Glutathiolation Regulates Tumor Necrosis Factor-α–Induced Caspase-3 Cleavage and Apoptosis
Key Role for Glutaredoxin in the Death Pathway
Caspase-3 cleavage and activation are known to play central roles in apoptosis. However, the mechanisms that regulate caspase-3 cleavage remain elusive. Glutaredoxin (Grx) is a ubiquitous redox molecule that is unique in its ability to regulate S-glutathiolation (glutathiolation) of proteins. Here we show the essential role of Grx in caspase-3 cleavage via regulation of caspase-3 glutathiolation. Grx activity was significantly upregulated by tumor necrosis factor-α in endothelial cells. Small interference RNA knock down of Grx significantly inhibited tumor necrosis factor-α–induced endothelial cell death because of attenuated caspase-3 cleavage concomitant with increased caspase-3 glutathiolation. Enhanced caspase-3 cleavage by wild-type Grx overexpression was reversed by catalytically inactive Grx (C22S), demonstrating a requirement for thioltransferase activity. Cysteine-to-serine mutations (C163S, C184S, and C220S) of caspase-3 that were predicted to prevent glutathiolation showed increased cleavage compared with wild-type caspase-3. This inverse correlation between caspase-3 glutathiolation and cleavage was further confirmed by the observation that in vitro glutathiolation of caspase-3 inhibited its cleavage with recombinant caspase-8. Furthermore, Grx association with caspase-3 was decreased by tumor necrosis factor-α. These findings demonstrate a novel mechanism of caspase-3 regulation by Grx in tumor necrosis factor-α–induced apoptosis.
Glutaredoxin (Grx) has been identified as a ubiquitous multifunctional thioltransferase. Among redox proteins that contain a dithiol in their conserved catalytic site sequence, Grx uniquely reduces glutathiolated proteins (protein–S-glutathione, or protein-SSG) because of its low pKa cysteine C22 and its ability to bind glutathione.1 Grx also catalyzes reversible oxidoreduction involving dithiol reactions, as in the reduction of ribonucleotide reductase.2,3 It is well established that, as an antioxidant enzyme, Grx protects cells from oxidative stress, whereas S-glutathiolation (glutathiolation), a posttranslational modification, protects proteins from irreversible oxidation. Recent studies have demonstrated important functional changes in glutathiolated proteins, including inactivation of kinases and phosphatases, changes in actin polymerization, and protein binding to DNA.4–9 However, the functional consequences of Grx-mediated protein deglutathiolation in specific signaling events remain unknown.
Because tumor necrosis factor-α (TNF-α)–induced inflammation and apoptosis contribute to vascular diseases including atherosclerosis and hypertension,10–13 we are interested in the role of glutathiolation in TNF-α–induced death signaling. TNF induces apoptosis through the recruitment of the death domain–containing adaptor proteins FADD and TRADD to its receptor, which in turn leads to the activation of the initiator caspase, caspase-8.14 The activated caspase-8 cleaves and activates caspase-3. Once caspase-3 is activated, it proteolyses important structural and regulatory components of the cell, leading to apoptosis.
In the present study, we identify a novel regulatory mechanism for caspase-3 activation. Specifically, we show that Grx plays a key role in caspase-3 cleavage by regulating caspase-3 glutathiolation in response to TNF-α, thereby modulating TNF-α–induced death signaling.
Materials and Methods
Antibodies and Reagents
Actin was from Santa Cruz Biotechnology; caspase-3 was from Cell Signaling; human Grx was from American Diagnostica Inc; and anti-SSG was from Virogen.
TNF-α was from Roche; pcDNA3-Grx was a generous gift from Dr Y. J. Lee (University of Pittsburgh, Pa); Flag-tagged caspase-3 was a generous gift from Dr Masayuki Miura (University of Tokyo, Japan); and recombinant caspase-8 was from Biomol.
Cell Culture, Small Interfering RNA Oligonucleotide Treatment, and Plasmid Transfection
Human umbilical vein endothelial cells (HUVECs) were purchased from Cascade Biologics and cultured in medium 200 supplemented with low-serum growth supplement (Cascade Biologics). Cells at passages 3 to 5 were used for experiments. To knock down Grx, we treated HUVECs with small interference RNA (siRNA) against human Grx. HUVECs at >90% confluence were used for transfection. Lipofectamine 2000 and Oligofectamine (Invitrogen) were used according to the instructions of the manufacturer.
