Induction of Complementary Function ... - Wiley Online Library

6 downloads 0 Views 1MB Size Report
Jul 25, 2014 - with or without 10 µM dicoumarol present in the as- say system. The dicoumarol-sensitive part of the ac- tivity was calculated as the slope of the ...
J BIOCHEM MOLECULAR TOXICOLOGY Volume 29, Number 1, 2015

Induction of Complementary Function Reductase Enzymes in Colon Cancer Cells by Dithiole-3-thione versus Sodium Selenite Melanie M. Erzinger,1 C´edric Bovet,1 Anuli Uzozie,2 and Shana J. Sturla1 1 Department

of Health Sciences and Technology, Institute of Food, Nutrition and Health, ETH Zurich 8092, Zurich, Switzerland; E-mail: [email protected] 2 Institute of Molecular Cancer Research, University of Zurich 8057, Zurich, Switzerland Received 14 May 2014; revised 23 July 2014; accepted 25 July 2014

ABSTRACT: Cellular induction of reductase enzymes can alter the susceptibility of cells toward drugs and chemicals. In this study, we compared the capacity of a single dose of sodium selenite and 3H-1,2dithiole-3-thione (D3T) to influence the drug-relevant reducing capacity of HT29 cells over time, and defined the protein-specific contribution to this activity on the basis of selected reaction monitoring mass spectrometry. Thioredoxin reductase 1 (TrxR1) protein levels and activity were inducible up to 2.2-fold by selenium. In contrast, selenium had only a minor influence on prostaglandin reductase 1 (PTGR1) and NAD(P)H:quinone oxidoreductase 1 (NQO1) activity and protein levels. D3T, a strong Nrf2 inducer, induced all the reductases and additionally increased the cytotoxicity of hydroxymethylacylfulvene, a bioreductive DNA-alkylating drug. The data and experimental approaches allow one to define induction potency for reductase enzymes PTGR1, TrxR1, and NQO1 in HT29 cells and link these to changes in drug cytotoxicity.  C 2014 Wiley Periodicals, Inc. J. Biochem. Mol. Toxicol. 29:10–20, 2015; View this article online at wileyonlinelibrary.com. DOI 10.1002/jbt.21601

Enzyme Induction; Reductase; Enzyme Activity; Biotransformation Efficiency; Quantitative Proteomics

KEYWORDS:

Correspondence to: Shana Sturla. Contract Grant Sponsor: Swiss National Science Foundation. Contract Grant Number: CRSII3_136247. Contract Grant Sponsor: US National Cancer Institute. Contract Grant Sponsor: R01 CA123007. Supporting information is available in the online issue at wileyonlinelibrary.com.  C 2014 Wiley Periodicals, Inc.

INTRODUCTION Reductase enzymes involved in detoxification or activation of xenobiotics can impact the susceptibility of cells toward drugs, chemicals, and carcinogens. Examples include prostaglandin reductase 1 (PTGR1), thioredoxin reductase 1 (TrxR1), and NAD(P)H:quinone oxidoreductase 1 (NQO1). Each catalyzes a complementary chemical transformation and is regulated in a different manner. PTGR1, a cytosolic 2e− enone reductase, inactivates prostaglandins and related eicosanoids [1, 2], and activates acylfulvenes, a group of bioreductive DNA-alkylating agents [3, 4]. The inducibility of PTGR1 by 3H-1,2-dithiole-3-thione (D3T) in rat liver was explained by transcription regulation via the Nrf2-Keap1 pathway [5, 6]. TrxR1, on the other hand, is a selenocysteine-containing disulfide oxidoreductase [7, 8] with endogenous substrates such as oxidized thioredoxins [8], vitamin K3 [9] and lipoic acid [10], and considered as a key in redox control of cellular function [11]. TrxR1 expression can be controlled on the transcriptional level, posttranscriptional, or posttranslational level [11], and depends on an adequate supply of selenium [11]. Lastly, NQO1 is a homodimeric flavoprotein with the following three main physiological functions: it detoxifies quinones, maintains endogenous antioxidants in their reduced and active forms, and regulates the stability of the tumor suppressor protein p53 [12]. Additionally, it is involved in the activation of mitomycin C (MMC), which is used for treating several cancers [13, 14]. The antioxidant response element and the xenobiotic response element, two transcription regulation elements found in the 5 flanking region of the NQO1 gene, are responsible for inducing or repressing its induction [12]. The activity of such reducing enzymes is of interest in cancer chemoprevention and modulated via 10

Volume 29, Number 1, 2015

INDUCTION OF REDUCTASES

transcriptional activation by environmental factors such as bioactive small molecules in the diet. Among these, the dietary element selenium (Se) modulates selenoprotein (e.g., TrxR1) function and there is an inverse relationship between balanced Se intake and the risk of human cancers, such as lung, prostate, and colon cancer [15–17]. Dithiolethiones of natural and synthetic origin, such as D3T or the widely studied chemopreventive agent oltipraz, are strong inducers of enzymes regulated through the Nrf2-Keap1 pathway, such as PTGR1 or NQO1 [18], and can therefore inhibit carcinogenesis in various organs, including colon, kidney, liver, and lung in rodents [19]. On the basis of chemical structures and properties, each compound can interact with cellular regulatory pathways differently, stimulating a complement of enzymes with varied xenobiotic substrate profiles. The potential for pharmacological induction of reductase enzymes suggests opportunities for combination therapies or chemoprevention [20], but there are knowledge gaps regarding quantitative relationships between biotransformation efficiency and how particular proteins give rise to a phenotype attributable to bioactive agents. Modern quantitative proteomicsoriented strategies allow one to quantify specific proteins in complex biological fluids [21, 22]. Selected reaction monitoring (SRM) has demonstrated a reproducible and accurate capability of multiplex protein quantification in complex biological samples [23–26]. In contrast, activities of NADPH-dependent cellular reductase enzymes are traditionally assayed on the basis of spectroscopic monitoring of the transformation of chemically reactive probes, mostly targeting other functionally complementary enzymes. In this study, we characterized the expression levels over time of PTGR1, TrxR1, and NQO1 in HT29 human colon cancer cells treated with one single dose of either D3T or selenite by integrating data from SRM versus general activity measurements. We describe how the impact of functionally and structurally distinct transcription-activating compounds may be distinguished on the basis of reductase activities versus protein expression profiles, and also analyzed their influence on cancer drug cytotoxicity.

