Identification of the Highly Reactive Cysteine 151 ... - ACS Publications

Feb 18, 2008 - Identification of the Highly Reactive Cysteine 151 in the. Chemopreventive Agent-Sensor Keap1 Protein is Method-Dependent. Aimee L. Egg...
0 downloads 0 Views 2MB Size
Download PDF
1878

Chem. Res. Toxicol. XXXX, xxx, 000

Identification of the Highly Reactive Cysteine 151 in the Chemopreventive Agent-Sensor Keap1 Protein is Method-Dependent Aimee L. Eggler,† Yan Luo,‡ Richard B. van Breemen,‡ and Andrew D. Mesecar*,† The Center for Pharmaceutical Biotechnology, and Department of Medicinal Chemistry and Pharmacognosy, UniVersity of Illinois at Chicago, Chicago, Illinois 60607 ReceiVed June 19, 2007

Upregulation of cytoprotective and detoxifying enzyme expression by small molecules is emerging as an important means of preventing carcinogenesis as well as other diseases. A proposed target of these agents is the Kelch-like ECH-associated protein 1 (Keap1). The vast majority of these agents contain electrophilic moieties, which react with a subset of the 27 cysteines of human Keap1. Modification of these cysteines is proposed to result in nuclear accumulation of transcription factor NF-E2-related factor-2 (Nrf2), a Keap1 binding partner, leading to upregulation of cytoprotective enzymes. The electrophilic agent biotinylated iodoacetamide (BIA) has been used by different laboratories to determine the most reactive cysteines in human Keap1, and the different methods used have generated very different results. In particular, our group has found C151 of human Keap1 to be highly reactive, while others have not identified this cysteine as being even weakly reactive. Nevertheless, C151 is the only cysteine of Keap1 shown thus far in the cell environment to be required to sense chemopreventive agents. In this work, we show that the BIA-modified C151 tryptic peptide is reproducibly detected by our method. We also investigated the key differences in the methods that have been used to prepare the protein for modification by BIA. Removal of the reducing agent from Keap1 before the addition of BIA did not significantly change the modification pattern of Keap1. However, treatment of Keap1 using an ultracentrifugation device in one method resulted in ∼99% of the protein remaining bound to the device at the time of BIA addition. In addition, the resulting pattern of cysteines identified as modified by BIA differed significantly from that obtained by our method. Notably, C151 was no longer detected as modified by BIA. We therefore recommend our method of Keap1 protein preparation for the detection of modified cysteines in proteomic studies. Introduction 1

The transcription factor NF-E2-related factor-2 (Nrf2) plays a crucial role in combating the effects of toxic agents on cells. Nrf2 accumulates in the nucleus in response to oxidative stress or treatment with chemical agents. A battery of cytoprotective enzymes is subsequently upregulated via Nrf2 binding to the antioxidant response element (ARE), a cis-acting sequence located in the 5′ flanking region of these genes. These cytoprotective enzymes include detoxifying enzymes that eliminate toxic agents, such as NAD(P)H:quinone oxidoreductase 1 and glutathione S-transferase, as well as antioxidant proteins, including heme oxygenase-1, that combat the inflammatory oxidizing effect of many toxic agents (1, 2). The Nrf2-deficient mouse has been shown to have a much greater susceptibility to toxic agents, including benzo[a]pyrene (3), the urinary bladderspecific carcinogen N-nitrosobutyl(4-hydroxybutyl)amine (BBN) (4), the hepatotoxin N-acetyl-p-aminophenol (APAP) (5), the * To whom correspondence should be addressed: Center for Pharmaceutical Biotechnology, University of Illinois––Chicago, 900 S. Ashland Ave. M/C 870, Chicago IL, 60607. Telephone: 312-996-1877. Fax: 312413-9303. E-mail: [email protected]. † The Center for Pharmaceutical Biotechnology. ‡ Department of Medicinal Chemistry and Pharmacognosy. 1 Abbreviations: Nrf2, NF-E2-related factor-2; ARE, antioxidant response element; BBN, N-nitrosobutyl(4-hydroxybutyl)amine; APAP, N-acetyl-paminophenol; PCP, pentachlorophenol; Keap1, Kelch-like ECH-associated protein 1; tBHQ, tert-butylhydroquinone; BIA, biotinylated iodoacetamide; TCEP, Tris[2-carboxyethyl]phosphine hydrochloride; LTQ-FT ICR MS, Thermo Finnigan linear ion-trap Fourier-transform ion-cyclotron resonance mass spectrometer.

