Structure−Activity Comparison of the Cytotoxic Properties of Diethyl

Nov 24, 2010 - Wooster, Wooster, Ohio 44691, United States; Department of Biochemistry, Vanderbilt ... Medicine, NashVille, Tennessee 37232, United St...
0 downloads 0 Views 5MB Size
Download PDF
Chem. Res. Toxicol. 2011, 24, 81–88

81

Structure-Activity Comparison of the Cytotoxic Properties of Diethyl Maleate and Related Molecules: Identification of Diethyl Acetylenedicarboxylate as a Thiol Cross-Linking Agent1 James D. West,*,† Chelsea E. Stamm,† and Philip J. Kingsley§ Program in Biochemistry and Molecular Biology, Departments of Biology and Chemistry, The College of Wooster, Wooster, Ohio 44691, United States; Department of Biochemistry, Vanderbilt UniVersity School of Medicine, NashVille, Tennessee 37232, United States ReceiVed August 28, 2010

Many R,β-unsaturated carbonyl compounds are used in biochemical and medical research. Their biological effects are due in large part to their electrophilic properties, whereby they undergo reaction with nucleophilic sites in proteins and nucleic acids. Here, we describe a structure-activity comparison of the cytotoxic properties of diethyl maleate (DEM) and closely related chemical analogs. All molecules that contained an R,β-unsaturated carbonyl group were cytotoxic to human colorectal carcinoma cells, causing apoptotic cell death. However, related molecules lacking this chemical moiety were not cytotoxic. One of the molecules screened, diethyl acetylenedicarboxylate (DAD), was considerably more cytotoxic than DEM and other analogues. Induction of cell death by DAD was significantly decreased following preincubation of cells with N-acetylcysteine, suggesting that its reactivity with thiols in cells might account for its cytotoxicity. By use of a model thiol compound, it was found that DAD can undergo addition reactions with two equivalents of thiol. When the reactivity of DAD with proteins was explored, it was determined that DAD induces oligomerization of Gpx3p, a yeast glutathione peroxidase with highly reactive cysteine residues in its active site. Our results suggest that DAD functions as a protein-thiol cross-linker, providing a potential chemical explanation for its cytotoxic potency. Introduction R,β-Unsaturated carbonyl compounds make up a broad class of natural products, industrial chemicals, environmental pollutants, contact allergens, and endogenous byproducts of lipid peroxidation and nonenzymatic amino acid decomposition (1-4). Regardless of their origin, most R,β-unsaturated carbonyl compounds are cytotoxic at various levels due to their electrophilic properties. Nucleophilic sites in DNA and proteins are common targets of these electrophiles (4-8). Electrophilic modification of protein targets, in particular, may result in inhibition of their function(s) and an increased propensity to misfold and aggregate, thereby promoting toxicity (4, 5, 9, 10). Several large-scale studies on R,β-unsaturated carbonyl compounds and other electrophiles have indicated a complex relationship between their predicted electrophilic character, hydrophobic properties, and reactivity with model nucleophiles (11-16). Because previous work has indicated that thiol groups present in small molecules and proteins are highly susceptible to Michael addition reactions, most of these reactivity studies are carried out with thiol-containing small molecules (e.g., glutathione) or peptides (3, 17). Whereas screens of R,βunsaturated carbonyl compound reactivity allow important predictions to be made concerning a molecule’s biological effects, side-by-side studies comparing the molecule’s reactivity with organismal and/or cell death are also important for understanding its toxicity (13, 18). * Corresponding author [phone (330) 263-2368; fax (330) 263-2378; e-mail [email protected]]. † The College of Wooster. § Vanderbilt University School of Medicine.

Among the classes of R,β-unsaturated carbonyl compounds studied, esterified derivatives of R,β-unsaturated dicarboxylic acids have received considerable attention for their biochemical utility and therapeutic potential. Maleate esters have traditionally been used in biochemical experiments to block thiol groups in proteins in vitro and deplete glutathione levels in cells (19, 20). Esterified derivatives of fumarate, the trans-isomer of maleate, have been developed as treatments for inflammatory skin and neurological disorders, and their therapeutic effects are purportedly derived from their electrophilic properties (21-23). The methyl ester of acetylenedicarboxylate is a known skin irritant (24), although it and related molecules have not been studied extensively in cellular experiments. Although many esterified derivatives of R,β-unsaturated carboxylic acids have known cytotoxicities, their ability to cause cell death (i.e., activate apoptotic signaling processes) in mammalian cells has most often been evaluated individually (25-27). Here, we tested a hypothesis that diethyl maleate (DEM)1 (1) and related molecules exhibit different toxicities as a result of their hydrophobicity or the bond type at the site of electrophilicaddition.Tothisend,weperformedastructure-activity comparison of DEM and analogous electrophiles to determine the cytotoxic potency of each molecule in human colorectal 1 Nonstandard abbreviations: DEM, diethyl maleate; DMM, dimethyl maleate; DBM, dibutyl maleate; MA, maleic acid; DEF, diethyl fumarate; DAD, diethyl acetylenedicarboxylate; DES, diethyl succinate; DAB, cis1,4-diacetoxy-2-butene; WST-1, 4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H5-tetrazolio]-1,3-benzene disulfonate; TBS-T, Tris-buffered saline with 0.1% Tween 20; PARP-1, poly(ADP-ribose) polymerase-1; NAC, N-acetylcysteine; NBM, 3-nitrobenzyl mercaptan; IA, iodoacetamide; NEM, Nethylmaleimide.