MATERIALS AND METHODS Chemicals and Reagents Chemicals and antibodies were purchased from Sigma–Aldrich (St. Louis, MO or Steinheim, Germany) unless otherwise specified. All solvents and chemicals were of analytical reagent grade or higher. The stable isotope-labeled peptides used as internal standard in the SRM experiments were obtained from Thermo Fisher Scientific (Ulm, Germany) at a concenJ Biochem Molecular Toxicology

DOI 10.1002/jbt

11

tration of 5 µM (AQUA peptide, concentration ± 25%) or JPT Peptide Technologies (SpikeTidesTM L, Berlin, Germany) at an unknown concentration. These peptides were modified with heavy labeled Lys8 or Arg10 at the respective C-terminus. iRT peptides were ob¨ tained from Biognosys (Zurich, Switzerland).

Cell Culture HT29 cells were provided by Professor Christophe Lacroix and Professor Leo Meile (ETH Zurich) or obtained commercially from the Leibnitz-Institut DSMZ (Braunschweig, Germany). They were grown in Roswell Park Memorial Institute 1640 medium with GlutaMAXTM or Dulbecco’s modified Eagle medium (Life Technologies, Grand Island, NY) supplemented with 10% (v/v) fetal bovine serum (Life Technologies), 1% (v/v) penicillin–streptomycin antibiotic solution (Life Technologies), and incubated at 37°C in a humidified atmosphere containing 5% CO2 . Stocks of Na2 SeO3 and D3T were prepared in water and dimethylsulfoxide (DMSO), respectively, and were added to the medium (final DMSO concentration ࣘ 0.1%). For all samples, 106 cells were seeded in 10-cm dishes and treated once 24 h later. Cells were then allowed to grow under these conditions for the time indicated. At the time of harvest, cells were ࣘ 80% confluent. Samples were prepared in duplicate and used for enzyme activity and SRM measurements (data presented in Table 1). We note that in the case of the 48 h time point for sodium selenite treatment, cell samples were derived from particularly distant cell passages (data presented in Table 1). Finally, in a smaller follow-up experiment concerning Western blot analysis, samples were prepared in triplicate and analyzed by both SRM and Western blot (data presented in Figures 1 and 2 and Table 2).

Cell Lysate Preparation HT29 cells were trypsinized, washed with 1× phosphate buffered saline (PBS), and incubated at 4°C for 15 min with Tris–HCl lysis buffer (pH 7.4) containing 150 mM NaCl, 1 mM EDTA, and one tablet per 10 mL of Complete Mini (Roche Diagnostics, Rotkreuz, Switzerland). The cell suspensions were sonicated with a sonicator tip (Sonics & Materials, Newtown, CT). The protein extracts were centrifuged (14,000 rcf) at 4°C for 15 min, and the total protein concentration of the supernatants was determined with the bicinchoninic acid protein assay (Thermo Fisher Scientific, Rockford, IL).

Enzyme Activity All activity assays were performed using a UV– vis spectrophotometer (Varian Cary-100, now Agilent Technologies, Santa Clara, CA) and performed in

12

ERZINGER ET AL.

Volume 29, Number 1, 2015

TABLE 1. Fold Changes Calculated from Enzyme Activity Data and Protein Levels (SRM Data)a Enone Reduction/PTGR1

Na2 SeO3 D3T a b

Activity SRMb Activity SRMb

Disulfide Reduction/TrxR1

Quinone Reduction/NQO1

24 h

48 h

72 h

24 h

48 h

72 h

24 h

48 h

72 h

1.2* 1.1 1.6*** 1.5***

1.0 1.2** 1.6*** 1.6***

1.6 1.1 1.5*** 1.9***

2.0*** 1.7*** 1.5*** 1.9***

1.7** 2.2*** 1.4 2.4***

2.2* 2.1*** 1.3*** 2.3***

1.3* 1.2*** 1.8** 1.4***

1.6* 1.4*** 2.1 1.9***

1.1 1.3*** 1.7* 2.3***

Numbers are fold changes between control and treated samples. *p value ࣘ 0.05, **p value ࣘ 0.01; ***p value ࣘ 0.001. Fold changes were obtained by statistical analysis of two proteotypic peptides per protein (see Materials and Methods).

triplicate. Statistical analysis was performed by using the unpaired two sample t-test.

Enone Reduction Activity Assay Enone reduction (ER) activity was targeted as a measurement for PTGR1 activity and was measured with trans-2-nonenal as a substrate and NADPH as a cofactor by adaptation of a previously reported method for evaluating PTGR1 enzyme kinetics [4]. The assay buffer consisted of 0.5× PBS, 150 µM NADPH (Calbiochem, San Diego, CA), and 500 µM trans-2nonenal. Fifty micrograms of total protein was added to 1 mL of assay buffer. The reaction was initiated by the addition of substrate. The decrease in absorbance at 340 nm, due to the consumption of NADPH, was measured over 5 min, and the activity was calculated as the slope of the decrease in absorbance per minute.