10.1021/tx700217c CCC: $37.00

environmental carcinogenic pollutant pentachlorophenol (PCP) (6), and diesel exhaust (7). Under basal conditions, Nrf2 is constitutively expressed at high levels but is also constitutively ubiquitinated and degraded such that the overall amount of Nrf2 is maintained at a low level (8). This low level of Nrf2 is sequestered largely in the cytoplasm, preventing constitutive activation of ARE genes. The cytoplasmic sequestration of Nrf2 and its ubiquitination are both mediated by the Keap1 protein (Kelch-like ECH-associated protein 1). A nuclear export signal sequence in Keap1 maintains the Keap1–Nrf2 protein complex in the cytoplasm (9). Keap1 also serves as a bridge between Nrf2 and the Cullin3-based E3ligase ubiquitination complex, leading to ubiquitination of Nrf2 and targeting Nrf2 for degradation by the 26S proteasome (10). Many agents have been identified that stabilize the Nrf2 protein, leading to its nuclear accumulation and subsequent activation of the ARE. These include both agents typically known to be toxic, such as those listed above to which the Nrf2 knockout mouse is susceptible, as well as agents that show cytoprotective activity at an appropriate dose. These agents, investigated most intensively for the ability to prevent carcinogenesis, include ones found in commonly consumed foods, such as sulforaphane from cruciferous vegetables (11). Notably, a common feature shared by this diverse group of agents is that they are overwhelmingly either electrophilic in nature or have metabolites that are electrophilic (12). A major mechanism by which electrophiles are proposed to induce upregulation of the ARE is by modification of a reactive  2007 American Chemical Society

ChemopreVentiVe Agent-Sensor Keap1 Protein

Chem. Res. Toxicol., Vol. xxx, No. xx, XXXX 1879

a very different pattern of modification of Keap1 and, in particular, no detection of the BIA-modified C151 peptide.

Experimental Procedures

Figure 1. Structure of BIA. The asterisk denotes the electrophilic carbon.

Figure 2. Schematic of the two different methods of preparing Keap1 for modification by BIA. In method A, BIA is added directly to the protein in the presence of a reducing agent. After incubation, unreacted BIA is quenched using excess DTT. In method B, the solution is removed from the protein by centrifugation using an ultrafiltration centrifuge device, followed by washing, and then the addition of BIA in the absence of a reducing agent. Unreacted BIA is removed by centrifugation, followed by washing, and finally, the protein on the filter is resuspended in buffer with a reducing agent.

subset of the Keap1 protein’s 27 cysteines (13). Some of the strongest in vivo evidence for this proposal is the fact that cells that express the human Keap1 C151S mutant are no longer responsive to the ARE inducers tert-butylhydroquinone (tBHQ), sulforaphane, and ebselen (10, 14, 15). Using the ARE inducer biotinylated iodoacetamide (BIA) (Figure 1) and mass spectrometric techniques coupled with trypsin digestion, we found cysteine 151 to be among the three most reactive cysteines in human Keap1 (16). Together, these data suggest that modification of cysteine 151 is a crucial part of the mechanism of upregulation of cytoprotective genes for ARE inducers. Independently, Hong et al. (17) reported 15 cysteines of human Keap1 to be modified by BIA, but they did not detect the BIA-modified C151-containing peptide or the corresponding unmodified peptide. One explanation offered by Hong et al. (18) as to why they did not detect modified C151 is that C151 might have quite low reactivity, resulting in a lower abundance of adducts and only sporadic detection of the BIA-modified C151containing peptide. They suggested that, although we observed the modified C151 peptide, our analyses might not have been performed with enough replicates to ascertain the abundance of this peptide. To address this concern and because the nonresponsiveness of the Keap1 C151S mutant to ARE inducers suggests an important electrophile-sensing role for C151, we wanted to evaluate the methods used by both groups to determine the most reactive cysteines of Keap1 to BIA (Figure 2). We show again that the BIA-modified C151-containing peptide is consistently and accurately detected using our method. In addition, we find that, when the Keap1 protein is desalted using an ultrafiltration device to remove the buffer as reported by Hong et al. (17), at least 99% of the protein remains bound to the filter in the device. Therefore, at the time of BIA modification of Keap1 cysteines, most of the Keap1 is not in solution but is instead bound to the filter. We show that this protein preparation method results in