10.1021/tx100292n  2011 American Chemical Society Published on Web 11/24/2010

82

Chem. Res. Toxicol., Vol. 24, No. 1, 2011

carcinoma (RKO) cells. In the process of conducting these studies, we determined that one molecule, diethyl acetylenedicarboxylate (DAD), was considerably more potent at inducing cell death than the other electrophiles tested, including DEM. Upon examining the thiol reactivity of DAD, we found that it functions as a thiol cross-linking agent capable of causing protein oligomerization, a property that may enhance its cytotoxicity.

Materials and Methods Chemicals. All chemicals and supplies, except where otherwise indicated, were purchased from Sigma-Aldrich (St. Louis, MO). DEM, dimethyl maleate (DMM), dibutyl maleate (DBM), diethyl fumarate (DEF), DAD, diethyl succinate (DES), cis-1,4-diacetoxy2-butene (DAB), and maleic acid (MA) were dissolved and diluted in 100% ethanol prior to use. Cell Culture. RKO colorectal carcinoma cells were originally obtained from the American Type Culture Collection and were kindly provided by the laboratory of Lawrence J. Marnett (Vanderbilt University School of Medicine). RKO cells were grown at 37 °C and 5% CO2 in Dulbecco’s Modified Eagle Medium (Invitrogen, Carlsbad, CA) supplemented with 10% fetal bovine serum (ThermoFisher, Rockford, IL), 2 mM L-glutamine (Invitrogen), and antibiotic/antimycotic (Invitrogen). Cells were treated at a confluence of ∼40-60%. Ethanol concentrations in treated cell cultures were e0.1%. Cytotoxicity Assays. The cytotoxic potency of individual molecules was determined using 4-[3-(4-iodophenyl)-2-(4-nitrophenyl)-2H-5-tetrazolio]-1,3-benzene disulfonate (WST-1) cell proliferation reagent (Roche Applied Science, Indianapolis, IN) according to the manufacturer’s instructions. Briefly, 50 µL of cells was seeded into 96-well plates and allowed to grow for approximately 24 h prior to reaching 50% confluence. Molecules were added to the wells as 2× stocks dissolved in 50 µL of DMEM, bringing the total culture volume to 100 µL. For experiments with NAC, 50 µL of cells was seeded as described above, followed by pretreatment with 25 µL of 3× NAC dissolved in DMEM for 3 h and treatment with 25 µL of 4× DAD dissolved in DMEM. Following a 24 h incubation (or a 21 h incubation, in experiments with NAC), 10 µL of WST-1 reagent was added to each well and was incubated for 1 h at 37 °C and 5% CO2. The absorbance of each well was read using a Thermo Multiskan Ascent plate reader (Hudson, NH) at 450 nm with background subtraction at 595 nm. Each treatment was performed in quadruplicate, and numbers were normalized to WST-1 conversion in the untreated control cultures. IC50 values (the concentration of a molecule at which the WST-1 conversion in a treated culture was 50% of the control) were calculated for a normalized response with a variable slope using GraphPad Prism (La Jolla, CA). DNA Fragmentation Analysis. DNA fragmentation analysis on treated cells was performed as described previously (28). Briefly, treated cells were scraped off cell culture dishes into media and centrifuged at 1000g for 3 min. Cell pellets were washed once with cold PBS and were lysed on ice for 30 min in a buffer containing 10 mM Tris (pH 8.0), 10 mM EDTA, and 0.2% Triton X-100. Lysates were centrifuged at 21000g for 10 min. The supernatant was recovered and incubated with 200 µg/mL RNase A for 30 min at 37 °C. Proteins in the lysate were digested by incubation with 1% SDS and 1 mg/mL proteinase K for 2 h at 55 °C. DNA was extracted once with phenol/chloroform/isoamyl alcohol (25:24:1), extracted once with chloroform, and precipitated in ethanol overnight at -80 °C. Samples were centrifuged for 20 min at 21000g at 4 °C to pellet DNA. The supernatant was removed. Isolated DNAs were vacuum-dried, resuspended in H2O, electrophoresed on a 1% agarose gel, and visualized with ethidium bromide. Immunoblot Analysis of Apoptotic Protein Cleavage. The protocol for immunoblotting was performed as described previously (29). Briefly, treated RKO cells were harvested by scraping cells

