Original Contribution

Critical cysteines in Akt1 regulate its activity and proteasomaldegradation: implications for neurodegenerative diseases

Faraz Ahmad 1, Prakash Nidadavolu 1, Lalitha Durgadoss, Vijayalakshmi Ravindranath n

Centre for Neuroscience, Indian Institute of Science, Bangalore 560012, India

a r t i c l e i n f o

Article history:Received 13 December 2013Received in revised form5 June 2014Accepted 6 June 2014Available online 14 June 2014

Keywords:GlutaredoxinOxidative stressParkinson diseaseProtein thiolsPP2AFree radicals

a b s t r a c t

Impaired Akt1 signaling is observed in neurodegenerative diseases, including Parkinson's disease (PD).In PD models oxidative modification of Akt1 leads to its dephosphorylation and consequent loss of itskinase activity. To explore the underlying mechanism we exposed Neuro2A cells to cadmium, a paninhibitor of protein thiol disulfide oxidoreductases, including glutaredoxin 1 (Grx1), or downregulatedGrx1, which led to dephosphorylation of Akt1, loss of its kinase activity, and also decreased Akt1 proteinlevels. Mutation of cysteines to serines at 296 and 310 in Akt1 did not affect its basal kinase activity butabolished cadmium- and Grx1 downregulation-induced reduction in Akt1 kinase activity, indicatingtheir critical role in redox modulation of Akt1 function and turnover. Cadmium-induced decrease inphosphorylated Akt1 correlated with increased association of wild-type (WT) Akt1 with PP2A, whichwas absent in the C296–310S Akt1 mutant and was also abolished by N-acetylcysteine treatment.Further, increased proteasomal degradation of Akt1 by cadmium was not seen in the C296–310S Akt1mutant, indicating that oxidation of cysteine residues facilitates degradation of WT Akt1. Moreover,preventing oxidative modification of Akt1 cysteines 296 and 310 by mutating them to serines increasedthe cell survival effects of Akt1. Thus, in neurodegenerative states such as PD, maintaining the thiolstatus of cysteines 296 and 310 in Akt1 would be critical for Akt1 kinase activity and for preventing itsdegradation by proteasomes. Preventing downregulation of Akt signaling not only has long-rangeconsequences for cell survival but could also affect the multiple roles that Akt plays, including in theAkt–mTOR signaling cascade.

& 2014 Elsevier Inc. All rights reserved.

Akt (a serine/threonine kinase, also known as protein kinase B) iscritical for cell survival [1], and its differential regulation influencessurvival of both neuronal [2,3] and nonneuronal [4,5] cells. Uponactivation of phosphoinositide 3-kinase, Akt1 is phosphorylated atThr308 in the activation loop of the kinase domain [6,7] and atSer473 in the C-terminal hydrophobic domain [8]. Dephosphoryla-tion of Akt1 predominantly by protein phosphatase 2A (PP2A)terminates the kinase activity [9]. Phosphorylated Akt1 (pAkt1)promotes cell survival by phosphorylating proapoptotic molecules,such as Bad [10,11] and caspase [12], resulting in their inactivation. Italso phosphorylates and inactivates glycogen synthase kinase 3β[13,14] and the forkhead family of transcription factors Foxo, whichmediate transcription of prodeath genes, such as Bim and Fas[15–17]. Akt1 also positively regulates transcription factors, CREBand NF-κB [18,19], which promote expression of several prosurvivalgenes, such as brain-derived neurotrophic factor and Bcl2, in addition

to affecting protein translation through the mammalian target ofrapamycin pathway.

Not surprisingly, enhancing Akt1 kinase activity has been shown tobe protective in cellular and animal models of neurodegeneration. Infact, upregulation of pAkt1 underlies neuroprotection mediated byestrogen [20–23], erythropoietin [24], preconditioning [25], andrapamycin [26]. On the other hand, dysregulation of Akt1 signalinghas been demonstrated in neurodegenerative conditions, such asHuntington disease [27], Alzheimer disease [28], amyotrophic lateralsclerosis [29], stroke, and, importantly, Parkinson disease (PD),a movement disorder involving degeneration of dopaminergic neu-rons in the substantia nigra pars compacta [30]. pAkt1 is down-regulated in the midbrain of PD patients [31] and specifically in themelanized neurons in the substantia nigra pars compacta [32], andthis is reflected in both neurotoxin-induced cellular [33] and geneticmodels of Parkinsonism [34]. More recently, we have shown thatoxidative modification of Akt1 increases its association with phospha-tase PP2A, resulting in dephosphorylation of pAkt1 and loss of itskinase activity in the ventral midbrain in a mouse model of PD [35].

