by phosphorylation on T231 within the N-terminal domain of Fps1 .
Another of the most resistant strains lacks the yeast aging gene Uth1. It is a member of the family of yeast genes termed the “SUN family”. It is the first indentified gene providing a link between oxidative stress response, aging and mitochondria . It has been shown to interfere with mitochondria biogenesis and it is involved in the autophagic degradation of mitochondria [46; 47]. It is also required for Bax-induced cell death in yeast . Since arsenic induces oxidative stress, it’s very likely that Uth1 is important in mediating arsenic-induced toxicity through oxidative stress.
The genes whose inactivation led to arsenic sensitivity were analyzed for various cellular interactions. Using the Cytoscape software, protein-protein or protein-DNA interactions were analyzed. Toxicity modulating subnetworks consisting of greater than 3 connected nodes, corresponding to sensitive strains, are shown in Figure 2.
Figure 2 illustrates the engagement of various cellular processes aiding the recovery of S. cerevisiae from arsenite exposure; the processes embraced by each subnetwork are indicated in Table 2. All of the proteins in each network confer recovery of S. cerevisiae from arsenite exposure. Subnetwork (1) contains cytoskeleton/structural maintaining proteins (Pfd1), as well as vacuolar transport proteins (Snf7, Vma6, Vma4). Subnetwork (2) contains a group of proteins involved in ubiquitination (Grr1 and Ubi4). Subnetwork (2) also contains proteins involved in budding, cell polarity and filament formation during endocytosis (Rvs167, Rvs161 and Sla1) as well as a protein of unknown function (YBR284W). Subnetwork (3) contains Ste50, which encodes for a protein that is involved in mating response, invasive/filamentous growth, and osmotolerance. Subnetwork (4) contains components of ADA and SAGA histone acetyltransferase complexes (Ada2, Gcn5 and Sgf29). Histone acetylation is a modification mark of active gene transcription. Histone acetyltransferase complexes may provide resistance by participating in transcriptional activation of genes whose products aid recovery. Subnetwork (5) is dominated by proteins involved in the high osmolarity MAPK signaling pathway, Hog1 (MAPK), Pbs2 (MEK) and Ssk2 (MAPKKK).
Several studies have been done to screen the S. cerevisiae gene deletion strains to assess the role of nonessential proteins in modulating toxicity upon exposure to arsenic compounds. Haugen et al.  identified two metabolic networks, L-threonine and L-homoserine synthesis/degradation and the sikimate pathway, that are important for sodium arsenite tolerance. Jin et al.  have shown that the mutants engaged in S. cerevisiae toxicity to sodium arsenite functioning in processes of stress-related transcription regulation, tubulin folding, signal transduction, secretory pathway, and response to stimulus. Dilda et al.,  identified the sensitive mutant involved in the processes to include the high osmolarity glycerol stress signaling pathway, storage carbohydrate metabolism, DNA repair, oxidative stress defense, ergosterol biosynthesis, actin function, vacuolar acidification, secretory pathway function and NADPH biosynthesis. The focus of our study was to elucidate previously unidentified mechanisms and cellular pathways important for regulating the toxicity of arsenic in human cells, and restrict our studies to those sensitive and resistant strains whose gene deletion product has a human homologue. In this study, we have identified 248 arsenite-sensitive and 5 arsenite-resistant mutants by performing a genome-wide screen of genes in yeast. Functional categorization and interactome mapping suggests that cells develop multiple pathways to defend against arsenic-induced toxicity. In addition to the previously identified genes and pathways that confer sensitivity to arsenic, we have identified pathways of acetylation and deacetylation processes, cell growth/morphogenesis, endocytosis, M phase, protein targeting, sorting and translocation, purine nucleotide/nucleoside/nucleobase anabolism, homoestasis of protons, budding, cell polarity and filatment formation. This knowledge can be utilized to determine and understand the molecular and biological mechanisms by which arsenic induces toxicity. Future studies will determine if the identified genes control the activity of arsenic uptake or efflux, by measuring the concentration of arsenic in the mutant yeast cells after arsenic exposure, and if the human homologue of the yeast protein whose absence renders the cells either more sensitive or resistant to arsenite and therefore may have a direct role in the toxicity of arsenic compounds to human cells.
This work was supported by grant numbers ES014454, ES005512, ES000260 from the National Institutes of Environmental Health Sciences, grant number CA16087 from the National Cancer Institute (to M.C.), and grants T32 ES07324-08 and T32 NIEHS 007267-16 (to T.P.E.).