Predictive toxicology of chemical mixtures using proteome-wide thermal profiling and protein target properties.

Our capability to predict the impact of exposure to chemical mixtures on environmental and human health is limited in comparison to the advances on the chemical characterization of the exposome. Current approaches, such as new approach methodologies, rely on the characterization of the chemicals and the available toxicological knowledge of individual compounds. In this study, we show a new methodological approach for the assessment of chemical mixtures based on a proteome-wide identification of the protein targets and revealing the relevance of new targets based on their role in the cellular function. We applied a proteome integral solubility alteration assay to identify 24 protein targets from a chemical mixture of 2,3,7,8-tetrachlorodibenzo-p-dioxin, alpha-endosulfan, and bisphenol A among the HepG2 soluble proteome, and validated the chemical mixture-target interaction orthogonally. To define the range of interactive capability of the new targets, the data from intrinsic properties of the targets were retrieved. Introducing the target properties as criteria for a multi-criteria decision-making analysis called the analytical hierarchy process, the prioritization of targets was based on their involvement in multiple pathways. This methodological approach that we present here opens a more realistic and achievable scenario to address the impact of complex and uncharacterized chemical mixtures in biological systems.

Insights into the mechanism of action of the chlorophyll derivative 13-2-hydroxypheophytine a on reducing neutral lipid reserves in zebrafish larvae and mice adipocytes.

Obesity is a worldwide epidemic and natural products may hold promise in its treatment. The chlorophyll derivative 13-2-hydroxypheophytine (hpa) was isolated in a screen with zebrafish larvae to identify lipid reducing molecules from cyanobacteria. However, the mechanisms underlying the lipid-reducing effects of hpa in zebrafish larvae remain poorly understood. Thus, investigating the mechanism of action of hpa and validation in other model organisms such as mice represents important initial steps. In this study, we identified 14 protein targets of hpa in zebrafish larvae by thermal proteome profiling, and selected two targets (malate dehydrogenase and pyruvate kinase) involved in cellular metabolism for further validation by enzymatic measurements. Our findings revealed a dose-dependent inhibition of pyruvate kinase by hpa exposure using protein extracts of zebrafish larvae in vitro, and in exposure experiments from 3 to 5 days post fertilization in vivo. Analysis of untargeted metabolomics of zebrafish larvae detected 940 mass peaks (66 increased, 129 decreased) and revealed that hpa induced the formation of various phospholipid species (phosphoinositol, phosphoethanolamine, phosphatidic acid). Inter-species validation showed that brown adipocytes exposed to hpa significantly reduced the size of lipid droplets, increased maximal mitochondrial respiratory capacity, and the expression of PPARy during adipocyte differentiation. In line with our data, previous work described that reduced pyruvate kinase activity lowered hepatic lipid content via reduced pyruvate and citrate, and improved mitochondrial function via phospholipids. Thus, our data provide new insights into the molecular mechanism underlying the lipid reducing activities of hpa in zebrafish larvae, and species overlapping functions in reduction of lipids.

Nonionic Surfactants can Modify the Thermal Stability of Globular and Membrane Proteins Interfering with the Thermal Proteome Profiling Principles to Identify Protein Targets.

The membrane proteins are essential targets for understanding cellular function. The unbiased identification of membrane protein targets is still the bottleneck for a system-level understanding of cellular response to stimuli or perturbations. It has been suggested to enrich the soluble proteome with membrane proteins by introducing nonionic surfactants in the solubilization solution. This strategy aimed to simultaneously identify the globular and membrane protein targets by thermal proteome profiling principles. However, the thermal shift assay would surpass the cloud point temperature from the nonionic surfactants frequently utilized for membrane protein solubilization. It is expected that around the cloud point temperature, the surfactant micelles would suffer structural modifications altering protein solubility. Here, we show that the presence of nonionic surfactants can alter protein thermal stability from a mixed, globular, and membrane proteome. In the presence of surfactant micelles, the changes in protein solubility analyzed after the thermal shift assay was affected by the thermally dependent modification of the micellar size and its interaction with proteins. We demonstrate that the introduction of nonionic surfactants for the solubilization of membrane proteins is not compatible with the principles of target identification by thermal proteome profiling methodologies. Our results lead to exploring thermally independent strategies for membrane protein solubilization to assure confident membrane protein target identification. The proteome-wide thermal shift methods have already shown their capability to elucidate mechanisms of action from pharma, biomedicine, analytical chemistry, or toxicology, and finding strategies, free from surfactants, to identify membrane protein targets would be the next challenge.

Prediction of Molecular Initiating Events for Adverse Outcome Pathways Using High-Throughput Identification of Chemical Targets.

The impact of exposure to multiple chemicals raises concerns for human and environmental health. The adverse outcome pathway method offers a framework to support mechanism-based assessment in environmental health starting by describing which mechanisms are triggered upon interaction with different stressors. The identification of the molecular initiating event and the molecular interaction between a chemical and a protein target is still a challenge for the development of adverse outcome pathways. The cellular response to chemical exposure studied with omics could not directly identify the protein targets. However, recent mass spectrometry-based methods are offering a proteome-wide identification of protein targets interacting with s but unrevealing a molecular initiating event from a set of targets is still dependent on available knowledge. Here, we directly coupled the target identification findings from the proteome integral solubility alteration assay with an analytical hierarchy process for the prediction of a prioritized molecular initiating event. We demonstrate the applicability of this combination of methodologies with a test compound (TCDD), and it could be further studied and integrated into AOPs. From the eight protein targets identified by the proteome integral solubility alteration assay after analyzing 2824 human hepatic proteins, the analytical hierarchy process can select the most suitable protein for an AOP. Our combined method solves the missing links between high-throughput target identification and prediction of the molecular initiating event. We anticipate its utility to decipher new molecular initiating events and support more sustainable methodologies to gain time and resources in chemical assessment.