The major objective in this study was to identify the mechanismunderlying the loss of Akt1 activity through oxidative modification. Animportant molecule that regulates redox status of protein thiols is

Contents lists available at ScienceDirect

journal homepage: www.elsevier.com/locate/freeradbiomed

Free Radical Biology and Medicine

http://dx.doi.org/10.1016/j.freeradbiomed.2014.06.0040891-5849/& 2014 Elsevier Inc. All rights reserved.

n Corresponding author. Fax: þ91 80 2360 3323.E-mail address: viji@cns.iisc.ernet.in (V. Ravindranath).1 These authors contributed equally to this work.

Free Radical Biology and Medicine 74 (2014) 118–128

glutaredoxin 1 (Grx1; a protein thiol disulfide oxidoreductase), whichhas glutathione (GSH; γ-glutamylcysteinyl glycine) disulfide transhy-drogenase activity, including conversion of glutathionylated proteins(PrSSG) to protein thiols (PrSH). We, therefore, perturbed protein thiolhomeostasis specifically by downregulating Grx1 using short hairpinRNA (shRNA) or by cadmium, a pan inhibitor of protein thiol disulfideoxidoreductases, including Grx1 and thioredoxin 1 [36,37]. The effectsof this experimental paradigm on Akt1 phosphorylation, kinaseactivity, and turnover were examined for endogenous Akt1 as wellas on recombinant wild-type (WT) Akt1 and compared to an Akt1mutant wherein cysteine residues at 296 and 310 were mutated toserines (C296–310S Akt1). Among others, these particular cysteineresidues were chosen because they lie in the Akt1 kinase domain,flanking the phosphorylation site at Thr308; hence their redoxmodulation could potentially affect phosphorylation and kinase activ-ity of Akt1.

Materials and methods

Antibodies and reagents

Primary antibodies against pAkt1 Ser473, pAkt1 Thr308, andAkt1 were obtained from Cell Signaling Technology. Antibodyagainst β-catenin was obtained from Santa Cruz Biotechnology,and anti-β-tubulin antibody from Sigma–Aldrich. Anti-V5 anti-body was purchased from Invitrogen. Secondary antibodies, anti-rabbit or anti-mouse conjugated to horseradish peroxidase, wereobtained from Vector Laboratories. Protease inhibitorcocktail was purchased from Calbiochem–Merck (Germany).An enhanced chemiluminescence kit was obtained from GEHealthcare Life Sciences. Akt1 kinase activity was assayed usinga kit from Invitrogen. A terminal deoxynucleotidyl transferasedUTP nick-end labeling (TUNEL) assay kit was procured fromRoche and mounting medium containing 40,6-diamidino-2-phenylindole (DAPI) from Vector Laboratories. All other reagentswere procured from either Sigma–Aldrich or Bio-Rad.

Plasmids

Following constructs were used in the study: mouse Akt1 gene(WT as well as C296–310S) codon optimized for optimal expres-sion and subcloned in-frame with a V5 tag into pcDNA 3.1 expres-sion vector (Invitrogen); pCMV scrambled RNA (scRNA) and shRNAagainst Grx1 were cloned into mU6 provector and used fordownregulation of Grx1 gene [38].

Cell culture

Neuro2A (N2A) cells were cultured in Dulbecco's modifiedEagle's medium supplemented with 10% fetal bovine serum. Fortransfection studies, cells were transfected using Lipofectamine(Invitrogen) with respective plasmids. In some experiments cellswere treated with 200 mM cadmium acetate for 3 h. To determineif cadmium-induced reactive oxygen species (ROS) generationcould be attenuated by ROS scavengers, cells were pretreated with1 mM N-acetylcysteine (NAC) for 2 h and then treated with200 mM cadmium acetate in the presence of 1 mM NAC. In otherexperiments cells were treated with 100 mM buthionine-L-sulfox-imine (BSO) every 24 h for 48 h to inhibit GSH synthesis or with100 mM H2O2 for 24 h to induce oxidative stress. Proteasomalactivity in cells was blocked by treatment with 10 μM MG132(N-(benzyloxycarbonyl)-Leu-Leu-leucinal) for 7 h.