Sample Preparation: “Solution Method”. The BIA (Pierce, trade name EZ-Link Iodoacetyl-LC-Biotin) was prepared as a stock solution at a concentration of 10 mM in dimethylsulfoxide (DMSO). The Keap1 protein [prepared as described previously (16)] stock was at 100 µM in 25 mM Tris-HCl buffer (pH 8.0) containing 2 mM Tris[2-carboxyethyl]phosphine hydrochloride (TCEP) and 20% glycerol (v/v). In the experiments in which the TCEP was removed from the protein before treatment with BIA, the buffer was exchanged to 25 mM Tris-HCl buffer (pH 8.0) using a Centricon 20 device [30 000 molecular weight cut-off (MWCO)] (Millipore) by performing three sequential 180-fold dilutions and concentrations of the protein, yielding a final TCEP concentration of 0.3 nM and a final Keap1 concentration of 100 µM. As a control, the Keap1 protein was also prepared in this manner using 25 mM Tris-HCl buffer (pH 8.0) containing 2 mM TCEP. For the reactions of Keap1 with BIA, Keap1 (10 µM) was incubated with increasing concentrations of BIA, from an equimolar amount to a 10-fold molar excess of BIA relative to Keap1. Where indicated, the Keap1 C151S mutant protein was substituted for wildtype Keap1. The Keap1 C151S mutant was prepared as for the wild-type Keap1 above, except that 50 µM ZnCl2 was added to the cell growth media. Incubations were performed in 25 mM TrisHCl buffer (pH 8.0) at room temperature for 2 h, followed by quenching with 1 mM dithiothreitol (DTT) for 30 min. The reaction was then treated with 3 mM iodoacetamide in excess of the DTT concentration, allowed to react for 30 min, and treated with a final 5 mM DTT addition. To determine the identity of the modification sites, samples were digested with 0.5 µg of trypsin for 3 h at 37 °C. Peptides were either analyzed directly by liquid chromatography–tandem mass spectrometry (LC–MS/MS), as described below, or those peptides that were biotinylated via the reaction of BIA with Keap1 cysteines were first isolated using an immobilizedavidin ICAT column (Applied Biosystems) and then analyzed. Sample Preparation: “Membrane Method”. To analyze the effect of treating the protein by the membrane method (Figure 2), the method and device reported by Hong et al. was used (17). Specifically, a Millipore Ultrafree MC low-binding regenerated cellulose centrifugal spin filter device with a 30 000 MWCO was used. All centrifugation steps were carried out at 12000g for 5 min at room temperature. The devices were first rinsed by adding 200 µL of methanol, followed by centrifugation, and then rinsed with 200 µL of distilled water, followed by centrifugation. Keap1 (70 µg, 10 µL of a 100 µM stock) was then applied to the center of the membrane, followed by centrifugation. A 7 µL aliquot of the filtrate was diluted to 60 µL with 20 mM Tris-HCl buffer (pH 8.0) and then scanned for UV absorption. The Keap1 protein on the membrane was rinsed again by adding 200 µL of 1 M ammonium bicarbonate buffer (pH 8.4), followed by centrifugation as above. Finally, 50 µL of 100 µM BIA in the same buffer [0.3% (v/v) final DMSO concentration] was added to the Keap1 on the membrane, and a pipette was used to gently wash the membrane 10 times to resuspend the protein. The device was then incubated in the dark at 37 °C for 2 h. For samples in which the protein recovery was analyzed, 42 µL of the solution was removed after incubation in the dark, diluted to 62 µL with buffer, and analyzed by UV spectroscopy. To determine any contribution to the absorbance from BIA, a matching dilution was made of BIA that had been incubated alongside the devices in the dark at 37 °C and analyzed by UV spectroscopy. The relative recovered Keap1 protein amounts for the samples were determined as the optical density (OD, 280 nm) measured for each sample, as well as that value minus the OD (280 nm) measured for BIA and the OD for each sample contributed by scattering (360 nm). The (+) control for untreated Keap1 protein was calculated as the OD (280 nm) measured that day for the protein before treatment, normalized to the dilution factor of the samples. The

1880

Chem. Res. Toxicol., Vol. xxx, No. xx, XXXX

results for the OD (280 nm) of the Keap1 that flowed through the device were normalized to the dilution factor of the samples. The corrected values for the (+) control and the filtrate were calculated by subtracting the OD (360 nm) from the OD (280 nm). Each sample analyzed and quantitated by UV was also analyzed by Western blot. Samples (15 µL of the solution analyzed by UV), along with 15 µL of 54 nM Keap1 as a control to represent 0.4% recovery, were loaded onto a freshly poured 10% sodium-dodecylsulfate-reducing polyacrylamide gel and electrophoresed at 190 V for 40 min. The proteins were electrophoretically transferred to polyvinylidene difluoride membranes and probed with Keap1 primary antibody (E-20) and horseradish-peroxidase-conjugated bovine antigoat secondary antibody (both from Santa Cruz Biotechnology) diluted in blocking buffer. Immunoreactive proteins were visualized by enhanced chemiluminescence and exposure to film. For samples that were analyzed by mass spectrometry, after the 2 h of incubation in the dark at 37 °C, the devices were centrifuged as above, followed by the addition of 200 µL of 0.1 M ammonium carbonate buffer and centrifugation. This buffer containing 40 mM TCEP (50 µL) was then added to the device, and the samples were incubated at 50 °C for 15 min. Iodoacetamide (20 µL of 0.2 M) was added, and the devices were incubated another 15 min at 25 °C. Trypsin digestion was performed by adding 1 µg of trypsin overnight at 37 °C. Finally, the peptides were obtained by centrifugation at 12000g, and 1 µL of concentrated formic acid was added before LC–MS/MS analysis. LC–MS/MS. Two different methods of LC–MS/MS analysis were used. In the first, peptides were analyzed using positive ion electrospray LC–MS/MS with a Surveyor high-performance liquid chromatography (HPLC) system interfaced to a Thermo Finnigan (San Jose, CA) LCQ Deca ion-trap mass spectrometer. In addition, where described, peptides were analyzed on a Thermo Finnigan linear ion-trap Fourier-transform ion-cyclotron resonance mass spectrometer (LTQ-FT ICR MS) equipped with a Dionex (Sunnyvale, CA) HPLC system, a Vydac (Hesperia, CA) 218MS LC–MS C18 reversed-phase HPLC column (5 µM, 2.1 × 150 mm, 300 Å), and a 218TP (5 µM, 2.1 mm) guard column. For both LC–MS/MS methods, the mobile phase consisted of a 40 min linear gradient from 95:4.9:0.1 to 4.9:95:0.1 water/acetonitrile/formic acid (v/v/v) at a flow rate of 200 µL/min. Data-dependent MS/MS spectra were acquired, in which the five most abundant peptide ions in each mass spectrum were acquired using collision-induced dissociation with a normalized collision energy of 35%. All LC–MS/MS data were processed using BioWorks 3.3.1 (Thermo Finnigan). Database searches were carried out, allowing 10 ppm mass deviation for each precursor ion. A maximum number of 2 missed cleavages was allowed. Oxidized methionine, carbamidomethylated cysteine, and BIA-labeled cysteine were set as optional amino acid modifications. The modification sites were identified on the basis of the tandem mass spectra of the biotinylated peptides and protein database searching using TurboSEQUEST (Thermo Finnigan). Peptide mass values were searched against the FASTA human protein database (International Protein Index, EMBL-EBI), using a minimum SEQUEST Xcorr score of 1.0.