West et al. into media. Cells were centrifuged at 1000g for 3 min and washed twice in cold PBS before lysis in a buffer containing 20 mM Tris (pH 7.5), 150 mM NaCl, 1 mM EDTA, 1 mM ethylene glycol bis(2aminoethyl ether)-N,N,N′,N′-tetraacetic acid (EGTA), 1% Triton X-100, and protease inhibitor cocktail (ThermoFisher). Lysates were centrifuged at 21000g for 5 min to clear debris, and protein concentrations in each sample were quantified using the bicinchoninic acid assay (ThermoFisher) with BSA as a standard. Equal amounts of proteins (20-30 µg) were treated with reducing SDSPAGE sample buffer [60 mM Tris, pH 6.8; 0.7 M 2-mercaptoethanol; 2% (w/v) SDS; 10% (v/v) glycerol; 0.33 mg/mL bromophenol blue] and were electrophoresed on a 4-20% gradient gel. Proteins were transferred to polyvinylidene fluoride membrane at 4 °C for 2 h at 250 mA. Membranes were blocked with Trisbuffered saline containing 0.1% Tween 20 (TBS-T) and 5% nonfat dry milk for 1 h at room temperature before being incubated with rabbit primary antibodies against R-tubulin, caspase-3, R-fodrin, and cleaved PARP-1 (Cell Signaling Technologies, Danvers, MA) for 2 h at room temperature. Subsequently, membranes were washed three times with TBS-T for 5 min, incubated with anti-rabbithorseradish peroxidase secondary antibody (Cell Signaling) for 45 min, washed four times with TBS-T, and visualized using enhanced chemiluminescence reagents (GE Healthcare, Piscataway, NJ). Formation and Characterization of Covalent Adducts between DAD and 3-Nitrobenzyl Mercaptan (NBM) Using LCMS. NBM (200 µM) was incubated with 40, 200, or 1000 µM DAD in 10 mM Tris (pH 8) at 37 °C for 24 h. HPLC-UV-MS analysis of the DAD/NBM reaction mixtures was performed on a Waters Acquity UPLC system in-line with a Waters Acquity photodiode array detector and a Thermo LTQ ion trap mass spectrometer equipped with an electrospray source. The photodiode array was set to 265 nm, and the LTQ detected ions from m/z 150 to 600 in positive ion mode. Chromatographic separation was achieved on an Agilent Zorbax C18 column (2.1 × 150 mm, 5 µm particle size) using the following gradient: 0-50% B over 5 min followed by 50-90% B over 12 min, where (A) was water and (B) was acetonitrile and each component contained 0.5% formic acid (v/v). The flow rate was 0.4 mL/min and the mobile phase was split 1:1 to the UV detector and the mass spectrometer. Aliquots of the reactions were injected without any pretreatment or purification. Protein Cross-Linking Analysis. Procedures for cloning, expressing, and purifying Gpx3p (a glutathione peroxidase from Saccharomyces cereVisiae) are included in the Supporting Information. Purified Gpx3p (180 µM) was incubated with 20 mM DTT for 15 min at room temperature to reduce any disulfide bonds formed from autoxidation. Reduced Gpx3p was centrifuged through a Micro Biospin 6 gel filtration column (Bio-Rad) equilibrated with 100 mM Tris (pH 8) for 1 min at 1000g to remove salts and DTT. Protein concentrations were determined on the basis of A280 values using a molar absorptivity of 16410 M-1 cm-1. For cross-linking experiments, reactions (20 µL) contained 100 mM Tris (pH 8), 2.5 mM EDTA, 150 mM NaCl, 10 µM reduced Gpx3p, and various concentrations of DEM or DAD dissolved in EtOH. Reactions were incubated at 37 °C for 1-24 h, at which time 1 µL of 1 M DTT was added to terminate the reaction, along with 10 µL of 3× reducing SDS-PAGE sample buffer. Some reactions were incubated for 1 h prior to the addition of DAD with 100 mM iodoacetamide (IA), 100 mM N-ethylmaleimide (NEM), or 100 mM dithiothreitol (DTT) to prevent reactivity of thiol groups in Gpx3p with DAD. Samples were heated at 95 °C for 5 min, electrophoresed on 4-20% acrylamide Tris-glycine gel (Invitrogen), and visualized by staining with Coomassie Brilliant Blue. Reactions between Gpx3p and DAD were also conducted in 100 mM HEPES (pH 8.0) instead of 100 mM Tris (pH 8.0) as described in the Supporting Information.

Results Cytotoxicities of DEM and Related Molecules. In this study, we compared the cytotoxic potency of DEM, a commonly used thiol-alkylating agent, with that of closely related molecules (Figure 1). RKO cells treated with these molecules for 24 h

Cytotoxicity of DEM and Structural Analogues

Figure 1. Structures of molecules used in this study.

were monitored for changes in viability using the tetrazolium salt, WST-1. All esterified derivatives of maleic acid (MA) decreased the number of viable RKO cells compared to control, whereas MA itself did not significantly decrease cell viability up to 1 mM (Figure 2A). The calculated IC50 values for causing decreased viability revealed only slight differences in the overall cytotoxicities of DEM, DMM, and DBM (Table 1). In testing molecules that differed at the site of the R,β-unsaturated bond or the regiochemistry of the carbonyl groups, three molecules (DEM, DEF, and DAD) decreased RKO cell viability (Figure 2B). However, molecules lacking a Michael acceptor site (DES and DAB) did not affect cellular viability at the doses tested (Figure 2B). The IC50 calculations for this series of molecules revealed the following order of cytotoxic potency: DAD > DEF > DEM . DES ≈ DAB (Table 1). Mechanism of Cell Death Initiated by DEM and Structural Analogues. To determine whether the cells treated with DEM and related molecules were dying through an apoptotic mechanism, DNA was isolated from treated cells to analyze the amount of apoptotic nucleosomal fragmentation. DNA fragmentation was observed for all molecules that caused pronounced cytotoxicity in our earlier experiments (Figure 3A,C). The dose response over which each molecule induced DNA fragmentation closely mirrored the dose-response profile observed in the cytotoxicity experiments with WST-1. Similarly, immunoblot analysis performed on lysates from treated cells indicated cleavage of the apoptotic protease caspase-3 and two of its targets, PARP-1 and R-fodrin, at the doses at which DNA fragmentation was observed (Figure 3B,D) (30). These results reveal comparable trends between the cytotoxic potency of these molecules and the dose ranges over which they induce apoptosis. Protection against DAD-Mediated Cell Death by NAC. Because DAD was the most potent molecule in both the cytotoxicity and apoptosis assays, we explored the possibility that its chemical reactivity accounts for its cytotoxicity. Since thiols in proteins and in glutathione are likely targets of DAD, we tested whether adding an excess of the thiol competitor NAC