Bovine aortic endothelial cells (BAECs) were purchased from Clonetics and cultured in medium 199 supplemented with 10% FETALCLONE III (Hyclone), basal modified Eagle’s medium vitamins, and amino acids (Invitrogen). Cells at passages 6 to 10 were used for experiments. BAECs at >80% confluence were used for transfection. Lipofectamine 2000 (Invitrogen) was used according to the instructions of the manufacturer.
The QuikChange site-directed mutagenesis kit (Stratagene, La Jolla, Calif) was used to generate the Grx-C22S mutation as well as caspase-3-C163S, caspase-3-C184S, and caspase-3-C220S mutations. PCRs and transformation were performed according to the instructions of the manufacturer. The mutations were confirmed by sequencing.
Western Blot, Immunoprecipitation, and BioGEE
Western blots (immunoblots) and immunoprecipitations were performed as described previously.10,11 The resulting autoradiograms were analyzed with NIH Image 1.60. Equal loading of protein was ensured by measuring actin expression. Glutathiolated protein labeling was performed as described.15 Briefly, endothelial cells (ECs) were preincubated with biotin-labeled, glutathione-reduced ethyl ester (BioGEE) in the presence of TNF-α (10 ng/mL) plus cycloheximide (CHX) (10 μg/mL). ECs were harvested, and glutathiolated proteins were collected by streptavidin-conjugated agarose beads. Proteins were released by incubation with 10 mmol/L dithiothreitol and separated by SDS-PAGE. Caspase-3 was then detected by Western blot.
In Vitro Glutathiolation of Caspase-3
HEK 293 cells were transfected with Flag-tagged caspase-3. After immunoprecipitation with Flag antibody, the agarose beads were washed with glutathiolation buffer containing 40 mmol/L KCl and 10 mmol/L KPO4. In vitro glutathiolation was performed using 500 μmol/L diamide as the oxidant in the presence of 10 mmol/L glutathione as described.4,16 Protein aliquots were subjected to SDS-PAGE. Caspase-3 glutathiolation was determined using an anti-SSG antibody. The remaining protein–agarose beads were subjected to extensive washing to remove the oxidant with cleavage buffer containing 50 mmol/L HEPES (pH 7.4), 100 mmol/L NaCl, 0.5% CHAPS (3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate), 1 mmol/L EDTA. In vitro cleavage of caspase-3 by caspase-8 was performed according to the instructions of the manufacturer. Cleaved caspase-3 was detected by Western blot.
Grx Activity Assay
BAECs were treated with TNF-α plus CHX. Grx activity was measured as described.17,18
Results represent 3 independent experiments if not mentioned specifically, and data are expressed as mean±SE. Where indicated, analysis of variance was performed, and P<0.05 was considered significant in all experiments.
Grx Is Activated by TNF-α, and Grx Knock-Down ECs Are Resistant to TNF-α–Induced Cell Death
We first assayed to determine whether TNF-α activates Grx. The results showed that TNF-α increased Grx activity by 70% at 6 hours (Figure 1A). These data suggest that TNF-α activates Grx, which can then deglutathiolate proteins during apoptosis.
To study the physiologic role of Grx in TNF-α–induced EC death, we used siRNA to specifically knock down Grx. Successful transfection was confirmed by fluorescence microscopy after transfection with fluorescein-conjugated control siRNA, which has no homology with any human gene. Two different Grx siRNA decreased EC Grx expression in a siRNA concentration–dependent manner, without affecting actin expression (Figure 1B). These results show that these siRNA were highly specific for Grx.
We next examined TNF-α–induced cell death in Grx knock-down ECs. Treatment with TNF-α increased cell death from 8±2% to 61±5% in cells transfected with control siRNA (Figure 1C). In contrast, TNF-α–induced cell death was significantly decreased to 18±6% in Grx knock-down ECs (Figure 1C). These data suggest that Grx plays an important role in TNF-α–induced EC death.