Disulfide Reduction Activity Assay Disulfide reduction (DR) activity was targeted as a measurement for TrxR1 activity. Assays were performed by modifying the procedures of Holmgren and ¨ Bjornstedt [7]. Briefly, the substrate, 5,5 -dithiobis(2nitrobenzoic acid) (DTNB), was reduced by DR to 5 thionitrobenzoic acid. The assay buffer consisted of 50 mM Tris–HCl (pH 7.4) containing 1 mM EDTA, 2 mM DTNB, 200 µM NADPH (Calbiochem), and 200 µg/mL bovine serum albumin. One hundred micrograms of total protein was added to 1 mL of assay buffer. The increase in absorbance at 412 nm, due to product formation, was measured over 3 min, and the activity was calculated as the slope of the increase in absorbance per minute.

Quinone Reduction Activity Assay Quinone reduction (QR) activity was targeted as a measurement for NQO1 activity. Assays were performed by modifying the procedures of Ernster [27]. Briefly, cytosolic NQO1 activity was measured using 2,6-dichloroindophenol (2,6-DCIP) as a substrate. The

decrease in absorbance at 600 nm, due to the reduction of 2,6-DCIP, was measured. The assay buffer consisted of 25 mM Tris–HCl (pH 7.4) containing 0.2 mM NADPH (Calbiochem), 5 µM flavin adenine dinucleotide, 0.1 mM 2,6-DCIP, 0.6 mg/L bovine serum albumin, and 0.1 µL/mL Tween-20. For each sample, 5 µg of total protein was added to 1 mL of assay buffer, with or without 10 µM dicoumarol present in the assay system. The dicoumarol-sensitive part of the activity was calculated as the slope of the decrease in absorbance per minute.

SRM Analysis Twenty five to seventy five micrograms of total protein was denatured by diluting the protein solution with 8 M urea dissolved in 50 mM ammonium bicarbonate (pH 7.8), reduced with 5 mM dithiothreitol (DTT) at 37°C for 30 min, and alkylated with 15 mM iodoacetamide at room temperature for 30 min in the dark. The reaction was stopped by adding DTT to get a final concentration of 15 mM and allowing the mixture to react at room temperature for 10 min in the dark. Samples were diluted with 50 mM ammonium bicarbonate (pH 7.8) to a final concentration of 0.6 M urea, and 3% (v/v) methanol was added. The proteins were digested at 37°C with sequencing grade modified trypsin (Promega, Madison, WI) at a final enzyme:substrate ratio of 1:20 for 15 h. The digestion was stopped by adding trifluoroacetic acid to a final concentration of 0.5% (v/v). Protein digests were spiked with the isotope-labeled peptides and desalted using reversed-phase cartridges Sep-Pak tC18 (Waters, Milford, MA). The peptides were lyophilized until dryness and dissolved at a concentration of 0.25 or 1 µg/µL with the iRT peptides diluted in 3% (v/v) acetonitrile acidified with 0.1% (v/v) formic acid. One microgram of total digest was analyzed in duplicate on a triple quadrupole mass spectrometer (TSQ Vantage, Thermo Fisher Scientific, San Jose, CA) coupled to a nanoLC system (nanoAcquity, Waters) and corresponding results were averaged. Peptides were trapped on a Symmetry C18 column (180 µm × J Biochem Molecular Toxicology

DOI 10.1002/jbt

Volume 29, Number 1, 2015

INDUCTION OF REDUCTASES

13

(a) PTGR 1 L/H peptide intensity

L/H peptide intensity

5 4 3 2 1 0

4 3 2 1 0

0 μM D3T

0 μM Na2SeO3 2 μM Na2SeO3

30 μM D3T

(b) TRXR 1 5

L/H peptide intensity

L/H peptide intensity

5 4 3 2 1 0

4 3 2 1 0

0 μM D3T

30 μM D3T

0 μM Na2SeO3 2 μM Na2SeO3

(c) NQO1 5

L/H peptide intensity

L/H peptide intensity

5 4 3 2 1 0

4 3 2 1 0

0 μM D3T

30 μM D3T

0 μM Na2SeO3 2 μM Na2SeO3

FIGURE 1. Light-to-heavy ratio (L/H) of proteotypic peptide intensities for (a) PTGR1 (HFVGYPTNSDEFELK), (b) TrxR1 (VVYENAYGQFIGPHR), and (c) NQO1 (EGHLSPDIVAEQK) measured for HT29 cells incubated 48 h with 0 versus 2 µM Na2 SeO3 or 0 versus 30 µM D3T. Cell samples were prepared in triplicate and measurements were performed in duplicate. For simplification, L/H ratio of only one proteotypic peptide per protein is shown. Horizontal lines show means; error bars indicated standard deviations from means.

20 mm, 5 µm, Waters) at a flow rate of 5 µL/min for 3 min, and separated on a 75-µm fused silica emitter packed with 8 cm Magic C18 AQ 3 µm (Michrom Bioresources, Auburn, CA) at a flow rate of 0.3 µL/min. A J Biochem Molecular Toxicology

DOI 10.1002/jbt

gradient from 100 to 70% mobile phase A in 45 min followed by 70 to 55% mobile phase A in 5 min was used. Mobile phases A and B were 3% (v/v) acetonitrile and 100% acetonitrile with 0.1% (v/v) formic acid,

14

ERZINGER ET AL.

Volume 29, Number 1, 2015

TABLE 2. Comparison of SRM Data with Values from Western Blot Analysis. HT29 Cells were Treated with either 2 µM Sodium Selenite or 30 µM D3T for 48 ha

Na2 SeO3 D3T

WBb SRMc WBb SRMc

PTGR1

TrxR1

NQO1

2.0 1.0 1.1 2.1***

1.3 1.6*** 3.1*** 2.2***

1.0 1.0 2.1** 1.9***

Numbers are fold changes between control and treated samples. *p value ࣘ 0.05, **p value ࣘ 0.01; ***p value ࣘ 0.001. b Representative Western blot data that was used to calculate fold changes are shown in Figure 2. c Corresponding SRM data that was used to calculate fold changes are shown in Figure 1. Fold changes were obtained by statistical analysis of two proteotypic peptides per protein (see Materials and Methods).