Chem. Res. Toxicol., Vol. 24, No. 1, 2011 83

Figure 2. Comparison of cytotoxicities of DEM and structural analogues. Molecules tested differed in the length of the aliphatic group attached to the carbonyl via an ester linkage (A) or the orientation or types of bonds present at the electrophilic center (B). Cell treatments and WST-1 cytotoxicity assays were conducted as described under Materials and Methods. The percent of WST-1 conversion in each case is compared with the control (100%). Results are the averages ( SD from two independent experiments, each conducted in quadruplicate for all molecules except DEM; results for DEM are from four independent experiments, each conducted in quadruplicate.

Table 1. Comparison of Cytotoxicities for the Molecules Studied compound

CAS Registry No.

IC50a (µM)

maleic acid (MA) dimethyl maleate (DMM) diethyl maleate (DEM) dibutyl maleate (DBM) diethyl fumarate (DEF) diethyl acetylenedicarboxylate (DAD) diethyl succinate (DES) cis-1,4-diacetoxy-2-butene (DAB)

110-16-7 624-48-6 141-05-9 105-76-0 623-91-6 762-21-0 123-25-1 25260-60-0

>1000 141 ( 12 138 ( 8 131 ( 22 81 ( 6 30 ( 1 >1000 >1000

a

IC50 values (the values at which WST-1 conversion is 50% of the control) ( SE were determined using the WST-1 cytotoxicity assay results depicted in Figure 2.

could prevent DAD-mediated toxicity. Preincubation of cells with NAC significantly decreased cell death caused by DAD (Figure 4A). In addition, nucleosomal DNA fragmentation was not observed in cells treated with NAC before the addition of DAD, suggesting that NAC protects against DAD-mediated apoptosis (Figure 4B). A similar decrease in DNA fragmentation was observed when cells were incubated with NAC prior to treatment with DEM or DEF (data not shown). These results suggest that cell death induced by DAD depends on its reactivity with cellular nucleophiles, including thiol groups in proteins. Analysis of Michael Adducts of DAD Using the Model Thiol Compound NBM. The preceding cell death studies revealed that DAD was considerably more toxic than related molecules that contain only a double bond at the Michael acceptor site and that free thiols prevent DAD-mediated cytotoxicity. One potential reason for its greater cytotoxicity in comparison to related molecules is that it contains a second Michael acceptor site following the addition of a nucleophile

84

Chem. Res. Toxicol., Vol. 24, No. 1, 2011

West et al.

Figure 3. Induction of apoptotic DNA fragmentation and caspase target proteolysis in cells treated with DEM and structural analogues. DNA fragmentation analysis was conducted for DEM analogues that differ in the length of the aliphatic ester group (A, B) or the orientation or types of bonds present at the electrophilic center (C, D). RKO cells were treated with the indicated concentrations of DEM analogues for 24 h. In panels A and C, soluble DNA was isolated as described under Materials and Methods. A DNA ladder is included in the left lane with the sizes of prominent bands indicated. In panels C and D, equal amounts of protein lysates (20 µg) from treated cells were probed for cleavage of apoptotic protein markers. R-Tubulin was included as a loading control. Results are representative of three independent experiments.

Figure 4. Protection against DAD-mediated cell death with NAC. RKO cells were incubated with 5 mM NAC for 3 h prior to treatment with the indicated concentrations of DAD for 21 h. (A) Cell viability was determined using the WST-1 assay. Results are the average ( SD of two experiments, each performed in quadruplicate. (B) Soluble DNA was isolated from treated cells, electrophoresed on a 1% agarose gel, and visualized using ethidium bromide staining. Results are representative of three independent experiments.

Scheme 1. Predicted Products from the Reaction between DAD and NBMa

a Numbers in parentheses refer to peaks eluting from the column (see Figure 5), and the calculated molecular weights (MW) for both products are shown. For simplicity, only a single stereochemical configuration is drawn.

to the triple bond (Scheme 1). To explore the possibility that a single molecule of DAD undergoes two Michael addition reactions, we incubated DAD and NBM in various ratios for 24 h at 37 °C and analyzed the reaction mixtures by HPLCUV-MS. NBM absorbs strongly at approximately 265 nm

(Figure 1 of the Supporting Information), providing a spectroscopic signature for the HPLC-UV detection of its reaction products. At least four different reaction products (peaks 3a, 3b, 4a, and 4b in Figure 5A, bottom chromatogram) were observed following analysis of an incubation of 1:1 DAD/NBM.

Cytotoxicity of DEM and Structural Analogues

Chem. Res. Toxicol., Vol. 24, No. 1, 2011 85

Figure 5. Adduct formation between DAD and the model thiol compound NBM: (A) HPLC-UV (265 nm) chromatograms from 15 µL injections of 200 µM DAD (top, peak 1), 200 µM NBM (middle, peak 2), and a reaction between 200 µM NBM and 200 µM DAD (24 h at 37 °C, bottom); (B) mass spectra corresponding to the retention times of 3a (11.1 min, top) and 3b (11.6 min, bottom); (C) mass spectra corresponding to the retention times of 4a (14.2 min, top) and 4b (14.5 min, bottom). LC-MS results are representative of two independent experiments.