Results To demonstrate that the C151 residue of Keap1 was indeed modified by BIA in our experiments and that the resulting peptide was reliably detected by mass spectrometry, experiments were performed similarly to those published previously (16). BIA was incubated with either wild-type Keap1 as before or Keap1 protein with a cysteine to serine mutation at position 151. The same ratios of BIA/Keap1 were tested again (1:1, 2:1, and 10:1), and the experiment was performed in triplicate. Two methods of peptide detection were used: (1) analysis after peptide purification on avidin columns as before or (2) analysis of the peptide mixture without avidin purification. In both cases, peptides were analyzed with the LCQ ion-trap mass spectro-

Eggler et al. Table 1. Detection of the BIA-Modified Keap1 C151 Peptidea with avidin

without avidin

[BIA]/[Keap1]

wt Keap1

Keap1 C151S

wt Keap1

Keap1 C151S

1:1 2:1 10:1

1,2,3b 1,2,3 1,2,3

0c 0 0

1,2,3 1,2,3 1,2,3

0 0 0

a Keap1 was modified with BIA following the solution method (Figure 2). The C151 peptide was detected using LC–MS/MS, with an LCQ Deca mass spectrometer. b Indicates that the peptide was found in all three replicates. c The “0” indicates that only the unmodified peptide and not the BIA-modified peptide was detected in the three replicates.

meter. Product ion tandem mass spectrometry was performed on each protonated peptide to confirm its identity. As shown in Table 1, the peptide containing the BIA modification at C151 was detected without exception in the 18 samples for the wildtype protein and was not detected in any of the 18 samples for Keap1 C151S. The mass spectrum of the modified C151 peptide, showing doubly and triply protonated molecules, and the product ion tandem mass spectrum of the doubly charged ion are shown in parts A and B of Figure 3. The mass error associated with the tandem mass spectrum was 1.1, and the SEQUEST Xcorr score was 6.01. These data indicate that the BIA-modified C151 peptide was reliably and accurately detected. Because BIA-modified C151 was consistently and readily detected by our method in these experiments and because of its importance in mediating the signaling mechanism for ARE inducers, we sought to determine the reason that C151 was not detected as being modified by the method of Hong et al. (17). Preparation of the protein for modification by BIA differs significantly between the two methods (Figure 2). Both methods begin with the Keap1 protein in buffer that contains a reducing agent, which should help prevent the oxidation of reactive cysteines and mimic the reducing environment of human cells. As shown in Figure 2A, our method, designated as the solution method, involves the addition of BIA directly to the Keap1 protein in solution without the removal of the reducing agent. The other method, designated here as the membrane method (Figure 2B), uses an ultracentrifugation device with a 30 000 Da MWCO membrane to first remove the solution containing the reducing agent. The BIA reagent is then added to the membrane to allow BIA to react with the Keap1 cysteines. One significant difference between the methods is the presence of a reducing agent at the time of Keap1 modification by BIA. To determine whether the removal of the reducing agent leads to the oxidation of C151 and, hence, the inability of this cysteine to be modified by an alkylating agent, the reducing agent TCEP was removed from the Keap1 protein preparation prior to the addition of BIA by buffer exchange using an ultrafiltration device. However, unlike in the membrane method, the Keap1 protein was not left on the filter after centrifugation but remained in solution at all times during the removal of the reducing agent. To allow sufficient time for any cysteine oxidation to occur, samples were exposed to the air for 3 h after the removal of TCEP before the addition of BIA. As a control, Keap1 was treated in the same manner, but TCEP was included in the buffer. Both samples were then modified immediately after TCEP removal with BIA at a 1:1 ratio relative to Keap1 and analyzed as before by LC–MS/MS, without avidin, with runs performed in triplicate. As shown in Table 2, very similar results were obtained whether or not TCEP was present at the time of BIA addition. Because the removal of the reducing agent was not found to cause the oxidation of C151, we decided to follow the membrane