a

FIGURE 2. Immunoblots of (a) PTGR1 and (b) TrxR1 and NQO1 in HT29 cells treated with 2 µM sodium selenite or 30 µM D3T for 48 h. Experiments have been performed in triplicate. For simplification, only one representative band per sample is shown. Treated samples and corresponding controls were run on the same gel. Actin served as loading control. For TrxR1, two bands were appearing only the upper band was quantified.

respectively. The targeted proteins were relatively quantified by selecting two proteotypic peptides per protein and two to four primary transitions for each proteotypic peptide. Proteotypic peptides had a unique sequence in the human proteome (according to the database from UniProtKB), had no miss-cleavage, and if possible no cysteine, methionine, and asparagine. If no proteotypic peptides without these amino acids were detected, peptides containing such amino acids were selected (i.e., LGFDVVFNYK (PTGR1), HFVGYPTNSDFELK (PTGR1), VMVLDFVTPTPLGTR (TRXR1), VVYENAYGQFIGPHR (TRXR1)). Primary transitions of the proteotypic peptides were selected based on their experimental high signal-to-noise ratio and used for a quantitative comparison. The mass spectrometer was operated in the intelligent SRM mode, triggering the acquisition of secondary transitions for identity confirmation [28]. iRT peptides were used as reference peptides for scheduling acquisition of the other peptides with on-the-fly RT recalibration [29]. The proteotypic peptides and their transitions used for SRM analysis are presented in Table S1 in the Supporting Information. The collision gas pressure was set at 1.5 mbar argon and the resolutions of the first and third quadrupoles were set at 0.7 mass unit. The scan time for each transition was set at 50 ms. Collision energy

for each peptide ion was calculated according to the default equation for the TSQ Vantage. The resulting raw data were processed with Skyline 1.4.0.4222, an open-source application for building SRM methods and analyzing the resulting SRM data [30]. Peak areas of the primary transitions of the light and heavy peptides were integrated and summed with Skyline. Statistical analysis of the generated light-toheavy peptide ratio (L/H) was performed with the R package SRMstats 1.6, a statistical framework that allows protein significance analysis [31]. The corresponding fold change in protein abundance was calculated by comparing the added peak areas of the primary transitions for the proteotypic peptides treated versus control cells of the cell sample replicate. For HT29 cells incubated for 72 h with 2 µM Na2 SeO3 , samples from each replicate were independently analyzed with the SRMstats package because the corresponding digested protein samples were spiked with different amounts of reference peptides; in this case, the corresponding fold change in protein expression from each replicate was averaged to obtain the value presented in Table 1.

Western Blot Analysis R Cell lysates were separated on NuPAGE 4–12% R Bis–Tris gel at 200 V for 60 min in a 1× NuPAGE MES SDS running buffer (Invitrogen, Carlsbad, CA) and electrophoretically transferred to an Amersham Hybond-P PVDF membrane (GE Healthcare, BuckingR hamshire, UK) at 30 V for 60 min in a 1× NuPAGE transfer buffer (Invitrogen). PVDF membranes were blocked with 5% milk–TBST at room temperature. The membranes were incubated with rabbit polyclonal antiPTGR1 antibody (1:500), anti-NQO1 antibody (1:2000), or anti-TrxR1 antibody (1:1000), respectively, followed by incubation with anti-rabbit IgG HRP antibody (1:5000; GE Healthcare). The proteins were detected usR ECL Western Blotting Substrate (Thermo ing a Pierce

J Biochem Molecular Toxicology

DOI 10.1002/jbt

Volume 29, Number 1, 2015

INDUCTION OF REDUCTASES

15

A lag time between treatment and response in enzyme expression changes was expected; therefore, analyses were performed up to 72 h after a single addition of the compounds.

Influence of Sodium Selenite on Enzyme Activity

SCHEME 1. Reduction reactions mediated by key reductase enzymes.

Scientific, Rockford, IL). Treated membranes were exposed to a X-ray film and developed using standard protocols. The protein bands were quantified using ImageJ software [32]. Actin served as loading control.

Drug Cytotoxicity Assay Hydroxymethylacylfulvene (HMAF), MMC, and cisplatin toxicity were analyzed in HT29 cells with and without Na2 SeO3 and D3T preconditioning. Cells were seeded in 96-well plates (1500 cells/well) and allowed to attach. Twenty four hours later they were exposed to 2 µM Na2 SeO3 , 30 µM D3T, or 0.1% (v/v) DMSO (vehicle control for D3T treatment). After 48 h, the medium was removed and increasing drug concentrations (HMAF: 0.01–50 µM, MMC: 0.001–15 µM, cisplatin: 0.1–1000 µM) were added. Twenty four hours after adding the drug, the medium was removed and fresh medium was added. Seventy two hours later, cell R Lumiviability was assayed using the CellTiter-Glo nescent Cell Viability Assay (Promega). Independent experiments were performed in triplicate. Statistical analysis was performed by using the unpaired two sample t-test.

RESULTS Enzyme Mediators of Reductase Activity Cellular responses were profiled on the basis of complementary chemical biotransformations of ER, DR, and QR capacities. Representative reductase enzymes anticipated to contribute to such processes and that have high relevance in the context of chemical induction, and cellular drug susceptibility was profiled. These include PTGR1 for ER, involved in the direct activation of HMAF [3, 4]; TrxR1 for DR, important for controlling the redox status of the cell [11]; and NQO1 for QR, involved in the activation of MMC [14] (Scheme 1). J Biochem Molecular Toxicology

DOI 10.1002/jbt

HT29 cells were incubated with a single dose of sodium selenite (2 µM) for 24, 48, or 72 h, and lysate activities corresponding to ER, DR, and QR capacity were profiled with trans-2-nonenal, DTNB, or 2,6-DCIP as substrates, respectively. At this concentration, there was no observable impact on cell viability on the basis of a full dose-response curve for sodium selenite in HT29 cells (data not shown). Enzyme activity data are presented in Table 1 and shown as fold changes between treated and control samples. For ER, a significant but small (1.2) increase in the activity was only detected after 24 h. On the other hand, DR activity significantly increased for all time points (24, 48, 72 h), with the highest fold change of 2.2 occurring after 72 h. Finally, QR activity increased from 24 to 48 h (1.3 and 1.6, respectively), but after 72 h, an elevation was no longer apparent. These data provided some indication of activity modulation, but lacked specificity and sensitivity, and we therefore sought a complementary strategy to determine changes in specific enzyme levels.