TUNEL assay

Cells were fixed in paraformaldehyde (4% w/v) for 20 min,followed by treatment with 0.02% (v/v) Triton X containing bovineserum albumin (BSA; 4% w/v) for 20 min. TUNEL assay was carriedout as per the manufacturer's protocol. Cells were mounted withmounting medium containing DAPI. Imaging of cells was per-formed on an Axio Imager M2 (Zeiss) and cells were counted by ablinded observer using ImageJ.

Sample processing

Cells were lysed in 1� phosphate-buffered saline (PBS; pH 7.4)containing 1% Igepal CA 630 (v/v), 1 mM EDTA, 50 mM sodiumfluoride, 1 mM sodium orthovanadate, 2 mg/ml aprotinin, 10 mg/mlleupeptin, 7 mg/ml pepstatin A, 100 mg/ml phenylmethanesulfonylfluoride, and 10 ml/ml protease inhibitor cocktail and centrifugedat 10,000g for 10 min at 4 1C. The supernatant obtained was usedfor kinase assay, ROS detection assay, immunoprecipitation, orimmunoblotting as described below. Protein concentration wasestimated by Bradford's assay or bicinchoninic acid (BCA) assay.

Immunoblotting

Samples were resolved on SDS–PAGE and blotted onto poly-vinylidene difluoride or nitrocellulose membranes and blockedwith 10 mM Tris, pH 7.5, containing 0.15 M NaCl and 0.05% (v/v)Tween 20 (TBST), supplemented with 5% BSA. Incubation withprimary and secondary antibodies was carried out according to thevendor's instructions. Chemiluminescent signals were detectedand the intensities of the bands were quantitated. For detection ofreduced Akt1, cell lysates were derivatized using a thiol alkylatingagent AMS (4-acetamido-40-maleimidylstilbene-2,20-disulfonicacid, disodium salt) before SDS–PAGE as described elsewhere[35]. For nonreducing gels, the reducing agent dithiothreitol(DTT) was omitted from the Laemmli loading buffer. Appropriateloading controls were incorporated in all immunoblottingexperiments.

Detection of ROS levels

20,70-Dichlorodihydrofluorescein diacetate (H2DCFDA), an oxidant-sensitive dye that is converted to 20,70-dichlorofluorescein (DCF) by theaction of esterases and oxidation, was employed to quantitativelydetermine ROS levels. Briefly, supernatants obtained after cell lysiswere incubated with 10 μM H2DCFDA for 15 min at 37 1C. Fluores-cence measurements were made at 490 nm (excitation) and 530 nm(emission) on an Infinite M200 Pro multimode microplate reader(Tecan Group Ltd.). A standard curve of DCF was used in all experi-ments to ensure linearity of the assay. Protein amounts in sampleswere estimated using BCA assay.

Total glutathione assay

Cells were lysed in 100 mM potassium phosphate buffer(pH 7.4) containing 1 mM EDTA with addition of 5-Sulfosalicylicacid solution such that its final concentration was 4% (w/v). Thesamples were vigorously vortexed and then centrifuged at 10,000gfor 10 min. The acid-soluble supernatant was used for estimationof total glutathione using an enzymatic recycling method [39] andexpressed as nmol glutathione/mg protein.

Immunoprecipitation

In experiments wherein recombinant V5-tagged WT or C296–310SAkt1 was expressed, cell lysates were incubated with anti-V5 antibody

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128 119

coupled to Dynabeads Protein G (Invitrogen) and incubated overnightat 4 1C. Immunoprecipitated pellet was washed with 1� PBS contain-ing 0.05% Igepal CA 630 (v/v) and used for assay of Akt1 kinase activity(as outlined below) or resuspended in Laemmli's sample buffer forSDS–PAGE followed by immunoblotting. In experiments designed tocoimmunoprecipitate PP2A, N2A cell lysates were incubated withpostnuclear supernatant obtained from mouse brain, overnight at 4 1Cand the above protocol was followed thereafter.