Peaks 3a and 3b (11.1 and 11.6 min, respectively, Figure 5A) displayed greater peak area than peaks 4a and 4b (14.2 and 14.5 min, respectively, Figure 5A) at 265 nm. Incubation of a 5-fold excess of NBM with DAD yielded greater amounts of molecules 4a and 4b, whereas incubation of a 5-fold excess of DAD with NBM yielded 3a and 3b almost exclusively (Figure 2 of the Supporting Information). Collectively, these results suggest that products 3a and 3b are predominant when NBM is limiting but that products 4a and 4b are preferentially formed when DAD is the limiting reagent. Concurrent mass spectrometric (MS) analysis was performed on all peaks eluting from the C18 column. The mass spectra of 3a and 3b displayed a base peak of m/z 340 (Figure 5B). This m/z value corresponded to the protonated compound formed by the addition of a single NBM molecule to DAD. In addition, the ammonium adduct (m/z 357, Figure 5B) and sodium adduct (m/z 362, Figure 5B) of the single-addition product were also observed. The mass spectra of peaks 4a and 4b showed a base peak at m/z 509 (Figure 5C), which corresponds to the protonated compound formed by the reaction of two molecules of NBM with a single molecule of DAD. The ammonium adduct (m/z 526, Figure 5C) and sodium adduct (m/z 531, Figure 5C) of this product were also observed. Adducts between DAD and

Tris (a primary amine used to buffer the reaction) were also formed in these assays (Figure 3 of the Supporting Information). However, the presence of Tris did not significantly decrease the ability of DAD to react with NBM incubated at 1:1 ratios (Figure 5A, bottom panel), as indicated by the depletion of NBM under these conditions. Taken together, these results indicate that DAD principally reacts with thiol-containing molecules and that it can add two thiol compounds to its electrophilic center. In addition, these findings suggest that DAD may function as a thiol cross-linker in intact proteins. Protein Cross-Linking by DAD. We tested whether proteins are cross-linked by DAD using purified Gpx3p, a gluthathione peroxidase from S. cereVisiae that functions in oxidant defense signaling and contains highly reactive Cys residues in its active site, as a model protein (31-33). Gpx3p formed higher molecular weight species following treatment with increasing concentrations of DAD (Figure 6A). Similar high molecular weight species were observed when Gpx3p was incubated with 1,8-bis(maleimidodiethylene glycol), a molecule that contains two maleimide moieties and can cross-link Cys residues in proteins (data not shown). However, no increase in oligomerization of Gpx3p was observed when incubated with DEM

86

Chem. Res. Toxicol., Vol. 24, No. 1, 2011

West et al.

Discussion

Figure 6. Cross-linking of Gpx3p by DAD is thiol-dependent. Gpx3p (10 µM) was exposed to various concentrations of DEM or DAD for 24 h at 37 °C (A) or to 2 mM DAD over 24 h (B). In panel C, Gpx3p was preincubated with the indicated concentrations of either DTT, IA, or NEM for 1 h at 37 °C prior to treatment with 2 mM DAD for 24 h. Samples were resolved using SDS-PAGE and visualized with Coomassie Blue. Numbers on the left side of each gel image reflect the position of molecular weight standards (in kDa). The positions of oligomers and highly cross-linked proteins are indicated on the right side of each gel image. Results are representative of three independent experiments.

(Figure 6A). The cross-linking of Gpx3p was time-dependent and increased steadily over an incubation period of 1-24 h (Figure 6B). We next sought to determine whether DAD-mediated crosslinking of Gpx3p was dependent on its ability to react with Cys residues in the protein. To do this, Gpx3p was incubated for 30 min with either 100 mM DTT (which would potentially compete for DAD with Gpx3p), 100 mM IA (which would principally react with thiol groups in Gpx3p), or 100 mM NEM (which would also principally react with thiol groups in Gpx3p) prior to treatment with DAD. Pronounced cross-linking of Gpx3p was observed only in the protein sample where DAD was incubated with Gpx3p alone (Figure 6C). Conversely, DTT, IA, and NEM significantly decreased the amount of Gpx3p cross-linking by DAD (Figure 6C). In addition, we conducted Gpx3p crosslinking experiments in the presence of either Tris buffer (which contains a nucleophilic primary amine) or HEPES buffer to determine whether protein cross-linking was influenced by the use of Tris in our assays. Comparable cross-linking of Gpx3p was observed when reactions were incubated in Tris or HEPES (Figure 4 of the Supporting Information), suggesting that the formation of adducts between DAD and primary amines does not interfere with protein cross-linking. Together, these results imply that protein cross-linking by DAD principally involves thiol groups.