Influence of Sodium Selenite on Enzyme Levels Protein levels in the same cell lysates as described above were evaluated with an SRM-based mass spectrometry assay. Results are shown as fold changes in Table 1. By this approach, the enone-reducing enzyme PTGR1 was specifically measured with SRM as opposed to general ER capacity measured with the activity-based assay. Selenite preconditioning had little impact on PTGR1 levels, with a 1.2-fold induction measured after 48 h and no change at 24 or 72 h. In contrast, TrxR1 levels, determined on the basis of SRM as opposed to DR activity measured with the activity-based assay, were highly modulated, doubling on average, with a slight peak in the 48 h levels versus the shorter (24 h) or longer (72 h) time points. Finally, NQO1 levels determined with SRM increased after 24, 48, as well as 72 h, with an average fold change of 1.3. As selenite differentially impacted the three enzymes, we aimed to compare how D3T, a chemopreventive organosulfur compound, influenced cellular protein levels and activities.

16

ERZINGER ET AL.

Influence of D3T on Enzyme Activity The activity of the three reductase enzymes was analyzed after incubating HT29 cells with a single dose of D3T (30 µM) for 24, 48, or 72 h, and corresponding fold changes were calculated (Table 1). Although a complete dose-response relationship could not be established due to the low toxicity and low solubility of D3T, at 30 µM, it had no impact on HT29 cell viability for the duration of the treatment (data not shown). In general, the tested activities were significantly induced without apparent dependence on induction time period. For ER, the average fold change for all three time points was 1.6, whereas it was 1.4 for DR after 24 and 72 h (48 h change was not statistically significant). We measured the largest D3T-mediated induction in activity for QR. In this case, there was a 1.8-fold change after 24 h and 1.7-fold change after 72 h. While a higher average value (2.1) was obtained after 48 h, the large variance in replicate measurements translated into a lack of statistical significance.

Volume 29, Number 1, 2015

SRM showed a different situation: Protein levels measured after D3T treatment were higher, with a maximal fold change of 2.4 after 48 h. These data suggest that transcriptional regulation via Nrf2 as well as selenium availability control expression as well as activity of TrxR1. For NQO1 levels, a maximum fold change of 2.3 was reached after a 72 h D3T treatment, with the corresponding value for sodium selenite being only 1.3. Likewise, in the case of QR activity, higher fold changes were measured after D3T treatment versus selenite, but values were not as high as was observed for protein levels. The highest QR activity fold change measured was 1.8, occurring already after a 24 h treatment with 30 µM D3T. NQO1 is well known to be regulated by the transcription factor Nrf2; therefore, the higher induction capacity of D3T over selenium measured in this study is consistent with previous findings.

Comparison of SRM Data with Values from Western Blot Analysis Influence of D3T on Enzyme Levels In the case of protein levels, D3T preconditioning had a significant impact on PTGR1 levels, with an average 1.6-fold change after 24 and 48 h and an even higher 1.9-fold increase after 72 h. TrxR1 levels were also influenced by D3T treatment, resulting in a 1.9fold induction after 24 h and an even higher increase for longer time points (2.4-fold change in average at 48 and 72 h). For D3T as an inducing agent, the NQO1 data were remarkable in showing the only clear response of increasing protein levels as a function of time. Thus, a maximal fold change of 2.3 was observed after 72 h, following 1.4- and 1.9-fold increases after 24 and 48 h, respectively.

Relative Influence of D3T versus Selenite After 24 h incubation of HT29 cells with D3T, a significant 1.6-fold increase in ER activity was measured, whereas the analogous value for selenite was 1.2. SRM data revealed a significant increase in PTGR1 levels in D3T-treated cells while there was little influence on selenite. Selenite is not known to be involved in the transcriptional regulation of PTGR1, consistent with our data. In contrast, the larger influence of D3T is consistent with regulation of PTGR1 by the Nrf2Keap1 pathway, as well as the significant contribution of PTGR1 to ER. After 72 h of sodium selenite treatment, DR activity was increased 2.2-fold, whereas the corresponding value for D3T was only 1.3. Analysis of TrxR1 levels by

In a second set of HT29 cell samples, SRM results for cells treated with 2 µM sodium selenite or 30 µM D3T for 48 h (Figure 1) were compared with values obtained by Western blot analysis performed with antibodies for PTGR1, TrxR1, and NQO1 (Figure 2). Calculated fold changes were compared (Table 2), and were similar, but changes detected by SRM mass spectrometry had higher statistical significance. For Western blot analysis, only the two highest fold changes (NQO1 and TrxR1 after D3T treatment) could be detected in a statistically significant manner. Of note, in this limited comparative assessment of a unique set of samples, replicates were obtained from experiments in which cell samples were prepared in triplicate and from passages close to each other. When comparing with the complete set of samples characterized by the data in Table 1, including where 48 h selenite samples were derived from distant cell passages, the variation in L/H ratios of proteotypic peptide intensities was smaller. As a general comparison between the analysis methods, both assays are similar in requiring methods to detect and then quantify a protein of interest in complex samples, but differ substantially in the statistical quality of results. The SRM approach is analytically accurate and amenable to high throughput, but the moderate discrepancies between biological repetitions underscores the requirement to analyze very large numbers of biological replicates to reliably understand subtle changes in complex biological samples. As recently discussed by Aebersold and colleagues [33], Western blotting detection usually depends solely on the specificity of the antibody used, whereas in SRM, J Biochem Molecular Toxicology

