Functional Impact of Protein Modification by Lipid Electrophiles
By: Carol A. Rouzer, VICB Communications
Published: January 30, 2014
Comprehensive proteomics combined with bioinformatics reveal that lipid electrophiles target proteins in functional networks.
An important mechanism of cellular damage for many toxicants is the modification of proteins by reactive electrophilic species. In some cases, the electrophile is generated directly through the metabolism of the toxicant. Indeed, this phenomenon is so well-recognized that pharmaceutical companies routinely evaluate metabolites of new drug candidates for their ability to modify proteins as part of toxicity screening. Alternatively, electrophiles may be generated indirectly through damage to endogenous cellular components during the oxidative stress that accompanies the response to many toxic exposures. Major targets of the reactive oxidizing species (ROS) generated during oxidative stress are the polyunsaturated fatty acids found in cellular membranes. ROS-mediated damage to these lipid species leads to the formation of fatty acid hydroperoxides, which may then decompose to lipid electrophiles. Proteomic analysis of the protein targets of lipid electrophiles have shown that large numbers of proteins are modified, although distinctive patterns of modification have emerged for each electrophile. Most of these studies have been carried out with isolated proteins or subcellular fractions, however, so the role of cellular defense mechanisms and the effects of intracellular architecture on protein modification patterns have not been assessed. To address these questions, and to better determine how electrophile-mediated protein modification can lead to cell death, Vanderbilt Institute of Biology members Dan Liebler, Larry Marnett, Ned Porter, and Bing Zhang joined forces to obtain a comprehensive view of cellular protein adduction by key lipid electrophiles [S. G. Codreaunu et al., (2014) Mol. Cell. Proteomics, published online January 15, DOI:10.1074/mcp.M113.032953].
Figure 1. Structures of HNE, ONE, aHNE, and aONE.
Two of the most important electrophiles generated from polyunsaturated fatty acids during oxidative stress are 4-hydroxy-2-nonenal (HNE) and 4-oxo-2-nonenal (ONE) (Figure 1). Both of these compounds form adducts with the nucleophilic sulfhydryl and amino groups of proteins through reaction of their α,β-unsaturated aldehyde moiety. Comprehensive elucidation of the protein targets of these electrophiles in intact cells required a novel method to efficiently and selectively isolate all adducted proteins from other cellular constituents. To achieve this goal, the Porter lab synthesized analogs of HNE and ONE bearing an alkynyl group at the terminal two carbons. These compounds, designated aHNE and aONE, respectively (Figure 1), retain the electrophilic α,β-unsaturated aldehyde functionality, and both exhibited the same toxicity profile as the parent compounds in the RKO human colorectal carcinoma cell line and in human THP-1 macrophage-like cells. The presence of the alkynyl group on aHNE and aONE provided a chemical “handle” that could be used to collectively isolate all proteins modified by the compounds. This was accomplished by using “click” chemistry to link the alkynyl group of the aHNE or aONE adduct to an azido-biotin tag, thus enabling the adducted proteins to be captured through the affinity of the biotin tag to streptavadin-coated beads. Key to the success of this approach was the Porter lab’s design of the azido-biotin tag, which included a linker that was readily cleaved upon exposure to ultraviolet light. This released the bound adducted proteins from the streptavadin beads, allowing improved recovery of the modified proteins, free from the bulk cellular material (Figure 2). The resulting adducted protein sample was then ready for proteomics analysis by mass spectrometry.
Figure 2. Method for isolating aHNE- and aONE-protein adducts. Following incubation with aHNE or aONE, cells were lysed, and the lysates were treated with NaBH4 to stabilize the adducts. The lysates were then incubated with an azido-biotin tag under click chemistry conditions to attach the tag to the alkynyl group of the aHNE or aONE adducts. Incubation with streptavadin beads resulted in binding of all biotin-tagged adducted proteins to the beads. All other cellular constituents were then washed away. Treatment of the beads with ultraviolet light cleaved the linker that connected the biotin tag to the adducted protein. The proteins were then easily isolated from the streptavadin beads, ready for proteomics analysis by mass spectrometry. Reprinted with permission from S. G. Codreaunu et al., (2014) Mol. Cell. Proteomics, published online January 15, DOI:10.1074/mcp.M113.032953]. Copyright 2014, ASBMB.
