Article pubs.acs.org/jpr
Systematic Proteomic Approach to Characterize the Impacts of Chemical Interactions on Protein and Cytotoxicity Responses to Metal Mixture Exposures Yue Ge,*,† Maribel Bruno,† Kathleen Wallace,† Sharon Leavitt,† Debora Andrews,† Maria A. Spassova,‡ Mingyu Xi,∥ Anindya Roy,∥ Najwa Haykal-Coates,† William Lefew,† Adam Swank,† Witold M. Winnik,† Chao Chen,‡ Jonne Woodard,† Aimen Farraj,† Kevin Y. Teichman,§ and Jeffrey A. Ross† †
National Health and Environmental Effects Research Laboratory, U.S. Environmental Protection Agency, 109 T. W. Alexander Drive, Research Triangle Park, North Carolina 27709, United States ‡ National Center for Environmental Assessment and §Office of Research and Development, U.S. Environmental Protection Agency, 1200 Pennsylvania Avenue, N. W., Washington, DC 20460, United States ∥ Department of Statistics, University of Maryland, Mathematics Building, College Park, Maryland 20742-4015, United States S Supporting Information *
ABSTRACT: Chemical interactions have posed a big challenge in toxicity characterization and human health risk assessment of environmental mixtures. To characterize the impacts of chemical interactions on protein and cytotoxicity responses to environmental mixtures, we established a systems biology approach integrating proteomics, bioinformatics, statistics, and computational toxicology to measure expression or phosphorylation levels of 21 critical toxicity pathway regulators and 445 downstream proteins in human BEAS-2B cells treated with 4 concentrations of nickel, 2 concentrations each of cadmium and chromium, as well as 12 defined binary and 8 defined ternary mixtures of these metals in vitro. Multivariate statistical analysis and mathematical modeling of the metal-mediated proteomic response patterns showed a high correlation between changes in protein expression or phosphorylation and cellular toxic responses to both individual metals and metal mixtures. Of the identified correlated proteins, only a small set of proteins including HIF-1α is likely to be responsible for selective cytotoxic responses to different metals and metals mixtures. Furthermore, support vector machine learning was utilized to computationally predict protein responses to uncharacterized metal mixtures using experimentally generated protein response profiles corresponding to known metal mixtures. This study provides a novel proteomic approach for characterization and prediction of toxicities of metal and other chemical mixtures. KEYWORDS: chemical interactions, dose response, mixture toxicity, chemical mixtures, risk assessment, proteomics, cytotoxicity, metals, systems biology
■
exposure to more than one metal or chemical.2 Research data on the effects of metal interactions on molecular and cellular toxic responses and their correlations in a multiple exposure or mixture can help fill data gaps and improve the available models and approaches that are currently used in risk assessments of metal and other chemical mixtures. The dilemma of lack of scientific information and research tools versus perception of high risk from exposures to mixtures by the exposed population poses an enormous challenge for the risk assessment community. To date, very few biology systems and technologies are available and suitable to address all
INTRODUCTION Metal contamination is a serious environmental concern as metals are generally nonbiodegradable, highly toxic, and frequently present in air, water, soil, and biota.1 Toxic metals such as nickel, cadmium, and chromium are lung carcinogens, included on the EPA’s list of priority pollutants, and found in tobacco smoke, diesel and jet fuel emissions, industrial wastes, metal ore and fumes, abandoned mine drainage, natural sources (volcanic ash and soil), and contaminated water.1 Humans are exposed to these metals mainly by inhalation, and the inhalations result in a variety of adverse health problems, particularly of the respiratory system such as acute lung injury caused by respiratory epithelial cell damage and loss of function.1,2 While the toxicity of each of these metals has been well-described, real-world exposures involve simultaneous © 2014 American Chemical Society
Special Issue: Environmental Impact on Health Received: August 1, 2014 Published: October 6, 2014 183
dx.doi.org/10.1021/pr500795d | J. Proteome Res. 2015, 14, 183−192
Journal of Proteome Research
Article
Table 1. List of 28 Single Metals and Metal Mixtures, Their Concentrations, and Cytotoxicity Levelsa
a
treatment group
Ni2+ (μM)
Cd2+ (μM)
Cr6+ (μM)
group label
cytotoxicity classification
relative cytotoxicity
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29
0 30 60 75 100 0 0 0 0 30 30 60 60 30 30 60 60 0 0 0 0 30 30 30 30 60 60 60 60
0 0 0 0 0 5.6 7.5 0 0 5.6 7.5 5.6 7.5 0 0 0 0 5.6 5.6 7.5 7.5 5.6 5.6 7.5 7.5 5.6 5.6 7.5 7.5
0 0 0 0 0 0 0 0.3 1.8 0 0 0 0 0.3 1.8 0.3 1.8 0.3 1.8 0.3 1.8 0.3 1.8 0.3 1.8 0.3 1.8 0.3 1.8