DOI 10.1002/jbt

Volume 29, Number 1, 2015

INDUCTION OF REDUCTASES

detection relies on multiple parameters [33]. A benefit of the targeted mass spectrometry approach is that multiple proteotypic peptides per protein are quantified and integrated with bioinformatic and statistical tools [26, 31]. Often better than standard Western blot analysis, SRM has the ability to discern highly similar protein sequences and do so with higher throughput potential [34]. In the present study, advantages of SRM over Western blot analysis were evident. For example, in the case of PTGR1, antibodies from different commercial providers were tested and in all cases, unspecific binding was observed, leading to multiple bands, and quantification results were inconsistent. The situation for TrxR1 was better, yet as reported by the commercial provider, two bands were still present in the region of interest. The upper band, reported as the target, was quantified. For NQO1 antibody, there was no problem with unspecific binding and the quantitative data correlates well with the SRM data.

Drug Cytotoxicity To evaluate possible impacts of the induced enzyme levels on drug cytotoxicity, two bioreductive anticancer drugs (HMAF and MMC) and the direct-acting drug (cisplatin) were tested in HT29 cells preconditioned with either 2 µM sodium selenite or 30 µM D3T. Results were compared to the corresponding values in vehicle-treated control cells and are shown in Figure 3. D3T is a known Nrf2 inducer and was expected to induce reductases regulated by this transcription factor, namely all three reductases of interest in this study (PTGR1, TrxR1, NQO1), and therefore potentially influence the toxicity of HMAF and MMC, both bioreductive anticancer drugs being activated by PTGR1 or NQO1, respectively. For sodium selenite, only an effect on TrxR1 was expected, because as a selenoprotein it relies on selenium supply in the medium. Because no direct effect of TrxR1 on activation of neither HMAF nor MMC is known, the outcome of pretreatment with sodium selenite on the cytotoxicity of the two drugs was unknown. As anticipated, pretreatment with D3T or sodium selenite and corresponding changes in reductase expression did not change the cytotoxicity of cisplatin, a drug that does not require bioreductive activation. Owing to the subtle changes in enzyme expression and activity that were detected in this study, the impact on cytotoxicity of bioreductive drugs was not expected to be very large but still a question of interest, because, in contrast to previous studies, doses of sodium selenite as well as D3T that were too low to induce appreciable toxicity on their own were used for the pretreatment. Although most of the tested combinations had J Biochem Molecular Toxicology

DOI 10.1002/jbt

17

no observable effect on the cytotoxicity of HMAF or MMC, combining D3T pretreatment with HMAF did result in a small but significant increase in drug cytotoxicity. HMAF relies on bioactivation by PTGR1, an enone-reducing enzyme [3, 4]. Therefore, induced increases in PTGR1 levels and increased enone-reducing activity upon D3T treatment presented in this study are very likely to be responsible for the increased drug cytotoxicity. For sodium selenite pretreatment, no direct influence on the activation of HMAF was expected, but it was previously reported in other cell lines that [35] sodium selenite pretreatment can influence HMAF cytotoxicity, possibly on an indirect basis related with changing cellular redox status or altering the enzyme. In this study, however, although DR as well as TrxR1 levels were induced, HMAF cytotoxicity was not. Reasons might be that the induction potency of sodium selenite in the study presented here is much lower compared to the previous study, where TrxR1 levels as well as activity were induced up to fourfold [35]. No changes in cytotoxicity for MMC were detected, even though NQO1 levels, an enzyme involved in the bioactivation of MMC, were induced with both compounds. Very likely the fold changes measured in this study are not large enough to induce a change in drug cytotoxicity. Doherty and colleagues [36] showed in an earlier study that it is difficult to define a threshold for the induction of NQO1 needed to induce changes in MMC cytotoxicity and they additionally showed that any theoretical threshold might also be dependent on the cell line.

DISCUSSION PTGR1, TrxR1, and NQO1 protein levels were analyzed with respect to compound- and time-dependent induction, and corresponding reductase activities were evaluated on the basis of activity probe based assays. From a methodological perspective, measuring protein levels by SRM mass spectrometry proved to be more precise and statistically reliable than by spectrophotometric or Western blot analysis. On the other hand, because activity data reflect a toxicity-relevant phenotype, i.e., enzyme function, the biological relevance is high. In this study, we consider the importance of accuracy and specificity of the method used on one side with biological relevance on the other side. The relative influence of D3T versus selenite on PTGR1 induction detected here suggests that selenite had little influence on PTGR1 levels and ER activity, whereas D3T significantly induced both. Selenite is not known to be involved in the transcriptional regulation of PTGR1, but the larger influence of D3T

18

ERZINGER ET AL.

Volume 29, Number 1, 2015

1.4

1.4 control

1 0.8

*

0.6 0.4

*

0.2

0.8 0.6 0.4

0 0.01

0.1 0.5 1 HMAF (μM)

5

0.01

0.1 0.5 1 HMAF (μM)

5

1.4

1.4 control

1.2 1 0.8 0.6 0.4

control

1.2

30 μM D3T

rel. survival

rel. survival

2 μM Na2SeO3

1

0.2

0

2 μM Na2SeO3

1 0.8 0.6 0.4 0.2

0.2

0

0 0.01

0.1 0.5 5 MMC (μM)

0.01

15

1.4

0.1 0.5 MMC (μM)

5

1.4 control

1.2 1 0.8 0.6 0.4 0.2

control

1.2

30 μM D3T

rel. survival

rel. survival

control

1.2

30 μM D3T

rel. survival

rel. survival

1.2

2 μM Na2SeO3

1 0.8 0.6 0.4 0.2

0

0 1

5 25 50 cispla (μM)

100

1

5 25 50 cispla (μM)

100

FIGURE 3. Drug cytotoxicity for HMAF, MMC, and cisplatin in HT29 cells pretreated with D3T or Na2 SeO3 compared to the corresponding control cells. Three independent experiments were performed. *p value ࣘ 0.05.

is consistent with regulation by the Nrf2-Keap1 pathway [5, 6]. TrxR1 protein expression and activity were significantly induced by both selenite and D3T; however, a larger increase in DR activity resulted from selenite versus D3T. TrxR1, a selenoprotein, relies on selenium supply and enriching media with selenium increased TrxR1 in different cell lines [11, 35]. Additionally, it was shown to be regulated via the Nrf2-Keap1 pathway [11, 37]. Finally, protein levels of NQO1 were significantly increased by both compounds, but with higher fold changes for D3T, consistent with the fact that NQO1 is regulated through the transcription factor Nrf2 [12].