Systematic analysis of chemical-protein interactions from zebrafish embryo by proteome-wide thermal shift assay, bridging the gap between molecular interactions and toxicity pathways.

The molecular interaction between chemicals and proteins often promotes alteration of cellular function. One of the challenges of the toxicology is to predict the impact of exposure to chemicals. Assessing the impact of exposure implies to understand their mechanism of actions starting from identification of specific protein targets of the interaction. Current methods can mainly predict effects of characterized chemicals with knowledge of its targets, and mechanism of actions. Here, we show that proteome-wide thermal shift methods can identify chemical-protein interactions and the protein targets from bioactive chemicals. We analyzed the identified targets from a soluble proteome extracted from zebrafish embryo, that is a model system for toxicology. To evaluate the utility to predict mechanism of actions, we discussed the applicability in four cases: single chemicals, chemical mixtures, novel chemicals, and novel drugs. Our results showed that this methodology could identify the protein targets, discriminate between protein increasing and decreasing in solubility, and offering additional data to complement the map of intertwined mechanism of actions. We anticipate that the proteome integral solubility alteration (PISA) assay, as it is defined here for the unbiased identification of protein targets of chemicals could bridge the gap between molecular interactions and toxicity pathways. SIGNIFICANCE: One of the challenges of the environmental toxicology is to predict the impact of exposure to chemicals on environment and human health. Our phenotype should be explained by our genotype and the environmental exposure. Genomic methodologies can offer a deep analysis of human genome that alone cannot explain our risks of disease. We are starting to understand the key role of exposure to chemicals on our health and risks of disease. Here, we present a proteomic-based method for the identification of soluble proteins interacting with chemicals in zebrafish embryo and discuss the opportunities to complement the map of toxicity pathway perturbations. We anticipate that this PISA assay could bridge the gap between molecular interactions and toxicity pathways.

Application of Bioactive Thermal Proteome Profiling to Decipher the Mechanism of Action of the Lipid Lowering 132-Hydroxy-pheophytin Isolated from a Marine Cyanobacteria.

The acceleration of the process of understanding the pharmacological application of new marine bioactive compounds requires identifying the compound protein targets leading the molecular mechanisms in a living cell. The thermal proteome profiling (TPP) methodology does not fulfill the requirements for its application to any bioactive compound lacking chemical and functional characterization. Here, we present a modified method that we called bTPP for bioactive thermal proteome profiling that guarantees target specificity from a soluble subproteome. We showed that the precipitation of the microsomal fraction before the thermal shift assay is crucial to accurately calculate the melting points of the protein targets. As a probe of concept, the protein targets of 132-hydroxy-pheophytin, a compound previously isolated from a marine cyanobacteria for its lipid reducing activity, were analyzed on the hepatic cell line HepG2. Our improved method identified 9 protein targets out of 2500 proteins, including 3 targets (isocitrate dehydrogenase, aldehyde dehydrogenase, phosphoserine aminotransferase) that could be related to obesity and diabetes, as they are involved in the regulation of insulin sensitivity and energy metabolism. This study demonstrated that the bTPP method can accelerate the field of biodiscovery, revealing protein targets involved in mechanisms of action (MOA) connected with future applications of bioactive compounds.

Impact of processing conditions on the phytoprostanes profile of three types of nut kernels.

The goal of this work was the identification and quantification of phytoprostanes (PhytoPs) in three types of nuts: "Walnut", "Macadamia", and "Pecan". This study represents a first approach to the relationship between the quantitative and qualitative PhytoP profiles in the "Macadamia" and "Pecan" nuts subjected to fried salt or fried honey processing. The kernels were found to contain 9-F1t-PhytoP, 9-epi-9-F1t-PhytoP, 16-B1-PhytoP, ent-16-B1-PhytoP, 9-L1-PhytoP, and ent-9-L1-PhytoP. "Macadamia" fried salt nuts were the only ones that produced 9-epi-9-D1-PhytoP and 9-D1-PhytoP. The total PhytoP concentration in raw nuts was in the range of 5541-7830 ng kg-1 fresh weight (FW); for most of the PhytoPs, the concentrations were lowest in raw walnuts, indicating that concentration of each PhytoP was influenced by the genotype. The frying process increased the total PhytoPs concentration to the range of 8903-33,727 ng kg-1 FW. Therefore, this is the first work describing PhytoPs in nuts and reinforces the capacity of these compounds to act as biomarkers to monitor the processing treatments that influence the final quality of nuts.

Impact of packaging atmosphere, storage and processing conditions on the generation of phytoprostanes as quality processing compounds in almond kernels.

The thermal processing of almond kernels implies the use of techniques that produce chemical changes such as oxidation. Phytoprostanes (PhytoPs) are considered biomarkers of the oxidative stress in plants. We studied the PhytoP profile in kernels of almond cultivars under different conditions, in relation to packaging, temperature and time of storage and processing. The most abundant PhytoP was the F1t series. The PhytoP levels increased significantly with the time of storage (3 and 6months) and the total PhytoP concentration was higher under air than in a vacuum packaging atmosphere. Storage at 24°C raised the concentrations of individual PhytoPs and the total sum of PhytoPs. The frying and roasting processes led to a strong reduction of the original concentration of most PhytoPs and promoted the synthesis of specific PhytoPs that were not detected in raw kernels and thus could be biomarkers of the degree of oxidative degradation of almonds.