Akt1 kinase activity assay

Akt1 kinase activity was assayed using lysates from cells trans-fected with WT or C296–310S Akt1 plasmids, both of which had a V5tag. The recombinant protein was immunoprecipitated using anti-V5antibody as mentioned above. The pellet was resuspended in kinaseassay buffer (Invitrogen) and the assay was performed. In brief,a reaction mixture containing assay buffer, 1 mM ATP, 0.2 mM DTT,and 10 mM OMNIA S/T peptide substrate was incubated for 5 min at30 1C. The resuspended immunoprecipitated pellet was then addedto the kinase reaction mixture. OMNIA S/T peptide, upon phosphor-ylation by Akt1 in the presence of ATP and Mg2þ , chelates Mg2þ toform a bridge between the Sox moiety and the phosphate group thatleads to an increase in fluorescence emission. Fluorescence wasmeasured at an excitation wavelength of 360 nm and emissionwavelength of 485 nm every 30 s for 2 h on an Infinite M200 Promultimode microplate reader (Tecan Group Ltd.). Relative fluores-cence units were plotted against time and the rate of increase influorescence, represented by the slope of the curve, served as themeasure of Akt1 activity. Reaction mixtures without enzyme orpeptide (substrate) were used as negative controls.

Quantitative real-time PCR

Total RNA was isolated from N2A cells and used for cDNAsynthesis. Quantitative real-time PCR (qRT-PCR) was performedusing the Power SYBR Green PCR master mix (Applied Biosystems)according to the manufacturer's instructions. 18s rRNAwas used asinternal control for normalization. Data were analyzed using thecompetitive threshold cycle (ΔΔCt) method [40].

Statistical analysis

Statistical differences between two groups were determinedusing Student's t test, whereas one-way analysis of variancefollowed by Newman–Keuls post hoc test was used to compareseveral groups. Correlation analysis was done using the Pearsoncorrelation method.

Results

Overexpression of WT Akt1 or Grx1 abolishes MPPþ-mediated celldeath

MPPþ , the toxic metabolite of 1-methyl-4-phenyl-1,2,3,6-tetra-hydropyridine (MPTP), is known to mediate neurodegenerationpresumably through inhibition of mitochondrial complex I andenhanced production of ROS among other mechanisms. N2A cellsurvival measured by TUNEL assay decreased significantly afterexposure to 2 mM MPPþ for 24 h (Fig. 1A). Consistent with its cellsurvival effects, overexpression of WT Akt1 was able to protect thecells from MPPþ-mediated cell death (Fig. 1B). Protein thioldisulfide oxidoreductases such as Grx1 also play an importantrole in cell survival by maintaining thiol homeostasis of proteins.Indeed, overexpression of Grx1 also significantly attenuatedMPPþ-induced cell death (Fig. 1C).

Grx1, but not GSH, is necessary for maintenance of bothphosphorylated Akt1 and total Akt1 levels

Because both Grx1 and WT Akt1 reversed MPPþ-mediated celldeath, and phosphorylation of Akt1 is known to promote cell survival,we examined the role of thiol antioxidant GSH and Grx1 in maintain-ing the phosphorylation status of endogenous Akt1 in N2A cells.Downregulation of Grx1 expression by about 40% using shRNA to Grx1was sufficient to cause a substantial loss of pAkt1 compared to cellstransfected with plasmid expressing scrambled oligonucleotidesequence. Phosphorylation of Akt1 at both Thr308 (Fig. 2A) andSer473 (Fig. 2B) showed comparable decrease. Accordingly, we studiedphosphorylation status of Akt1 only at Ser473 in subsequent experi-ments. Unexpectedly, we also observed a decrease in total endogenousAkt1 levels after shRNA-mediated downregulation of Grx1 (Fig. 2A andB). These observations were replicated upon treatment of cells withcadmium. Because cadmium is a pan inhibitor of protein thiol disulfideoxidoreductases, including Grx1 [36,41], cadmium treatment led to agreater loss of pAkt1 and total Akt1 (Fig. 2C) than shRNA-mediatedGrx1 knockdown. To assess if this phenomenon was a genericresponse to redox perturbation, we depleted the major cellularantioxidant, GSH, using BSO, an inhibitor of γ-glutamylcysteinesynthetase. Exposure to BSO (100 μM) every 24 h for 48 h depletedtotal cellular GSH by about 90% (Fig. 2F). Surprisingly, we did not find asignificant change in phosphorylation status or total levels of Akt1(Fig. 2D), indicating that GSH levels were not critical for maintenanceof pAkt1 or total Akt1 levels. To understand the lack of effect of BSO onpAkt1 and Akt1 we estimated ROS levels in cells exposed to BSO orcadmium using H2DCFDA (Fig. 2E). Consistent with previous studies,cadmium treatment led to a substantial increase in ROS levels, whichwas scavenged by NAC [42]. Although BSO treatment did result in asmall increase in ROS levels, it did not reach statistical significance.Indeed, reduction of total glutathione by more than 95% by 1mM BSO(compared to 100 μM BSO used in the present study) has been shownto only modestly increase ROS levels by about 15–20% [43]. It shouldbe noted that H2DCFDA based ROS assay does not reflect local changesin oxidizing species that can contribute to local protein oxidation,including those of critical cysteine moieties in proteins resulting inalteration of their function, such as that seen in Akt1. Nevertheless, theassay measures hydroperoxide levels in cells and provides an indica-tion of increased global oxidative stress, if any.