Caspase-3 Glutathiolation Inversely Correlates With TNF-α–Induced Cleavage
A critical step in TNF-α–induced cell death is caspase-3 cleavage, which transforms caspase-3 to an active cysteine protease.19,20 Caspase-3 normally exists as an inactive 32-kDa proenzyme consisting of an amino-terminal prodomain, a large subunit and a small subunit. Once caspase-8 is activated, caspase-3 is cleaved by proteolysis between these domains to become activated. The first cleavage is between the large subunit and the small subunit to generate 19 kDa and 12 kDa (termed p19 and p12) fragments. The second cleavage within the 19-kDa fragment generates a p17 kDa subunit (termed p17). Finally, the p12 and p17 subunits bind together to generate active caspase-3.
During TNF-α–induced apoptosis caspase-3 is cleaved in a time-dependent manner. Because Grx is a thioltransferase that specifically catalyzes protein deglutathiolation, we hypothesized that caspase-3 cleavage was regulated by glutathiolation. To assay caspase-3 glutathiolation, we treated cells with BioGEE (biotin-labeled glutathione-reduced ethyl ester) to label glutathiolated proteins.15 Glutathiolated proteins were precipitated by streptavidin beads, and procaspase-3 was detected by Western blot (Figure 2A). Caspase-3 cleavage was assayed by Western blot using an antibody that recognizes procaspase-3 and cleaved caspase-3 (p17) specifically. TNF-α treatment decreased caspase-3 glutathiolation (Figure 2A and 2C) concomitant with increased caspase-3 cleavage (Figure 2B and 2C), suggesting that deglutathiolation of caspase-3 enhances its cleavage.
Thioltransferase Activity of Grx Is Required to Regulate Caspase-3 Cleavage
Because Grx catalyzes protein deglutathiolation, we hypothesized that Grx regulation of caspase-3 cleavage requires thioltransferase activity. To test our hypothesis, we made a thioltransferase deficient (C22S) Grx.21 Vector control, Grx-wild-type (WT), and Grx-C22S were transfected into ECs, and TNF-α–induced caspase-3 cleavage was determined using an antibody recognizing cleaved caspase-3 specifically (Figure 3). In ECs transfected with vector, TNF-α induced caspase-3 cleavage. In ECs transfected with Grx-WT, caspase-3 cleavage was increased (≈3-fold relative to vector). In contrast, cells transfected with Grx-C22S showed significantly decreased caspase-3 cleavage relative to Grx-WT (≈0.2-fold relative to vector) (Figure 3). These results support the idea that Grx regulates TNF-α–induced caspase-3 cleavage in a thioltransferase-dependent manner.
Grx Affects Caspase-3 Cleavage by Regulating Caspase-3 Glutathiolation
If Grx-mediated deglutathiolation increases caspase-3 cleavage, we predicted that decreasing Grx expression would increase caspase-3 glutathiolation and reduce caspase-3 cleavage. In Grx knocked down ECs, caspase-3 glutathiolation was increased both basally and after TNF-α treatment (Figure 4A, compare control with Grx siRNA). Most dramatic was the decrease in caspase-3 cleavage in Grx knock-down cells, as detected by Western blot using an antibody recognizing cleaved caspase-3 (Figure 4B).
To further confirm our observations, we exposed Grx knock-down ECs to different TNF-α concentrations. TNF-α decreased procaspase-3 in a concentration-dependent manner. However, more procaspase-3 remained in samples treated with Grx siRNA compared with control siRNA, indicating that Grx siRNA inhibited caspase-3 cleavage (Figure 4C and 4D).
Similar to TNF-α, H2O2 also induced caspase-3 cleavage (Figure IA in the online data supplement, available at http://circres.ahajournals.org). There was decreased caspse-3 glutathiolation in ECs treated with H2O2 (300 μmol/L) for 3 hours and 6 hours (supplemental Figure IB). Furthermore, in ECs treated with Grx siRNA, H2O2-mediated caspase-3 cleavage was inhibited concomitant with increased caspase-3 glutathiolation (supplemental Figure IC, lane 4*). These results support the hypothesis that Grx regulates caspase-3 cleavage through glutathiolation.