ChemopreVentiVe Agent-Sensor Keap1 Protein

Chem. Res. Toxicol., Vol. xxx, No. xx, XXXX 1881

Figure 3. Detection of the modified peptide C151VLHVMNGAVMYQIDSVVR by mass spectrometry. (A) Positive ion electrospray mass spectrum obtained using the LCQ mass spectrometer showing the doubly and triply protonated peptides. It should be noted that the mass of BIA was 510.1 units, and alkylation of a protein cysteine sulfhydryl group by BIA would result in an increase of 382.2 units (BIA – HI). The mass of the unmodified peptide was 2133 units, and modification by BIA resulted in the m/z values shown. (B) Product ion tandem mass spectrum of the doubly protonated C151 peptide with the b and y sequence ions indicated, obtained using the LCQ mass spectrometer. Note that, because C151 occurred at the N terminus of this peptide, all of the b-type ions contained modified C151. None of the y-type fragment ions that were detected contained modified C151. (C) Positive ion electrospray mass spectrum obtained using the LTQ-FT mass spectrometer showing the triply protonated peptide. (D) Product ion tandem mass spectrum of the triply protonated C151 peptide of m/z 839.4321 obtained using the LTQ-FT mass spectrometer.

Table 2. Detection of BIA-Modified Peptides after the Removal of TCEP from Keap1 Cys-151 Cys-226 Cys-257 Cys-273 Cys-288 Cys-297 Cys-319 Cys-613 Cys-23 Cys-38 Cys-14 Cys-241 Cys-368 Cys-622 Cys-624

TCEP present

TCEP removed

1,2,3,4a 1,2,3,4 1,2,3,4 2 1,2,3,4 1,2,3,4 1,2,3,4 1,2,3,4 3 2,4 1,3 1,2,3,4

1,2,3,4 1,2,3,4 1,2,3,4 4 1,2,3,4 1,2,3,4 1,2,3,4 1,2,3,4 1,4 1,3 1,2,3 1,2,3,4 2,3 3,4 1

2,4 1,3

a Indicates in which experiments, out of four replicates, the BIA-modified peptide was found. Modified peptides were detected using LC–MS/MS with a LCQ Deca mass spectrometer.

method to determine whether, in our hands, C151 was no longer detected as being modified by BIA. After following the method and analyzing the recovered solution using electrospray mass spectrometry with a LCQ ion-trap instrument, we were surprised to find that no peptides were detected, either as modified or unmodified. This suggested that either Keap1 had flowed through the device or bound to the membrane. Therefore, the amount of Keap1 protein present in the filtrate and in solution at the time of BIA modification in the membrane method was analyzed. The method was followed to the point at which the protein had incubated with BIA for 2 h. The samples were then analyzed by measuring their UV absorbances at 280 nm to detect Keap1 protein, as described in the Experimental Procedures. As shown in Figure 4A, surprisingly, at the time of BIA

addition, most of the Keap1 protein remained on the membrane. Only ∼1% of the protein was detected in the filtrate, and less than 1% was detected in the recovered samples. To further examine Keap1 protein amounts present in the samples at the time of incubation with BIA, the recovered samples were analyzed by Western blot using an antibody against the Keap1 protein. To obtain a rough estimate of the amount of Keap1 protein present in the samples, an amount of Keap1 equivalent to 0.4% of the initial protein amount was included in the analysis. As shown in Figure 4B, Keap1 protein could be detected in four of the five samples but only at very low amounts, well below 0.4% of the initial protein amount. The amount of recovered peptide from the membrane method was too low to detect by using electrospray mass spectrometry with the LCQ ion-trap instrument. However, the sensitivity of this circular ion trap is relatively low compared to that of a linear ion-trap mass spectrometer. Therefore, the highly sensitive linear ion trap of the LTQ-FT ICR MS was used in an attempt to detect the peptides recovered from the membrane method. We also analyzed the peptides resulting from our method, omitting avidin purification of peptides. A BIA/Keap1 ratio of 5:1 was used in both experiments. As shown in Table 3, we were indeed successful in detecting numerous peptides after the membrane method using the LTQ-FT ICR. Results are reported as peptides detected in three replicate experiments. For each experiment, peptides could be detected as modified by BIA, modified by IA, or detected as the unmodified peptide (NM). Using the solution method, all 27 cysteines were detected as modified by either BIA or IA and all but two were also detected as unmodified. C151 was detected as modified by BIA in all three experiments and was also detected as modified by IA or unmodified in two experiments. The mass spectrum of the triply charged C151-modified peptide obtained by using the LTQ-FT