Mutation of critical cysteine residues in Akt1 does not affect itskinase activity

Akt1 has two cysteines in the kinase domain, flanking thephosphorylation site Thr308, at residues 296 and 310. Because oftheir proximity to the Akt1 phosphorylation site, redoxmodification ofthese cysteines could potentially cause modulation of Akt1 phosphor-ylation and kinase activity. We hypothesized that mutating thesecysteine moieties could prevent Grx1 downregulation-mediated lossof Akt1 phosphorylation, presumably caused by oxidation of thiolgroups of these residues. Mutant construct of Akt1 with cysteineresidues at 296 and 310 substituted by serines (C296–310S) wassynthesized. We then examined if these mutations interfered withbasal phosphorylation status and kinase activity of Akt1 by comparingthem with those of recombinant WT Akt1. Because both the WT andthe C296–310S Akt1 constructs had a V5 tag, we immunoprecipitatedthe recombinant proteins with anti-V5 antibody and estimated theirkinase activity without interference from endogenous Akt1. Neitherphosphorylated Akt1 levels (normalized to total Akt1; Fig. 3A) norkinase activity (Fig. 3B) of C296–310S Akt1 were statistically differentfrom those ofWTAkt1, indicating that mutation of the cysteine groupsdid not interfere with either phosphorylation status or kinase activityunder basal conditions.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128120

Grx1 downregulation has differential effects on phosphorylation andkinase activity of WT and C296-310S Akt1

ShRNA-mediated downregulation of Grx1 expression resultedin a significant reduction in the pAkt/Akt1 ratio (Fig. 4A) and Akt1kinase activity (Fig. 4D) compared to treatment with scRNA in cellsoverexpressing recombinant WT Akt1, indicating that Grx1 iscritical for maintaining Akt1 kinase activity. Further, we found apositive correlation between Grx1 mRNA levels (measured by qRT-PCR, Fig. 4B and E) and pAkt1 levels (Fig. 4C); and kinase activity ofWTAkt1 (Fig. 4F).

On the other hand, overexpression of C296–310S Akt1 inconjunction with Grx1 downregulation using shRNA did not affecteither the pAkt/Akt1 ratio (Fig. 5A) or its kinase activity (Fig. 5C).These experiments indicate that preventing oxidation of cysteineresidues at 296 and 310 in Akt1 can preserve its activity duringredox perturbation. The extent of Grx1 downregulation by shRNAfor each experiment was measured by qRT-PCR (Fig. 5B and D).

Cadmium exposure leads to increased association of Akt1 with PP2A

We then estimated the amount of reduced Akt1 after shRNA-mediated Grx1 knockdown. The levels of reduced recombinant WTAkt1, estimated by derivatizing the thiol groups with AMS, decreasedsignificantly after Grx1 knockdown (Fig. 6A), whereas that of C296–310S Akt1 did not change (Fig. 6B), indicating that cysteines at 296 and310 are the major sites of oxidation in Akt1 after Grx1 knockdown. Tovalidate our findings, we also subjected cell lysates obtained fromcadmium treated cells overexpressing WT or C296–310S Akt1 tononreducing SDS–PAGE followed by immunoblotting using antibodyagainst Akt1. Similar to Grx1 knockdown, cadmium treatment led to a

significant increase in the oxidized forms of WT Akt1 (Fig. 6C), but notof C296–310S Akt1 (Fig. 6D).