In Vitro Glutathiolation of Caspase-3 Inhibits Its Cleavage by Recombinant Caspase-8
To demonstrate that caspase-3 glutathiolation directly inhibits its cleavage by caspase-8, we performed in vitro glutathiolation of caspase-3 and assayed caspase-3 cleavage by recombinant caspase-8. Overexpressed caspase-3 was immunoprecipitated from HEK 293 cells, and equal amounts of caspase-3 were pulled down (Figure 5A). In vitro glutathiolation was performed in the presence of diamide and glutathione (Figure 5B, top). Caspase-3 cleavage was assayed with the addition of recombinant caspase-8 after glutathiolation (Figure 5B, bottom). Caspase-3 glutathiolation decreased with an increased time of exposure to diamide because of further oxidation (Figure 5B, top), whereas caspase-3 cleavage increased (Figure 5B, bottom). Correlation analysis of the results showed a clear negative relationship between caspase-3 cleavage and glutathiolation (Figure 5C). These results strongly indicate that caspase-3 cleavage is inhibited by caspase-3 glutathiolation.
Cysteine-to-Serine Mutations of Caspase-3 Are More Sensitive to TNF-α–Induced Cleavage
To further study the regulation of caspase-3 cleavage by Grx, we made cysteine-to-serine mutations of 3 cysteines in caspase-3 (C163, C184, C220) that are exposed at the surface of the molecule and are likely to be glutathiolated. These three cysteines are highly conserved across species and within the caspase family. We predicted that the cysteine-to-serine mutations would increase TNF-α–induced cleavage, as the mutated caspase-3 molecules could not be protected by glutathiolation. Caspase-3 mutants were transfected into ECs (Figure 6A), and caspase-3 cleavage was assayed after TNF-α stimulation for 4 hours. Cleavage of C184S, and C220S proteins in cells exposed to TNF-α was increased compared with vector control and WT caspase-3 (Figure 6B). Caspase-3 mutant C163S, which is a mutation of the catalytic site cysteine, showed cleavage to a similar extent as WT caspase-3, which could be explained by a decrease in its auto cleavage by the mutation of catalytic cysteine. Cleavage of C220S occurred even without TNF-α treatment, suggesting a particularly key role for this cysteine in caspase-3 cleavage. These data indicate that C220 and C184, which are located in caspase-3 p12 subunit after cleavage, might be essential for glutathiolation-mediated inhibition of caspase-3 cleavage.
TNF-α Regulates Caspase-3 Interaction With Grx
To further study the molecular mechanism of Grx regulation of caspase-3 cleavage, we performed immunoprecipitation to assay the interaction between Grx and caspase-3. Our results showed that Grx interacted with caspase-3 as demonstrated by both coprecipitation of endogenous proteins and transfected Grx and Flag-tagged caspase-3 (Figure 7A and 7B). Because caspase-3 translocates to the nucleus during apoptosis,22,23 we speculated that TNF-α dissociates Grx from caspase-3. Our results showed that TNF-α treatment decreased the amount of Grx bound to caspase-3, suggesting that Grx deglutathiolated caspase-3 and then was dissociated from caspase-3 under TNF-α stimulation (Figure 7).
In this study, we define a critical role for Grx in TNF-α–mediated caspase-3 activation through regulation of caspase-3 cleavage and glutathiolation. Data to support this notion include (1) caspase-3 cleavage by TNF-α inversely correlated with glutathiolation; (2) TNF-α activated Grx. In Grx knock-down cells, TNF-α–induced apoptosis was attenuated and caspase-3 cleavage was significantly reduced, concomitant with increased caspase-3 glutathiolation; (3) enhanced caspase-3 cleavage by Grx overexpression was reversed by a thioltransferase inactive Grx (C22S); (4) in vitro glutathiolation experiments showed that caspase-3 cleavage by caspase-8 was inversely correlated with caspase-3 glutathiolation; (5) mutations of cysteines (C184 and C220) in caspase-3 show significant increase in cleavage; and (6) Grx binds caspase-3. Our data suggest a model in which TNF-α activates Grx, Grx deglutathiolates caspase-3, and then Grx dissociates from caspase-3, leading to caspase-3 activation. The present study therefore identifies a novel regulatory mechanism in the death pathway mediated by Grx-dependent caspase-3 binding and deglutathiolation. Our observations are consistent with a study showing protein glutathiolation is decreased in apoptotic cells,24 suggesting a general role for Grx in regulating protein glutathiolation during apoptosis.