control Ni30 Ni60 Ni75 Ni100 Cd5.6 Cd7.5 Cr0.3 Cr1.8 Ni30Cd5.6 Ni30Cd7.5 Ni60Cd5.6 Ni60Cd7.5 Ni30Cr0.3 Ni30Cr1.8 Ni60Cr0.3 Ni60Cr1.8 Cd5.6Cr0.3 Cd5.6Cr1.8 Cd7.5Cr0.3 Cd7.5Cr1.8 Ni30Cd5.6Cr0.3 Ni30Cd5.6Cr1.8 Ni30Cd7.5Cr0.3 Ni30Cd7.5Cr1.8 Ni60Cd5.6Cr0.3 Ni60Cd5.6Cr1.8 Ni60Cd7.5Cr0.3 Ni60Cd7.5Cr1.8
N L L M H N N N M N L L M L M L H N L N M M L M M M H M H
1.000 0.9388 0.8815 0.7253 0.5208 1.1494 1.1353 1.0423 0.7703 0.9975 0.9446 0.8352 0.7473 0.9635 0.6845 0.8435 0.4355 1.1222 0.8132 1.0006 0.7641 0.7594 0.8344 0.7490 0.6801 0.7106 0.4170 0.6590 0.4030
N: noncytotoxicity, L: low cytotoxicity, M: moderate toxicity, and H: high cytotoxicity.
from both single and mixture exposures in these genomic studies at gene expression levels,10 it is still not clear how metal interactions and mixture concentrations affect protein or pathway responses and thereby cellular toxicity responses. To address these issues, we measured expression or phosphorylation levels of 21 key regulators involved in metal toxicity pathways using multiplexed ELISA and 454 downstream proteins identified by 2-D gel electrophoresis (2-DE), and mass spectrometry (MS) in human airway epithelial cells treated with mixtures of nickel (II) chloride (Ni2+), cadmium(II) chloride (Cd2+), and potassium dichromate (Cr6+). Highthroughput and quantitative proteomics is a powerful approach for identifying protein changes/responses/signatures/patterns associated with mixtures and individual components within mixtures. Effects of metal interactions on protein responses and relevant biochemical/toxicity pathways were characterized by qualitatively and quantitatively comparing protein response profiles associated with both metal mixtures and individual mixture components. The identified protein responses were then correlated with cellular toxicity, including cell viability, apoptosis, and energy metabolism observed in the treated BEAS-2B cells in vitro to establish qualitative and quantitative relationships between exposures to metal mixtures including individual mixture components and concentrations and protein and cytotoxicity responses. The central aim of this study was to develop and utilize innovative proteomic methods including screening individual metals and the mixtures of the metals for distinct unique exposure and toxicity signatures of mixtures and each chemical in the mixtures and for potential points of
interactions among complex mixtures at molecular, cellular, and organ levels, which are critical to the risk assessment of chemical mixtures. Therefore, there is an important need to develop novel and biology-focused methodologies and approaches for efficient analysis of environment toxicity pathways, biomarkers, and toxic mechanisms associated with exposure to metal mixtures and for accurate distinction of chemicals in the mixture that present little or no concern from those with the greatest likelihood of causing an adverse effect in the target species. High-throughput and high-content proteomic methods especially applied to predictive toxicology provide opportunities for addressing these challenges. When a human cell is exposed to metal mixtures, the protein profile including protein expressions, post-translational modifications, and activities in the exposed cell usually changes in response to the exposure.3 Changes in the human cell proteome or protein responses are the molecular basis of cellular toxic responses and usually precede before the presence of cellular toxicity end points and phenotypic changes resulting from the exposures. Therefore, health states of the exposed human cells are largely determined by the expression, posttranslational modification, and activity levels of thousands of proteins. Many proteins involved in cell growth, apoptosis, oxidative stress, and inflammation, including p53, hypoxia inducible factor 1 alpha (HIF-1α), and NFκB, are differentially modulated by exposure to metals and metal mixtures,4,5 and the mixtures induce unique protein or cytotoxicity changes that are not observed from single metal exposures.6−9 Because of the lack of overlapped protein responses or pathways identified 184
dx.doi.org/10.1021/pr500795d | J. Proteome Res. 2015, 14, 183−192
Journal of Proteome Research
Article
at room temperature in 96-well plate according to manufacturer’s instructions. After this, the wells containing the fixed cells were washed with Binding Buffer (1XPBS/1%BSA) and then aspirated. F-actin in cells was then stained with 50 μL of Alexa Fluor 488 Phalloidin after a 20 min of incubation at room temperature and were then rinsed with 100 μL of 1× PBS.
biologic/toxic conversion such as shared and critical protein and toxicity pathway activations to enhance the identification and exploration of potential interactions of metals in a mixture. This study, therefore, provides a systematic and efficient proteomic approach to characterize chemical interactions and mixture toxicity to support risk assessment of environmental mixtures.