In our previous study [35] we showed that oxidation of thiolgroups in Akt1 leads to its increased association with PP2A, whichmight potentially be responsible for loss of pAkt1 in a mousemodel of Parkinsonism. Hence, we decided to employ C296–310SAkt1 to address the question of whether oxidation of criticalcysteine residues 296 and 310 contributes to its increased associa-tion with PP2A. Our initial attempts to coimmunoprecipitate PP2Awith Akt1 from N2A cells were unsuccessful (Fig. 7A, top). Hence,we used mouse brain lysate as a more abundant pool of PP2A forcoimmunoprecipitation with recombinant Akt1 expressed in N2Acells. Use of anti-V5 antibody for these experiments excluded theinteraction of endogenous Akt1 (both from mouse brain and fromN2A cells) with PP2A (Fig. 7A, bottom). N2A cells overexpressingV5-tagged WT or C296–310S Akt1 were treated with saline orcadmium acetate (200 μM) for 3 h and cell lysates were used tocoimmunoprecipitate mouse brain PP2A. After cadmium treat-ment, we observed increased association of PP2A with WTAkt1 (Fig. 7B), which was abolished by NAC pretreatment(Fig. 7D). On the other hand, interaction of C296–310S Akt1 withPP2A remained unaffected by cadmium treatment (Fig. 7C), imply-ing that oxidation of cysteine thiols at 296 and 310 in Akt1 indeedcontributes to its increased association with PP2A. This mightconstitute a potential mechanism for Akt1 dephosphorylation and,consequently, decrease in its kinase activity.

Cadmium-induced proteasomal degradation of Akt1 is dependenton its thiol oxidation state

As discussed before, cadmium treatment resulted insignificant loss of endogenous Akt1 protein (Fig. 2C). Because

Fig. 1. Cytotoxic effect of MPPþ on N2A cells and its attenuation by overexpression of Akt1 or Grx1. (A) N2A cells were treated with saline or MPPþ (2 mM) for 24 h. TUNEL-positive cells (expressed as percentage of total DAPI-stained nuclei) increased significantly after MPPþ treatment. (B) The cytotoxic effect of MPPþ on N2A cells wassignificantly attenuated by overexpression of WT Akt1 compared to cells transfected with empty vector. (C) MPPþ-mediated cytotoxicity was also attenuated byoverexpression of Grx1. Representative images and quantification thereof are shown. Data are presented as mean 7 SEM (n ¼ 3–6 individual experiments). *p o 0.001 and#p o 0.001, significantly different from controls and cells treated with MPPþ , respectively.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128 121

Akt1 is known to be degraded through a proteasomalpathway [44,45], we treated cells with MG132, a potent andcell-permeative proteasome inhibitor, and then assayed the levelsof endogenous Akt1. Exposure of cells to MG132 (10 mM) alone for

7 h led to enhanced endogenous Akt1 levels (Fig. 8A), indicatingthat Akt1 is indeed degraded through the proteasome. Theefficiency of proteasome inhibition was assessed by enhancedaccumulation of β-catenin (Fig. 8A) [46]. Next, we tested

Fig. 2. Inhibition or knockdown of Grx1 but not GSH depletion leads to loss of endogenous pAkt1 and Akt1 protein. (A) N2A cells were transfected with scRNA or shRNA toGrx1 and the levels of pAkt1 (Thr308) were measured by immunoblotting. Significant loss of endogenous pAkt1 and total Akt1 protein was seen after Grx1 knockdowncompared to scRNA control. (B) In a parallel set of experiments using scRNA or shRNA against Grx1, pAkt1 (Ser473) levels were assayed. Loss of pAkt1 (Ser473) was similar tothat of Thr308. (C) Cells were treated with saline (control) or cadmium acetate (200 mM) for 3 h and endogenous Akt1 and pAkt1 levels were measured. pAkt1, pAkt/Akt1ratio, and total Akt1 decreased significantly. (D) Cells were treated with BSO (100 mM) every 24 h for 48 h and endogenous pAkt1 and Akt1 levels were analyzed. Nosignificant change in the levels of pAkt1, pAkt/Akt1, or Akt1 was observed. Representative immunoblots and quantitation are depicted in (A–D). (E) Cells were treated withsaline (control), cadmium acetate (200 μM) for 3 h, BSO (100 μM) every 24 h for 48 h, or H2O2 (100 μM) for 24 h and ROS levels were assayed. To scavenge cadmiumgenerated ROS, cells were pretreated with 1 mM NAC followed by exposure to cadmium in the presence of 1 mM NAC. ROS levels are represented as DCF fluorescence/mgprotein. ROS levels were significantly higher in cells treated with cadmium or H2O2, which could be abolished by NAC treatment. A small increase in ROS levels was seen incells treated with BSO; however, it did not reach statistical significance with respect to control. (F) Total glutathione level was quantified in cells treated with saline or BSO(100 μM) every 24 h for 48 h and is represented as total glutathione/mg protein. BSO treatment led to a substantial decrease in the total glutathione levels compared to thecontrol. Data are presented as mean 7 SD (n ¼ 3–9 independent experiments). *p o 0.05 and #p o 0.05, significantly different compared to respective control andbetween groups, respectively.