Caspase-3 is modified by glutathiolation, and Grx regulates TNF-α–induced apoptosis through caspase-3 glutathiolation (Figures 3 through 7⇑⇑⇑⇑). These data coincide with previous observations that caspase-3 undergoes cysteine modification.25–27 It is also known that recombinant caspase-3 is inactivated by low doses of hydrogen peroxide and reactivated by dithiothreitol treatment. These observations, combined with our data, further strengthen the idea that caspase-3 is modified by glutathiolation.25
Glutathiolation is a reversible posttranslational modification that protects proteins from irreversible oxidation and loss of function. Proteins that are known to be transiently inactivated by glutathiolation include phosphatases, kinases, and transcription factors.4–9 Here, we show that when caspase-3 is glutathiolated, it is inactivated and resistant to TNF-α–induced cleavage. Because caspase-3 activation is required for TNF-α–induced apoptosis, it is predictable that caspase-3 is deglutathiolated by TNF-α.
We observed that Grx also regulated H2O2-mediated caspase-3 cleavage (supplemental Figure I). Other groups found that Grx inhibited oxidative stress–induced cell death.28,29 The discrepancy between their data and ours might be attributable to the differences in cell types, extent and duration of oxidative stress, and experimental systems. In particular, they overexpressed Grx, whereas most of our data were generated by transient transfection of Grx siRNA.
As a cysteinease with low pKa, caspase-3 is a likely target for glutathiolation, although the precise mechanisms by which glutathiolation inhibits caspase-3 cleavage are unclear. As cysteine residues play a fundamental role in protein structure and function, it is tempting to speculate that caspase-3 glutathiolation could decrease its accessibility to caspase-8 and block the cleavage site through a structural change caused by this modification. Our mutation experiments suggest that C220 and C184 (located in the p12 subunit after cleavage) might be important sites for glutathiolation. Interestingly, it appears that glutathiolation increases caspase-3 stability, as the cysteine mutant caspase-3 molecules (C184S and C220S) are unstable with increased cleavage even without TNF-α (Figure 6).
Our study reveals a novel strategy for the inhibition of caspase-3 activation by showing that the formation of a mixed disulfide between glutathione and caspase-3 (glutathiolation) inhibits caspase-3 activation. Our model is supported by a recent publication demonstrating that a small thiol-containing molecule forms a disulfide with caspase-1 and inhibits its activation.30 Further studies on the structure of glutathiolated caspase-3 are required to shed light on the molecular mechanism for the inhibition.
In recent years, glutathiolation has emerged as a redox-sensitive posttranslational modification that plays a regulatory role in both signal transduction and protein function.5,6,31–33 Recently, Adachi and colleagues showed that SERCA is regulated by glutathiolation during arterial relaxation by nitric oxide.33 Inactivation of SERCA in atherosclerosis reduced SERCA-dependent relaxation. These findings indicate that glutathiolation of key proteins represents not only a marker of reactive oxygen species/reactive nitrogen species, but also related directly to cardiovascular function and diseases.
Oxidative stress and inflammation have become prominent mechanisms for EC dysfunction and progression of vascular diseases such as hypertension and atherosclerosis.34 Recent data show that EC apoptosis is associated with increased atherosclerosis.13 Our finding that Grx regulates caspase-3 cleavage represents a novel mechanism for redox control of vascular function, suggesting that Grx may play a critical role in the progression of vascular inflammatory diseases.
We thank Drs Mark Taubman and Junichi Abe for reading the manuscript and helpful discussion.
Sources of Funding
This work was supported by American Heart Association Scientist Development Grant 0530195N (to S.P.) and NIH grant HL077789 (to B.C.B.).
Original received July 17, 2006; revision received November 16, 2006; accepted December 8, 2006.
Casagrande S, Bonetto V, Fratelli M, Gianazza E, Eberini I, Massignan T, Salmona M, Chang G, Holmgren A, Ghezzi P. Glutathionylation of human thioredoxin: a possible crosstalk between the glutathione and thioredoxin systems. Proc Natl Acad Sci U S A. 2002; 99: 9745–9749.
Caplan JF, Filipenko NR, Fitzpatrick SL, Waisman DM. Regulation of annexin A2 by reversible glutathionylation. J Biol Chem. 2004; 279: 7740–7750.