By combining D3T pretreatment with HMAF, a significant increase in cytotoxicity of the drug was detected. In another study, pretreating HepG2 or SW620 cells with curcumin, resveratrol, or D3T reduced IC50 values for HMAF to a larger extent [38]. In the earlier case, however, PTGR1 was induced up to fourfold [38]. Furthermore, the sensitivity of HeLa cells toward acylfulvenes was increased about 50% upon selenite preconditioning, which was associated with a fourfold induction of TrxR1 [35]. The small cytotoxicity differences observed in the system described here may be related to the low induction capacities of D3T or selenite in HT29 cells. HepG2 cells are liver cells known to be highly J Biochem Molecular Toxicology

DOI 10.1002/jbt

Volume 29, Number 1, 2015

INDUCTION OF REDUCTASES

responsive to exogenous inducers. Although HT29 cells were selected as one of the most common cancer cell line models, harboring typical APC and p53 mutations and lacking mismatch DNA repair deficiencies, basal levels of PTGR1 appear to be higher than in other colon cancer cells such as SW620. Thus, the enzyme response triggered by D3T may be attenuated when the levels are high. Nonetheless, future research employing the methodology established here but involving a large number of cell lines could clarify how protein levels, enzyme activities, and drug cytotoxicity vary on the basis of inherent cell characteristics. We showed that for each of the three investigated enzymes, time period of treatment as well as chemical properties of the bioactive compound used lead to different specific induction profiles, albeit with a small but measurable degree of change. When our new reductase-targeting SRM assay was applied, specificity was high, whereas for the activity data there is potentially more biological relevance. In this study, we established a way to balance this information by combining analytical tools for probing reductive drugmetabolizing enzymes and defined how selenite and D3T induce the activity, as well as protein levels, of PTGR1, TrxR1, and NQO1 in HT29 cells.

4.

5. 6.

7. 8. 9. 10.

11.

12.

SUPPORTING INFORMATION 13.

Table S1. includes the proteotypic peptides and their mass transitions measured by SRM analysis. This material is available free of charge via the Internet.

14. 15.

ACKNOWLEDGMENTS We thank Nathalie Ziegler (ETH Zurich) for technical assistance and support and Dr. Giancarlo Marra (University of Zurich) for helpful discussions and technical support. We acknowledge Professors Christophe Lacroix and Leo Meile (ETH Zurich) for providing HT29 cells.

REFERENCES

17. 18.

19.

1. Tai H-H, Ensor CM, Tong M, Zhou H, Yan F. Prostaglandin catabolizing enzymes. Prostaglandins Other Lipid Mediat 2002;68–69:483–493. 2. Yu X, Egner PA, Wakabayashi N, Yamamoto M, Kensler TW. Nrf2-mediated induction of cytoprotective enzymes by 15-deoxy-delta12,14-prostaglandin J2 is attenuated by alkenal/one oxidoreductase. J Biol Chem 2006;281:26245–26252. 3. Dick RA, Yu X, Kensler TW. NADPH alkenal/one oxidoreductase activity determines sensitivity of cancer cells J Biochem Molecular Toxicology

16.

DOI 10.1002/jbt

20. 21. 22.

19

to the chemotherapeutic alkylating agent irofulven. Clin Cancer Res 2004;10(4):1492–1499. Gong JC, Neels JF, Yu X, Kensler TW, Peterson LA, Sturla SJ. Investigating the role of stereochemistry in the activity of anticancer acylfulvenes: synthesis, reductasemediated bioactivation, and cellular toxicity. J Med Chem 2006;49(8):2593–2599. Primiano T, Gastel JA, Kensler TW, Sutter TR. Isolation of cDNAs representing dithiolethione-responsive genes. Carcinogenesis 1996;17(11):2297–2303. Dick RA, Kwak M-K, Sutter TR, Kensler TW. Antioxidative function and substrate specificity of NAD(P)Hdependent alkenal/one oxidoreductase. J Biol Chem 2001;276(44):40803–40810. ¨ Holmgren A, Bjornstedt M. Thioredoxin and thioredoxin reductase. Methods Enzymol 1995;252:199–208. Mustacich D, Powis G. Thioredoxin reductase. Biochem J 2000;346:1–8. Holmgren A. Reduction of disulfides by thioredoxin. J Biol Chem 1979;254:9113–9119. Arn´er ESJ, Nordberg J, Holmgren A. Efficient reduction of lipoamide and lipoic acid by mammalian thioredoxin reductase. Biochem Biophys Res Commun 1996;225:268– 274. ¨ A-K, Arn´er ESJ. Regulation of the mammalian Rundlof selenoprotein thioredoxin reductase 1 in relation to cellular phenotype, growth, and signaling events. Antioxid Redox Signaling 2004;6:41–52. Nioi P, Hayes JD. Contribution of NAD(P)H:quinone oxidoreductase 1 to protection against carcinogenesis, and regulation of its gene by the Nrf2 basic-region leucine zipper and the arylhydrocarbon receptor basic helixloop-helix transcription factors. Mutat Res 2004;555:149– 171. Verweij J, Pinedo HM. Mitomycin C: mechanism of action, usefulness and limitations. Anti-Cancer Drugs 1990;1:5–13. Cummings J, Spanswick VJ, Tomasz M, Smyth JF. Enzymology of mitomycin C metabolic activation in tumour tissue. Biochem Pharmacol 1998;56:405–414. Rayman MP. Selenium in cancer prevention: a review of the evidence and mechanism of action. Proc Nutr Soc 2005;64:527–542. Zeng H, Combs GF Jr. Selenium as an anticancer nutrient: roles in cell proliferation and tumor cell invasion. J Nutr Biochem 2008;19:1–7. Allan CB, Lacourciere GM, Stadtman TC. Responsiveness of selenoproteins to dietary selenium. Annu Rev Nutr 1999;19:1–16. Kwak M-K, Itoh K, Yamamoto M, Sutter TR, Kensler TW. Role of transcription factor Nrf2 in the induction of hepatic phase 2 and antioxidative enzymes in vivo by the cancer chemoprotective agent, 3H-1, 2-dithiole-3-thione. Mol Med 2001;7:135–145. Zhang Y, Munday R. Dithiolethiones for cancer chemoprevention: where do we stand? Mol Cancer Ther 2008;7:3470–3479. Erzinger MM, Sturla SJ. Bioreduction-mediated fooddrug interactions: opportunities for oncology nutrition. Chimia 2011;65:411–415. Lange V, Picotti P, Domon B, Aebersold R. Selected reaction monitoring for quantitative proteomics: a tutorial. Mol Syst Biol 2008;4(222):1–14. Ross PL, Huang YN, Marchese JN, Williamson B, Parker K, Hattan S, Khainovski N, Pillai S, Dey S, Daniels S. Multiplexed protein quantitation in Saccharomyces cerevisiae