Fig. 3. The cysteine mutant of Akt1 (C296–310S) does not show altered phosphorylation status or kinase activity at basal states. (A) N2A cells were transfected withrecombinant V5-tagged WT Akt1 or C296–310S Akt1 construct and pAkt1 levels were assessed by immunoblotting. Representative immunoblot and quantification areshown. The pAkt/Akt1 ratio of C296–310S Akt1 was similar to that of WT Akt1. (B) Akt1 kinase activity was assayed after immunoprecipitation with V5 antibody and isdepicted as the rate of increase in relative fluorescence units (RFU) in cells overexpressing WT or C296–310S Akt1. A representative curve and quantification are depicted. Thekinase activity of C296–310S Akt1 was similar to that of WT Akt1. Data are presented as mean 7 SD (n ¼ 6 individual experiments).

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128122

whether Akt1 loss induced by cadmium could be rescuedby MG132. Indeed, inhibition of proteasomal activity rescuedN2A cells from cadmium-induced Akt1 loss (Fig. 8B). BecauseNAC could efficiently scavenge ROS generated after cadmiumtreatment (Fig. 2E), we examined if it could also abolishcadmium-induced loss of Akt1. Indeed, NAC could efficientlyabolish cadmium-induced Akt1 loss (Fig. 8C), indicating thatoxidation of thiols in Akt1 leads to its increased proteasomaldegradation.

During our experiments, we noticed that, unlike endogenousand overexpressed WT Akt1, C296–310S Akt1 seemed to beresistant to protein loss induced by Grx1 knockdown or bycadmium. Hence, we hypothesized that Akt1 in its oxidized statecould be a target for proteasomal degradation. Because criticalcysteines that are oxidatively modified after cadmium treatmentare mutated in C296–310S Akt1, presumably it could be protectedfrom proteasomal degradation induced by cadmium. To this end,we overexpressed WT or C296–310S Akt1 in N2A cells and then

Fig. 4. Grx1 downregulation leads to loss of phosphorylation and kinase activity of overexpressed WT Akt1. N2A cells were transfected with WT Akt1 along with scrambledor shRNA against Grx1, and (A) pAkt1 levels and (D) Akt1 kinase activity were measured. Downregulation of Grx1 resulted in a significant reduction in both (A) pAkt/Akt1ratio and (D) kinase activity. Extent of downregulation of Grx1 mRNA for the respective experiments is depicted in (B) and (E). Relative levels of (C) pAkt1 and (F) Akt1 kinaseactivity of recombinant WT Akt1 correlated significantly with the extent of Grx1 downregulation. Data are presented as mean 7 SD (n ¼ 5 or 6 independent experiments).*p o 0.01, significant difference between the groups.

Fig. 5. Grx1 knockdown has no effect on phosphorylation status or kinase activity of C296–310S Akt1. N2A cells were transfected with C296–310S Akt1 along with scrambledor shRNA against Grx1, and (A) pAkt1 levels and (C) Akt1 kinase activity were measured. Downregulation of Grx1 had no effect on (A) pAkt/Akt1 ratio or (C) kinase activity ofC296–310S Akt1. Downregulation of Grx1 mRNA levels for the respective experiments is depicted in (B) and (D). Data are presented as mean 7 SD (n ¼ 5 or 6 independentexperiments). *p o 0.05, significant difference between the groups.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128 123

exposed the cells to cadmium acetate (200 μM) for 3 h. Whereascadmium exposure induced a reduction in WT Akt1 (Fig. 8D),mutation of critical cysteine residues protected C296–310S Akt1from this degradation (Fig. 8E). This indicates that WT Akt1 isdegraded under oxidative stress, and mutation of the criticalcysteines, and hence preventing their oxidation, protects C296–310S Akt1 from proteasomal degradation.