Reddy S, Jones AD, Cross CE, Wong PS, Van Der Vliet A. Inactivation of creatine kinase by S-glutathionylation of the active-site cysteine residue. Biochem J. 2000; 347 (pt 3): 821–827.
Surapisitchat J, Hoefen RJ, Pi X, Yoshizumi M, Yan C, Berk BC. Fluid shear stress inhibits TNF-α–alpha activation of JNK but not ERK1/2 or p38 in human umbilical vein endothelial cells: inhibitory crosstalk among MAPK family members. Proc Natl Acad Sci U S A. 2001; 98: 6476–6481.
Yamawaki H, Lehoux S, Berk BC. Chronic physiological shear stress inhibits tumor necrosis factor-induced proinflammatory responses in rabbit aorta perfused ex vivo. Circulation. 2003; 108: 1619–1625.
Durand E, Scoazec A, Lafont A, Boddaert J, Al Hajzen A, Addad F, Mirshahi M, Desnos M, Tedgui A, Mallat Z. In vivo induction of endothelial apoptosis leads to vessel thrombosis and endothelial denudation: a clue to the understanding of the mechanisms of thrombotic plaque erosion. Circulation. 2004; 109: 2503–2506.
Humphries KM, Juliano C, Taylor SS. Regulation of cAMP-dependent protein kinase activity by glutathionylation. J Biol Chem. 2002; 277: 43505–43511.
Yang Y, Jao S, Nanduri S, Starke DW, Mieyal JJ, Qin J. Reactivity of the human thioltransferase (glutaredoxin) C7S, C25S, C78S, C82S mutant and NMR solution structure of its glutathionyl mixed disulfide intermediate reflect catalytic specificity. Biochemistry. 1998; 37: 17145–17156.
Kamada S, Kikkawa U, Tsujimoto Y, Hunter T. Nuclear translocation of caspase-3 is dependent on its proteolytic activation and recognition of a substrate-like protein(s). J Biol Chem. 2005; 280: 857–860.
Feng Y, Hu J, Xie D, Qin J, Zhong Y, Li X, Xiao W, Wu J, Tao D, Zhang M, Zhu Y, Song Y, Reed E, Li QQ, Gong J. Subcellular localization of caspase-3 activation correlates with changes in apoptotic morphology in MOLT-4 leukemia cells exposed to X-ray irradiation. Int J Oncol. 2005; 27: 699–704.
Di Stefano A, Frosali S, Leonini A, Ettorre A, Priora R, Di Simplicio FC, Di Simplicio P. GSH depletion, protein S-glutathionylation and mitochondrial transmembrane potential hyperpolarization are early events in initiation of cell death induced by a mixture of isothiazolinones in HL60 cells. Biochim Biophys Acta. 2006; 1763: 214–225.
Mannick JB, Hausladen A, Liu L, Hess DT, Zeng M, Miao QX, Kane LS, Gow AJ, Stamler JS. Fas-induced caspase denitrosylation. Science. 1999; 284: 651–654.
Song JJ, Rhee JG, Suntharalingam M, Walsh SA, Spitz DR, Lee YJ. Role of glutaredoxin in metabolic oxidative stress. Glutaredoxin as a sensor of oxidative stress mediated by H2O2. J Biol Chem. 2002; 277: 46566–46575.
Murata H, Ihara Y, Nakamura H, Yodoi J, Sumikawa K, Kondo T. Glutaredoxin exerts an antiapoptotic effect by regulating the redox state of Akt. J Biol Chem. 2003; 278: 50226–50233.
Scheer JM, Romanowski MJ, Wells JA. A common allosteric site and mechanism in caspases. Proc Natl Acad Sci U S A. 2006; 103: 7595–7600.
Clavreul N, Adachi T, Pimental DR, Ido Y, Schoneich C, Cohen RA. S-Glutathiolation by peroxynitrite of p21ras at cysteine-118 mediates its direct activation and downstream signaling in endothelial cells. FASEB J. 2006; 20: 518–520.
Adachi T, Pimentel DR, Heibeck T, Hou X, Lee YJ, Jiang B, Ido Y, Cohen RA. S-Glutathiolation of Ras mediates redox-sensitive signaling by angiotensin II in vascular smooth muscle cells. J Biol Chem. 2004; 279: 29857–29862.