Here, we have conducted a direct cytotoxicity comparison of DEM and closely related structural analogues to determine what structural differences influence cell death. Of the molecules that we studied, DAD, which contains a triple bond at the site of Michael addition, was the most potent inducer of apoptotic cell death (Figures 2 and 3). In searching for a chemical explanation for DAD’s potency, we tested the hypothesis that DAD could react with two thiol groups (Scheme 1) and found that, when incubated with NBM, DAD is capable of adding two thiols to its triple bond (Figure 5). These results, taken together with the observation that DAD causes oligomerization of the Cys-containing protein Gpx3p (Figure 6), suggest that DAD functions as a thiol cross-linking agent, a property that potentially influences its ability to induce cell death. Although DAD was the most potent molecule identified in our cytotoxicity studies, several differences in the cytotoxic potencies of the other molecules included in this study were also observed. Specifically, we found that there is an absolute requirement for a Michael acceptor site in these molecules to cause cell death. Sites for Michael addition are not present in DES or DAB, neither of which caused pronounced cytotoxicity in our experiments. These results imply that reactivity with cellular nucleophiles is key to bringing about the observed cytotoxicities. In addition, examining the apoptotic responses caused by MA, DMM, DEM, and DBM revealed that cellular permeability of the electrophile is potentially important. MA, which would likely exist as a charged species and be relatively impermeable to cells, did not show pronounced cytotoxicity in our experiments. In contrast, the apoptotic response triggered by molecules that varied in the length of the aliphatic chain attached to the ester bond revealed that more hydrophobic molecules (i.e., DBM) caused apoptosis at lower concentrations than DEM or DMM (Figure 3A,C). One possible explanation for this result is that DBM is more hydrophobic and therefore more likely to cross the cell membrane. The relative hydrophobicity of an electrophile has been proposed to contribute to the toxicity of some, but not all, R,β-unsaturated carbonyl compounds (34-36). These molecules may also differ in cytotoxicity due to their reactivity with cellular nucleophiles. In this case, DAD, with the ability to undergo two Michael addition reactions at its electrophilic center, exhibited the greatest difference in cytotoxicity. A recent paper suggests that the major amino acid modified by R,β-unsaturated carbonyl compounds in the proteome is Cys (17). Therefore, we developed an experimental approach that used the model compound NBM to study the products formed between DAD and free thiols. When DAD was incubated with NBM, two single-addition products (3a and 3b) and at least two products in which two molecules of NBM were added to one molecule of DAD (4a and 4b) were observed. One likely explanation for two different peaks being observed for both groups of Michael addition products is cis-trans isomers eluting with slightly different retention times (in the case of 3a and 3b) and a combination of stereoisomers eluting with different retention times (in the case of 4a and 4b). Although the exact configuration of the products formed remains unresolved, the existence of products 4a and 4b in these experiments provides evidence that DAD functions as a Cys cross-linker in vitro and potentially in vivo. In support of a role for protein cross-linking in DAD-mediated cell death, other published studies reveal comparable differences in the cytotoxic potencies of protein cross-linkers and structur-

Cytotoxicity of DEM and Structural Analogues

ally related molecules that cannot induce covalent protein oligomerization. Comparison of either 2-nonenoic acid or nonanal, neither of which would likely undergo cross-linking reactions, to the protein cross-linkers 4-hydroxy-2-nonenal and 2-nonenal reveals at least a 3-4-fold enhancement in cytotoxicity afforded by putative protein cross-linkers (34, 37). Likewise, acrolein, which can form reversible Cys-Lys crosslinks, is a more potent inhibitor of synaptosome function than acrylamide, which does not induce appreciable protein crosslinking (38-40). These previous findings, in addition to our finding that DAD is approximately 5-fold more potent at inducing cell death than DEM, suggest that one molecular property that can influence the cytotoxic efficacy of an R,βunsaturated carbonyl compound is its ability to form protein cross-links. Although the cellular proteins modified by DAD and other DEM-related analogues are still unknown, many protein targets are likely to undergo inactivation and/or aggregation following their reaction with these electrophiles. Recent evidence suggests that protein-modifying electrophiles cause the aggregation of a variety of disease-related proteins and cytoskeletal proteins, as well as inactivation of specific molecular chaperones, the proteasome, signaling enzymes, and metabolic enzymes that utilize thiol groups to carry out their function (4, 5, 10, 41-47). The additive effect of these protein damage events is most likely a major reason for the cytotoxic properties of R,β-unsaturated carbonyl compounds. Because DAD was the only electrophile studied that can modify target proteins to form intramolecular or intermolecular cross-links, this property may afford an increased propensity to cause cell death. Although DAD is likely to be highly toxic to most types of cells and organisms as a result of its cross-linking properties, it may serve as a useful chemical tool for cell biological and biochemical experiments in which protein aggregation is desired or transient proteinprotein interactions require chemical cross-linking to facilitate characterization. Acknowledgment. This work was supported by start-up funds, faculty development funds, and William H. Wilson Research Funds provided by the College of Wooster and by a grant from the Howard Hughes Medical Institute Undergraduate Science Education Program to The College of Wooster. C.E.S. was supported through the College’s Sophomore Research Program and the Howard Hughes Medical Institute. We thank Mark Snider and Paul Edmiston (The College of Wooster) for assistance with HPLC experiments and Larry Marnett (Vanderbilt University School of Medicine) and the staff of the Vanderbilt University Mass Spectrometry Center for assistance with mass spectrometry experiments. Supporting Information Available: Additional experimental procedures and figures. This material is available free of charge via the Internet at http://pubs.acs.org.

References (1) Drahl, C., Cravatt, B. F., and Sorensen, E. J. (2005) Protein-reactive natural products. Angew. Chem., Int. Ed. Engl. 44, 5788–5809. (2) Liebler, D. C. (2008) Protein damage by reactive electrophiles: targets and consequences. Chem. Res. Toxicol. 21, 117–128. (3) Aptula, A. O., and Roberts, D. W. (2006) Mechanistic applicability domains for nonanimal-based prediction of toxicological end points: general principles and application to reactive toxicity. Chem. Res. Toxicol. 19, 1097–1105. (4) Marnett, L. J., Riggins, J. N., and West, J. D. (2003) Endogenous generation of reactive oxidants and electrophiles and their reactions with DNA and protein. J. Clin. InVest. 111, 583–593.