The Marnett lab applied this approach to isolate protein adducts from RKO cells, and the Liebler lab applied it to THP-1 cells incubated with three different concentrations of aHNE or aONE. The investigators discovered that glutathione (GSH) levels were much higher in the RKO cells (5 mM) than in the THP-1 cells (0.02 mM) and that reaction of the electrophiles with GSH partially protected the RKO cell proteome from adduction. Thus, to achieve similar levels of adducts required higher concentrations of electrophiles (0, 10, 20, and 50 μM) in RKO cells than were required in THP-1 cells (0, 5, 10, and 20 μM).
Proteomic analysis performed in the Liebler lab revealed that, of the 3,012 proteins identified in the total THP-1 proteome, 1,634 proteins were modified by aHNE, aONE, or both. Similarly, of the 2,820 proteins in the RKO proteome, 1,119 proteins were targets of one or both electrophiles. In both cell lines, about 50% of all adducted proteins were targets of both electrophiles. The Liebler lab used spectral counts to quantify the level of adduction of each of the target proteins as a function of electrophile concentration. Analysis of the results by the Zhang laboratory revealed three protein classes demonstrating different patterns of electrophile concentration-dependency. For Class I proteins, adduction occurred only at the highest electrophile concentrations, while Class III proteins were readily adducted, even at very low concentrations. Class II proteins exhibited intermediate behavior, with adduction levels linearly correlated to electrophile concentration (Figure 3).
Figure 3. Analysis of the level of protein adduction as a function of electrophile concentration. The results are shown here for aHNE treatment of THP-1 cells. Three classes of proteins with distinct patterns of concentration-dependence are observed. Class III proteins are adducted at all concentrations. Class I proteins are adducted only at the highest concentration. Class II proteins exhibit intermediate behavior. Reprinted with permission from S. G. Codreaunu et al., (2014) Mol. Cell. Proteomics, published online January 15, DOI:10.1074/mcp.M113.032953]. Copyright 2014, ASBMB.
The investigators hypothesized that proteins that were targeted by both electrophiles in both cell types were the most relevant to cytotoxicity. Therefore, they focused on these 447 core target proteins (Figure 4). The Zhang lab used the Netgestalt tool to place the core target proteins into protein-protein interaction networks. The results showed that, even though large numbers of proteins (approximately half of the proteome of each cell type) were adducted, the target proteins were enriched in just ten network modules. The majority of these were related to mRNA processing and translation. This apparent selectivity led the investigators to explore the impact of protein adduction on specific biological processes. They used WebGestalt (a WEB-based GEne SeT AnaLysis Toolkit) to evaluate Gene Ontology (GO) biological process enrichment for each of the three classes of target proteins. Although there were clear differences between the two cell lines with regard to biological processes associated with each class for each electrophile, an important trend emerged for both electrophiles in both cell types. Class II and III targets were enriched for processes involved in cytoskeletal regulation, while class I and II targets were enriched for processes involved in protein synthesis and turnover.
Figure 4. Venn diagram showing the relationships of proteins modified by aHNE (green) and aONE (red) in THP-1 (top left) and RKO (bottom left) cells. Of the 877 proteins and 595 proteins modified by both electrophiles in THP-1 and RKO cells, respectively, 447 were shared by both cell types. Bioinformatics analysis focused on these core target proteins. Reprinted with permission from S. G. Codreaunu et al., (2014) Mol. Cell. Proteomics, published online January 15, DOI:10.1074/mcp.M113.032953]. Copyright 2014, ASBMB.
Despite differences in reactivity between aHNE and aONE, and despite differences in GSH-mediated protection against electrophile damage between the two cell lines, both electrophiles exhibit the same toxicity (LD50 = 20 μM) in both RKO and THP-1 cells. Adduction of class II and class III proteins occurs at concentrations that are not toxic to the cells, so the researchers concluded that it is most likely modification of the class I proteins that contributes to cell death. These considerations suggest that cells can tolerate substantial damage to cytoskeletal components, but interference with protein synthesis and turnover is much more likely to be fatal. In this regard, it is interesting to note that some proteins in both cells lines were not adducted even at the highest electrophile concentrations. These proteins were enriched for functions related to mitochondrial electron transport and oxidative phosphorylation.
Together the results demonstrate that electrophile-mediated cytotoxicity cannot be explained simply on the basis of total protein adduction. In fact, the researchers hypothesize that there is a hierarchy of protein susceptibility to electrophile damage, such that a selected group of proteins that are readily adducted can tolerate electrophile modification without deleterious effects to the cells. This concept is exemplified by class III proteins that react under conditions of low electrophile concentration that also activate stress responses in the cells, thus enabling them to ultimately overcome the damage and survive. In contrast, high electrophile concentrations that lead to disruption of protein synthesis and processing or mitochondrial energy metabolism cannot be tolerated and will result in cell death.