1882

Chem. Res. Toxicol., Vol. xxx, No. xx, XXXX

Eggler et al. Table 3. Detection of Peptides after Keap1 Treatment by the “Solution” or “Membrane” Methoda solution method

membrane method

BIAb

IAb

NMb

BIA

IA

NM

1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 2,3 2,3 2 2 2 0 0 0 0 0 0 0 0 0 0 0 0 0

1,2 1,2 0 1,2 1,2 2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 2 2 1,2 1,2 1,2 2

2,3 3 2,3 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 0 0 3 3

0 1,2,3 1,2,3 1,2 1,2,3 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2 1,2,3 1,2 1,2 1,3 1,2,3 1,2 1,2 1,2 0 0 0 0 0 0

3 1,2,3 0 1,3 1,2,3 2,3 1 1,2,3 1,3 1,3 1,3 1,3 1,3 1,2,3 1,2,3 2,3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 3 3 0 0 1,2,3 2,3

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

Cys-151 Cys-226 Cys-257 Cys-273 Cys-288 Cys-297 Cys-319 Cys-434 Cys-613 Cys-23 Cys-38 Cys-13 Cys-14 Cys-77 Cys-241 Cys-249 Cys-368 Cys-489 Cys-518 Cys-622 Cys-624 Cys-171 Cys-196 Cys-395 Cys-406 Cys-513 Cys-583

a A 5:1 molar ratio of BIA/Keap1 was used in the experiments, which were performed in triplicate. Modified peptides were detected using LC– MS/MS with a LTQ-FT ICR MS. b Indicates in which experiments the BIA-modified peptide, the IA-modified peptide, and the unmodified peptide, respectively, were found.

Table 4. Order of Reactivity of Keap1 Cysteines by the Solution Method Using a LTQ-FT ICR MS BIA/Keap1 0.025

Figure 4. Detection of Keap1 protein using the membrane method at the time of incubation with BIA. (A) Keap1 concentrations were detected by UV spectroscopy. The empty bars indicate the measured OD at 280 nm, and the gray bars show the corrected OD, with contributions from scattering at 360 nm and BIA at 280 nm (0.012 OD) subtracted. The (+) control indicates the protein concentration before the method. FT indicates the filtrate from the device. (B) Recovered protein samples were analyzed by SDS–PAGE/Western blot and probed with Keap1 antibody. As a standard, 0.81 pmol of Keap1 was included in the far right lane.

ICR is shown in Figure 3C, and the tandem mass spectra, with a mass error of 0.02 and a SEQUEST Xcorr score of 6.09, is shown in Figure 3D. In contrast, the membrane method produced a very different pattern of modification compared to the solution method (Table 3). Most notably, C151 was not detected as modified by BIA in any of the replicate experiments. The C151-containing peptide was detected only once, and this was as the IA-modified peptide. This finding agrees with what was observed previously, that the BIA-modified C151-containing peptide is absent when the membrane method is used to prepare the protein for modification (17). In addition, no peptides were detected in their unmodified form, whereas in our experiment, all but two cysteine-containing peptides were detected in their unmodified form.

C23 C38 C77 C151 C226 C241 C257 C273 C288 C297 C319 C368 C434 C489 C613 C622 C624

0.05

0.1

0.5

1

2

5 1,2,3a 1,2,3

1,2,3

1,2,3

1,2,3

1,2,3 1,2,3 1,2,3

1,2,3 1,2,3 3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3

1,2,3 1,2,3 3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3

1,2,3

1,3

1,3

1,2,3

1,2,3

1,2,3

1,2,3 1,2,3 1,2,3

1,2,3

1,2,3

1,2,3

1,2,3

2,3

1,2,3

1,2,3

1,2,3

1,2,3 1,2,3 3 1,2,3 1,2,3 1,2,3 1,2,3 1,2,3 3 1,2,3 1,2,3 3 3

a Indicates in which experiments, out of three replicates, the BIA-modified peptide was found.

Finally, we wanted to verify that C151 was still among the most reactive of the Keap1 cysteines when the samples were analyzed using the LTQ-FT ICR MS and avidin purification was omitted. We analyzed the modification pattern of Keap1 by BIA at ratios ranging from 0.025 to 5 molecules of BIA per Keap1 molecule. As shown in Table 4, even at the lowest ratio, three cysteines were detected as modified by BIA, C151, C288, and C319. At the highest ratio of 5:1, 15 cysteines were detected as modified by BIA, and there is good agreement with the data obtained in a separate experiment for Table 3 by our method.