■
Isolation of Total Proteins from BEAS-2B Cells for 2-DE
After the removal of the control and treated BEAS-2B cells were from the incubator, 10 mL of ice-cold washing buffer (250 mM sucrose/10 mM Tris solution, pH 7.4) was added to the cell-cultured dishes. Cells were then scraped off from the dishes, transferred to 15 mL conical tubes, and centrifuged at 1300g for 5 min at 4 °C. Cells were washed three times by suspending in ice-cold washing buffer, centrifuging at 1300g for 10 min, and then lysed in 0.5 mL of 2-DE sample buffer containing 30 mM Tris (pH 8.8), 2 M urea, 7 M thiourea, 4% CHAPS (w/v), 0.5% Triton X100, and 10 μL of protease/ phosphatase inhibitor cocktail. The dissolved samples were sonicated for 10 s at a constant duty cycle of 20%, with a 30 s interval between the sonication. The cell lysates were then placed on ice for 1 h and were then centrifuged for 5 min at 14 000g. The supernatants containing total proteins were used for 2-DE analysis. The protein concentration was determined using the 2D Quant kit from GE Healthcare.
MATERIAL AND METHODS
Materials
BEAS-2B cells were purchased from American Type Culture Collection (Rockville, MD). Trypsin Gold (MS-grade) was obtained from Promega (Madison, WI). Dithiothreitol (DTT) was purchased from Bioanalytical (Natick, MA). CyDyes (Cy3, Cy2, and Cy5), Typhoon 9410 scanner, nonlinear immobilized pH gradient (IPG) strips (pH 3−11), Ettan IPGPhor apparatus, Decyder software, and 2D Image Quant (version 5.1) were manufactured by GE Healthcare (Piscataway, NJ). Methanol (HPLC grade) and glacial acetic acid were purchased from Fisher Scientific (Fair Lawn, NJ). Protean II apparatus was a product from Bio-Rad (Hercules, CA). IPA Software is a product from Ingenuity Systems (Redwood City, CA). LHC-9 medium, HEPES buffer, saline solution (HBSS), and trypsinneutralizing solution (TNS) were obtained from Lonza (Walkersville, MD). Versene, TrypLE, Dulbecco’s phosphate buffer solution (DPBS), and Sypro Ruby stain were obtained from Life Technologies (Carlsbad, CA). Precast 8−16% gradient bisacrylamide gels were from Jule (Milford, CT). The AKT, Human Proinflammatory-9 Plex, HIF1α, apoptosis, MAPK, and EGFR ELISA kits for measurement of protein changes at expression and phosphorylation levels in the treated BEAS-2B cells were from Meso Scale Discovery (Gaithersburg, MD). ELISA kits for validation of protein changes of ATP synthase, enolase 1, pyruvate dehydrogenase, SOD2, and catalase that were identified by 2-DE gel electrophoresis were purchased from Abcam (Cambridge, MA). Trypan blue, neutral red (NR), and all other chemicals were obtained from SigmaAldrich (St. Louis, MO).
Protein Labeling, 2-DE, and MS Analysis
To determine how protein changes may actually lead to the cytotoxic outcome and toxicity pathway alterations, we also analyzed protein expression profiles in the treated BEAS-2B cells using DIGE as previously described.11 Briefly, each CyDye was mixed with 50 μg of protein sample, and the pooled protein samples were diluted further in the sample buffer (7 M urea, 2 M thiourea, 2% w/v CHAPS, 1% IPG, 13 mM DTT) to the volume of 340 μL for 2-DE analysis. For the 2-DE analysis, triplicate gels from control and treated protein sample were run. First-dimension IEF was carried out in an Ettan IPGphor, and precast IPG strips (18 cm, pH 3−11 nonlinear) were employed for the first dimension separation at 20 °C using a three-phase electrophoresis program. The proteins in IPG strips were reduced and alkylated by sequential incubation in the equilibration buffer (0.05 M Tris-HCl, pH 8.8, 2% SDS; 30% glycerol, 6 M urea, 0.002% bromophenol blue) plus 0.3 g DTT/10 mL of equilibration buffer for 15 min, then the equilibration buffer containing 0.9 g/10 mL iodoacetamide for 15 min of incubation before the 2-D electrophoresis. The 2-D SDS-PAGE was performed overnight at 80 V and 4 °C. The resulted 2-DE gels were scanned using a Typhoon 9410 scanner to produce digital images. The scanned 2-DE gel images were imported into Decyder software version 6 for protein quantitation analysis. Protein spots showing 1.2-fold change or greater, either increase or decrease, and with a p value