Chem. Res. Toxicol., Vol. 24, No. 1, 2011 87 (5) Rudolph, T. K., and Freeman, B. A. (2009) Transduction of redox signaling by electrophile-protein reactions. Sci. Signal. 2, re7. (6) LoPachin, R. M., Gavin, T., Petersen, D. R., and Barber, D. S. (2009) Molecular mechanisms of 4-hydroxy-2-nonenal and acrolein toxicity: nucleophilic targets and adduct formation. Chem. Res. Toxicol. 22, 1499–1508. (7) Burcham, P. C. (1998) Genotoxic lipid peroxidation products: their DNA damaging properties and role in formation of endogenous DNA adducts. Mutagenesis 13, 287–305. (8) Gates, K. S. (2009) An overview of chemical processes that damage cellular DNA: spontaneous hydrolysis, alkylation, and reactions with radicals. Chem. Res. Toxicol. 22, 1747–1760. (9) Bieschke, J., Zhang, Q., Bosco, D. A., Lerner, R. A., Powers, E. T., Wentworth, P., Jr., and Kelly, J. W. (2006) Small molecule oxidation products trigger disease-associated protein misfolding. Acc. Chem. Res. 39, 611–619. (10) Petersen, D. R., and Doorn, J. A. (2004) Reactions of 4-hydroxynonenal with proteins and cellular targets. Free Radical Biol. Med. 37, 937–945. (11) McCarthy, T. J., Hayes, E. P., Schwartz, C. S., and Witz, G. (1994) The reactivity of selected acrylate esters toward glutathione and deoxyribonucleosides in Vitro: structure-activity relationships. Fundam. Appl. Toxicol. 22, 543–548. (12) Schultz, T. W., Yarbrough, J. W., Hunter, R. S., and Aptula, A. O. (2007) Verification of the structural alerts for Michael acceptors. Chem. Res. Toxicol. 20, 1359–1363. (13) Yarbrough, J. W., and Schultz, T. W. (2007) Abiotic sulfhydryl reactivity: A predictor of aquatic toxicity for carbonyl-containing R,βunsaturated compounds. Chem. Res. Toxicol. 20, 558–562. (14) Bohme, A., Thaens, D., Paschke, A., and Schuurmann, G. (2009) Kinetic glutathione chemoassay to quantify thiol reactivity of organic electrophiles: application to R,β-unsaturated ketones, acrylates, and propiolates. Chem. Res. Toxicol. 22, 742–750. (15) Roberts, D. W., and Natsch, A. (2009) High throughput kinetic profiling approach for covalent binding to peptides: application to skin sensitization potency of Michael acceptor electrophiles. Chem. Res. Toxicol. 22, 592–603. (16) Schwobel, J. A., Wondrousch, D., Koleva, Y. K., Madden, J. C., Cronin, M. T., and Schuurmann, G. (2010) Prediction of Michaeltype acceptor reactivity toward glutathione. Chem. Res. Toxicol. 23, 1576–1585. (17) Weerapana, E., Simon, G. M., and Cravatt, B. F. (2008) Disparate proteome reactivity profiles of carbon electrophiles. Nat. Chem. Biol. 4, 405–407. (18) Bohme, A., Thaens, D., Schramm, F., Paschke, A., Schuurmann, G. (2010) Thiol reactivity and its impact on the ciliate toxicity of R,βunsaturated aldehydes, ketones, and esters. Chem. Res. Toxicol. doi: 10.1021/tx100226n(in press). (19) Plummer, J. L., Smith, B. R., Sies, H., and Bend, J. R. (1981) Chemical depletion of glutathione in ViVo. Methods Enzymol. 77, 50–59. (20) Mulder, G. J., and Ouwerkerk-Mahadevan, S. (1997) Modulation of glutathione conjugation in ViVo: how to decrease glutathione conjugation in ViVo or in intact cellular systems in Vitro. Chem.-Biol. Interact. 105, 17–34. (21) Rostami Yazdi, M., and Mrowietz, U. (2008) Fumaric acid esters. Clin. Dermatol. 26, 522–526. (22) Linker, R. A., Kieseier, B. C., and Gold, R. (2008) Identification and development of new therapeutics for multiple sclerosis. Trends Pharmacol. Sci. 29, 558–565. (23) Mrowietz, U., and Asadullah, K. (2005) Dimethylfumarate for psoriasis: more than a dietary curiosity. Trends Mol. Med. 11, 43–48. (24) Slovak, A. J., and Payne, A. R. (1984) Delayed dermal burns caused by dimethyl acetylenedicarboxylate. Contact Dermatitis 11, 29–30. (25) Coffey, R. N., Watson, R. W., Hegarty, N. J., O’Neill, A., Gibbons, N., Brady, H. R., and Fitzpatrick, J. M. (2000) Thiol-mediated apoptosis in prostate carcinoma cells. Cancer 88, 2092–2104. (26) O’Neill, A. J., O’Neill, S., Hegarty, N. J., Coffey, R. N., Gibbons, N., Brady, H., Fitzpatrick, J. M., and Watson, R. W. (2000) Glutathione depletion-induced neutrophil apoptosis is caspase-3-dependent. Shock 14, 605–609. (27) Treumer, F., Zhu, K., Glaser, R., and Mrowietz, U. (2003) Dimethylfumarate is a potent inducer of apoptosis in human T cells. J. InVest. Dermatol. 121, 1383–1388. (28) Ji, C., Amarnath, V., Pietenpol, J. A., and Marnett, L. J. (2001) 4-Hydroxynonenal induces apoptosis via caspase-3 activation and cytochrome c release. Chem. Res. Toxicol. 14, 1090–1096. (29) West, J. D., Ji, C., and Marnett, L. J. (2005) Modulation of DNA fragmentation factor 40 nuclease activity by poly(ADP-ribose) polymerase-1. J. Biol. Chem. 280, 15141–15147. (30) Fischer, U., Janicke, R. U., and Schulze-Osthoff, K. (2003) Many cuts to ruin: a comprehensive update of caspase substrates. Cell Death Differ. 10, 76–100.