ChemopreVentiVe Agent-Sensor Keap1 Protein

Discussion Our main finding is that detection of modified cysteines of the Keap1 protein is highly dependent upon whether the protein is in solution at the time of modification or bound to an ultrafiltration membrane. In particular, the highly reactive and biologically significant cysteine, C151, is not detected as modified by the alkylation agent BIA after the membrane method is used to prepare the protein for the reaction with BIA. We also find that the removal of the reducing agent 3 h prior to the reaction of BIA with Keap1 cysteines does not alter the modification pattern significantly. Therefore, because of the relatively harsh treatment of Keap1 in the membrane method, Keap1 is probably predominantly in an unfolded or misfolded state at the time of BIA modification, resulting in a different modification pattern. The Keap1 protein in our experiments is likely in its properly folded state, because it is capable both of binding to Nrf2 and targeting Nrf2 for Cul3-based ubiqutination (data not shown). Therefore, when the identity of reactive cysteines in properly folded Keap1 by various agents of interest is to be determined, we recommend using the solution method over the membrane method to identify modification patterns of Keap1. Another observed difference between the solution and membrane methods is that no unmodified peptides were detected in the membrane method. A likely explanation for this result is that we quench the iodoacetamide with excess DTT before performing a tryptic digest, so that no free iodoacetamide is present to modify cysteines exposed by digestion, whereas iodoacetamide is still present throughout the tryptic digest in the membrane method, allowing modification of exposed cysteines. Our detection of C151 (16) as one of the most highly reactive cysteines in human Keap1 agreed well with the known biological data at the time of publication for a key role for C151 in sensing electrophilic ARE inducers. The Hannink laboratory showed that cells transfected with the C151S Keap1 mutant were largely unresponsive to two ARE inducers, tBHQ, a quinone, and sulforaphane, a promisingly chemopreventive agent with an isothiocyanate reactive moiety (14). Recently, two other groups have found similar results for ebselen, a seleno-organic drug (15), and neurite outgrowth-promoting prostaglandin compounds (19), which protect neurons from oxidative insults. The importance of C151 in detection of ARE inducers with a wide range of chemistries is supported by our recent findings on three ARE inducers from food sources, xanthohumol from hops, isoliquiritigenin from licorice, and 10-shogaol from ginger (21). While these three Michael acceptors were found to modify slightly different subsets of Keap1 cysteines, for all three C151 was found to be highly reactive. The downstream effect of the modification of C151 is still unknown. C151 lies in the BTB domain of Keap1, which is responsible for both homodimerization and mediating the interaction with the Cul3 protein. It was previously thought that modification of Keap1 cysteines lead to disruption of the interaction of Keap1 with the Neh2 domain of Nrf2. However, we showed that there is no disruption upon the modification of Keap1 cysteines, including C151 (16). A second possibility is that the modification of C151 alters the ubiquitination of either Keap1 or Nrf2. The interaction of the BTB domain with the Cul3 protein has been observed to weaken upon treatment of cells with ARE inducers (10, 20), and importantly, the Hannink group showed that this disruption was dependent upon C151. Therefore, it seems likely that the Keap1–Cul3 interaction is

Chem. Res. Toxicol., Vol. xxx, No. xx, XXXX 1883

altered by C151 modification. We are currently investigating this possibility. Acknowledgment. This work was supported by NIH grants 5 P01 CA48112 and P50 AT000155. We acknowledge the use of the CBC/UIC RRC Proteomics and Informatics Services Facility, which was established by a grant from The Searle Funds at The Chicago Community Trust to the Chicago Biomedical Consortium. Note Added after Print Publication. Because of a production error, the data in Table 3 were incorrect in the version posted on the Web October 13, 2007 (ASAP) and published in the December 2007 issue (Vol. 20, No. 12, pp 1878–1884); the correct electronic version of the paper was published on February 18, 2008, and an Addition and Correction appears in the February 2008 issue (Vol. 21, No. 2).