20

23.

24. 25.

26.

27. 28.

29.

30.

ERZINGER ET AL.

using amine-reactive isobaric tagging reagents. Mol Cell Proteomics 2004;3(12):1154–1169. Addona TA, Abbatiello SE, Schilling B, Skates SJ, Mani DR, Bunk DM, Spiegelman CH, Zimmermann LJ, Ham A-JL, Keshishian H. Multi-site assessment of the precision and reproducibility of multiple reaction monitoringbased measurements of proteins in plasma. Nat Biotechnol 2009;27(7):633–641. Picotti P, Bodenmiller B, Mueller LN, Domon B, Aebersold R. Full dynamic range proteome analysis of S. cerevisiae by targeted proteomics. Cell 2009;138(4):795–806. Picotti P, Rinner O, Stallmach R, Dautel F, Farrah T, Domon B, Wenschuh H, Aebersold R. High-throughput generation of selected reaction-monitoring assays for proteins and proteomes. Nat Methods 2010;7(1): 43–48. ¨ Reiter L, Rinner O, Picotti P, Huttenhain R, Beck M, Brusniak M-Y, Hengartner MO, Aebersold R. mProphet: automated data processing and statistical validation for largescale SRM experiments. Nat Methods 2011;8:430–435. Ernster L. DT diaphorase. Methods Enzymol 1967;10:309–317. Kiyonami R, Schoen A, Prakash A, Peterman S, Zabrouskov V, Picotti P, Aebersold R, Huhmer A, Domon B. Increased selectivity, analytical precision, and throughput in targeted proteomics. Mol Cell Proteomics 2011;10(2):M110. 002931. Escher C, Reiter L, MacLean B, Ossola R, Herzog F, Chilton J, MacCoss MJ, Rinner O. Using iRT, a normalized retention time for more targeted measurement of peptides. Proteomics 2012;12:1111–1121. MacLean B, Tomazela DM, Shulman N, Chambers M, Finney GL, Frewen B, Kern R, Tabb DL, Liebler DC, MacCoss MJ. Skyline: an open source document editor for

Volume 29, Number 1, 2015

31.

32. 33.

34. 35.

36.

37. 38.

creating and analyzing targeted proteomics experiments. Bioinformatics 2010;26:966–968. Chang CY, Picotti P, Huettenhain R, HeinzelmannSchwarz V, Jovanovic M, Aebersold R, Vitek O. Protein significance analysis in selected reaction monitoring (SRM) measurements. Mol Cell Proteomics 2012;11:M111.014662. Abramoff MD, Magalhaes PJ, Ram SJ. Image processing with ImageJ. Biophotonics Int 2004;11(7):36–42. Aebersold R, Burlingame AL, Bradshaw RA. Western blots versus selected reaction monitoring assays: time to turn the tables? Mol Cell Proteomics 2013;12(9):2381– 2382. Picotti P, Bodenmiller B, Aebersold R. Proteomics meets the scientific method. Nat Methods 2013;10:24–27. Liu X, Pietsch KE, Sturla SJ. Susceptibility of the antioxidant selenoenzymes thioredoxin reductase and glutathione peroxidase to alkylation-mediated inhibition by anticancer acylfulvenes. Chem Res Toxicol 2011;24:726– 736. Doherty GP, Leith MK, Wang X, Curphey TJ, Begleiter A. Induction of DT-diaphorase by 1,2-dithiole-3-thiones in human tumour and normal cells and effect on antitumour activity of bioreductive agents. Br J Cancer 1998;77:1241–1252. Kim Y-C, Masutani H, Yamaguchi Y, Itoh K, Yamamoto M, Yodoi J. Hemin-induced activation of the thioredoxin gene by Nrf2. J Biol Chem 2001;276(21):18399–18406. Yu X, Erzinger MM, Pietsch KE, Cervoni-Curet FN, Whang J, Niederhuber J, Sturla SJ. Upregulation of human prostaglandin reductase 1 (PTGR1) improves efficacy of hydroxymethylacylfulvene, an antitumor chemotherapeutic agent. J Pharmacol Exp Ther 2012;343:426–433.

J Biochem Molecular Toxicology

DOI 10.1002/jbt