88

Chem. Res. Toxicol., Vol. 24, No. 1, 2011

(31) Delaunay, A., Pflieger, D., Barrault, M. B., Vinh, J., and Toledano, M. B. (2002) A thiol peroxidase is an H2O2 receptor and redoxtransducer in gene activation. Cell 111, 471–481. (32) Ma, L. H., Takanishi, C. L., and Wood, M. J. (2007) Molecular mechanism of oxidative stress perception by the Orp1 protein. J. Biol. Chem. 282, 31429–31436. (33) Paulsen, C. E., and Carroll, K. S. (2009) Chemical dissection of an essential redox switch in yeast. Chem. Biol. 16, 217–225. (34) Haynes, R. L., Szweda, L., Pickin, K., Welker, M. E., and Townsend, A. J. (2000) Structure-activity relationships for growth inhibition and induction of apoptosis by 4-hydroxy-2-nonenal in RAW264.7 cells. Mol. Pharmacol. 58, 788–794. (35) Schultz, T. W., Netzeva, T. I., Roberts, D. W., and Cronin, M. T. (2005) Structure-toxicity relationships for the effects to Tetrahymena pyriformis of aliphatic, carbonyl-containing, R,β-unsaturated chemicals. Chem. Res. Toxicol. 18, 330–341. (36) Chan, K., Poon, R., and O’Brien, P. J. (2008) Application of structureactivity relationships to investigate the molecular mechanisms of hepatocyte toxicity and electrophilic reactivity of R,β-unsaturated aldehydes. J. Appl. Toxicol. 28, 1027–1039. (37) Pillon, N. J., Soulere, L., Vella, R. E., Croze, M., Care, B. R., Soula, H. A., Doutheau, A., Lagarde, M., and Soulage, C. O. (2010) Quantitative structure-activity relationship for 4-hydroxy-2-alkenal induced cytotoxicity in L6 muscle cells. Chem.-Biol. Interact. 188, 171–180. (38) Burcham, P. C., Raso, A., Thompson, C., and Tan, D. (2007) Intermolecular protein cross-linking during acrolein toxicity: efficacy of carbonyl scavengers as inhibitors of heat shock protein-90 crosslinking in A549 cells. Chem. Res. Toxicol. 20, 1629–1637.

West et al. (39) Cai, J., Bhatnagar, A., and Pierce, W. M., Jr. (2009) Protein modification by acrolein: formation and stability of cysteine adducts. Chem. Res. Toxicol. 22, 708–716. (40) Lopachin, R. M., Barber, D. S., Geohagen, B. C., Gavin, T., He, D., and Das, S. (2007) Structure-toxicity analysis of type-2 alkenes: in Vitro neurotoxicity. Toxicol. Sci. 95, 136–146. (41) Stewart, B. J., Doorn, J. A., and Petersen, D. R. (2007) Residue-specific adduction of tubulin by 4-hydroxynonenal and 4-oxononenal causes crosslinking and inhibits polymerization. Chem. Res. Toxicol. 20, 1111–1119. (42) Siegel, S. J., Bieschke, J., Powers, E. T., and Kelly, J. W. (2007) The oxidative stress metabolite 4-hydroxynonenal promotes Alzheimer protofibril formation. Biochemistry 46, 1503–1510. (43) Kuhla, B., Haase, C., Flach, K., Luth, H. J., Arendt, T., and Munch, G. (2007) Effect of pseudophosphorylation and cross-linking by lipid peroxidation and advanced glycation end product precursors on τau aggregation and filament formation. J. Biol. Chem. 282, 6984–6991. (44) Qahwash, I. M., Boire, A., Lanning, J., Krausz, T., Pytel, P., and Meredith, S. C. (2007) Site-specific effects of peptide lipidation on β-amyloid aggregation and cytotoxicity. J. Biol. Chem. 282, 36987–36997. (45) Carbone, D. L., Doorn, J. A., Kiebler, Z., Sampey, B. P., and Petersen, D. R. (2004) Inhibition of Hsp72-mediated protein refolding by 4-hydroxy-2-nonenal. Chem. Res. Toxicol. 17, 1459–1467. (46) Carbone, D. L., Doorn, J. A., Kiebler, Z., Ickes, B. R., and Petersen, D. R. (2005) Modification of heat shock protein 90 by 4-hydroxynonenal in a rat model of chronic alcoholic liver disease. J. Pharmacol. Exp. Ther. 315, 8–15. (47) Uchida, K., and Shibata, T. (2008) 15-Deoxy-∆ (12, 14) -prostaglandin J2: an electrophilic trigger of cellular responses. Chem. Res. Toxicol. 21, 138–144.

TX100292N