References (1) Kobayashi, M., and Yamamoto, M. (2005) Molecular mechanisms activating the Nrf2–Keap1 pathway of antioxidant gene regulation. Antioxid. Redox Signaling 7, 385–394. (2) Talalay, P., Dinkova-Kostova, A. T., and Holtzclaw, W. D. (2003) Importance of phase 2 gene regulation in protection against electrophile and reactive oxygen toxicity and carcinogenesis. AdV. Enzyme Regul. 43, 121–134. (3) Ramos-Gomez, M., Kwak, M. K., Dolan, P. M., Itoh, K., Yamamoto, M., Talalay, P., and Kensler, T. W. (2001) Sensitivity to carcinogenesis is increased and chemoprotective efficacy of enzyme inducers is lost in Nrf2 transcription factor-deficient mice. Proc. Natl. Acad. Sci. U.S.A. 98, 3410–3415. (4) Iida, K., Itoh, K., Kumagai, Y., Oyasu, R., Hattori, K., Kawai, K., Shimazui, T., Akaza, H., and Yamamoto, M. (2004) Nrf2 is essential for the chemopreventive efficacy of oltipraz against urinary bladder carcinogenesis. Cancer Res. 64, 6424–6431. (5) Chan, K., Han, X. D., and Kan, Y. W. (2001) An important function of Nrf2 in combating oxidative stress: Detoxification of acetaminophen. Proc. Natl. Acad. Sci. U.S.A. 98, 4611–4616. (6) Umemura, T., Kuroiwa, Y., Kitamura, Y., Ishii, Y., Kanki, K., Kodama, Y., Itoh, K., Yamamoto, M., Nishikawa, A., and Hirose, M. (2006) A crucial role of Nrf2 in in vivo defense against oxidative damage by an environmental pollutant, pentachlorophenol. Toxicol. Sci. 90, 111– 119. (7) Aoki, Y., Sato, H., Nishimura, N., Takahashi, S., Itoh, K., and Yamamoto, M. (2001) Accelerated DNA adduct formation in the lung of the Nrf2 knockout mouse exposed to diesel exhaust. Toxicol. Appl. Pharmacol. 173, 154–160. (8) Nguyen, T., Sherratt, P. J., Huang, H.-C., Yang, C. S., and Pickett, C. B. (2003) Increased protein stability as a mechanism that enhances Nrf2-mediated transcriptional activation of the antioxidant response element. Degradation of Nrf2 by the 26 S proteasome. J. Biol. Chem. 278, 4536–4541. (9) Velichkova, M., and Hasson, T. (2005) Keap1 regulates the oxidationsensitive shuttling of Nrf2 into and out of the nucleus via a Crm1dependent nuclear export mechanism. Mol. Cell. Biol. 25, 4501–4513. (10) Zhang, D. D., Lo, S.-C., Cross, J. V., Templeton, D. J., and Hannink, M. (2004) Keap1 is a redox-regulated substrate adaptor protein for a Cul3-dependent ubiquitin ligase complex. Mol. Cell. Biol. 24, 10941– 10953. (11) Fahey, J. W., Zhang, Y., and Talalay, P. (1997) Broccoli sprouts: An exceptionally rich source of inducers of enzymes that protect against chemical carcinogens. Proc. Natl. Acad. Sci. U.S.A. 94, 10367–10372. (12) Talalay, P., De Long, M. J., and Prochaska, H. J. (1988) Identification of a common chemical signal regulating the induction of enzymes that protect against chemical carcinogenesis. Proc. Natl. Acad. Sci. U.S.A. 85, 8261–8265. (13) Dinkova-Kostova, A. T., Holtzclaw, W. D., Cole, R. N., Itoh, K., Wakabayashi, N., Katoh, Y., Yamamoto, M., and Talalay, P. (2002) Direct evidence that sulfhydryl groups of Keap1 are the sensors regulating induction of phase 2 enzymes that protect against carcinogens and oxidants. Proc. Natl. Acad. Sci. U.S.A. 99, 11908–11913. (14) Zhang, D. D., and Hannink, M. (2003) Distinct cysteine residues in Keap1 are required for Keap1-dependent ubiquitination of Nrf2 and for stabilization of Nrf2 by chemopreventive agents and oxidative stress. Mol. Cell. Biol. 23, 8137–8151. (15) Sakurai, T., Kanayama, M., Shibata, T., Itoh, K., Kobayashi, A., Yamamoto, M., and Uchida, K. (2006) Ebselen, a seleno-organic antioxidant, as an electrophile. Chem. Res. Toxicol. 19, 1196–1204.

1884

Chem. Res. Toxicol., Vol. xxx, No. xx, XXXX

(16) Eggler, A. L., Liu, G., Pezzuto, J. M., van Breemen, R. B., and Mesecar, A. D. (2005) Modifying specific cysteines of the electrophilesensing human Keap1 protein is insufficient to disrupt binding to the Nrf2 domain Neh2. Proc. Natl. Acad. Sci. U.S.A. 102, 10070–10075. (17) Hong, F., Sekhar, K. R., Freeman, M. L., and Liebler, D. C. (2005) Specific patterns of electrophile adduction trigger Keap1 ubiquitination and Nrf2 activation. J. Biol. Chem. 280, 31768–31775. (18) Hong, F., Freeman, M. L., and Liebler, D. C. (2005) Identification of sensor cysteines in human Keap1 modified by the cancer chemopreventive agent sulforaphane. Chem. Res. Toxicol. 18, 1917–1926. (19) Satoh, T., Okamoto, S. i., Cui, J., Watanabe, Y., Furuta, K., Suzuki, M., Tohyama, K., and Lipton, S. A. (2006) From the cover: Activation

Eggler et al. of the Keap1/Nrf2 pathway for neuroprotection by electrophillic phase II inducers. Proc. Natl. Acad. Sci. U.S.A. 103, 768–773. (20) Gao, L., Wang, J., Sekhar, K. R., Yin, H., Yared, N. F., Schneider, S. N., Sasi, S., Dalton, T. P., Anderson, M. E., Chan, J. Y., Morrow, J. D., and Freeman, M. L. (2007) Novel n-3 fatty acid oxidation products activate Nrf2 by destabilizing the association between Keap1 and Cullin3. J. Biol. Chem. 282, 2529–2537. (21) Luo, Y., Eggler, A. L., Liu, D., Liu, G., Mesecar, A. D., and van Breemen, R. B. (2007) Sites of alkylation of human Keap1 by natural chemoprevention agents. J. Am. Soc. Mass Spectrom., in press.

TX700217C