Mutation of critical cysteine residues in Akt1 protects againstcadmium-induced cell death

Finally, we examined the degree of protection afforded by cysteinedouble-mutant C296–310S Akt1 in comparison to WT Akt1 againstcadmium-mediated cell death [36]. N2A cells overexpressing WT orC296–310S Akt1 were treated with saline or cadmium acetate

Fig. 6. Downregulation of Grx1 leads to loss of reduced WT Akt1 but not C296–310S Akt1. Cells were transfected with V5-tagged (A) WT or (B) C296–310S Akt1 along withscRNA or shRNA to Grx1 and lysates were derivatized with the thiol alkylating agent AMS to measure the levels of reduced Akt1. Samples were run on reducing PAGE andimmunostained with anti-V5 antibody. Whereas the levels of reduced WT Akt1 were significantly lowered by Grx1 knockdown, they remained unaffected in the cellstransfected with the C296–310S Akt1. The thinner arrows in the representative immunoblots denote AMS-derivatized Akt1 and thicker arrows mark unmodified Akt1 (A andB). Cells overexpressing V5-tagged (C) WT or (D) C296–310S Akt1 were treated with cadmium acetate (200 μM) for 3 h and lysates were separated on SDS–PAGE undernonreducing conditions. The oxidized high-molecular-weight Akt1 forms are shown by thinner arrows and monomeric Akt1 is denoted by thicker arrows in therepresentative immunoblots. Reduced samples were also subjected to SDS–PAGE for comparison in each case (C and D). Data are presented as mean 7 SEM (n ¼ 6–9independent experiments). *p o 0.05, significant difference compared to respective control.

Fig. 7. Mutation of the critical cysteines in Akt1 abolishes cadmium-induced increase in its association with PP2A. (A) N2A cells overexpressing V5-tagged WT Akt1 wereharvested, and immunoprecipitation was performed with antibody against V5 followed by immunoblotting. Coimmunoprecipitation of PP2A with WT Akt1 was not clearlydiscernable in cell lysates. However, addition of brain lysate before immunoprecipitation of N2A cell lysate led to coimmunoprecipitation of PP2A with WT Akt1. Cellsexpressing (B) WT or (C) C296–310S Akt1 were treated with cadmium acetate (200 mM) for 3 h and the lysates were used for coimmunoprecipitation of PP2A along withmouse brain homogenate using anti-V5 antibody. Immunoblots were probed for both Akt1 and PP2A. (D) Lysate from cells overexpressing V5-taggedWT Akt1 and pretreatedwith 1 mM NAC for 2 h followed by cadmium incubation in presence of NAC were used for coimmunoprecipitation of PP2A. Association of PP2A is depicted as PP2A levelsrelative to Akt1 in the immunoprecipitate. Representative blots and quantification are shown. Cadmium exposure significantly increased association of PP2A with WT Akt1but not with C296–310S Akt1. Cadmium-induced increased association of PP2Awith WT Akt1 was abolished by NAC. Data are presented as mean 7 SEM (n ¼ 3 independentexperiments). *p o 0.05, significantly different from the respective control.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128124

(200 μM) for 3 h and cell viability was assayed by TUNEL staining.Expression of C296–310S Akt1 attenuated cytotoxicity induced bycadmium as seen by reduced number of TUNEL-positive cells com-pared to WT Akt1 (Fig. 9). Hence, preventing oxidation of Akt1 at thecritical cysteine residues at 296 and 310 can reduce both its depho-sphorylation and its proteasomal degradation, which translates intoenhanced cell survival under cytotoxic stress.

Discussion

Protein thiol homeostasis is essential for normal functioning ofseveral proteins including enzymes. During oxidative stress, whenreducing equivalents are limiting, ROS generation leads to oxida-tive modification of proteins, and one of the main targets are thethiol groups of cysteines. Protein thiols can be oxidized by ROS tothiyl radicals or progressively to sulfenic, sulfinic, and sulfonicacids. Sulfenic acids react with GSH to form PrSSG, which can befurther oxidized to form intramolecular and intermolecular pro-tein mixed disulfides (PrSSPr). Cellular antioxidant systems thathelp maintain thiol homeostasis are (i) the most abundant cellularnonprotein thiol, GSH, and (ii) protein thiol disulfide oxidoreduc-tases, which include, among others, cytosolic and mitochondrialglutaredoxins and thioredoxins [47]. Although GSH can scavengeROS directly and regulate protein thiol oxidation by their rever-sible glutathionylation to form PrSSG, protein thiol disulfideoxidoreductases catalyze the formation of disulfides and theirreduction to thiols. Cytosolic Grx1 catalyzes both glutathionylationand deglutathionylation reactions and its activity is maintained byGSH reductase, utilizing reducing equivalents from NADPH andGSH. Grx1 can act both as a glutathionylating enzyme in presenceof an oxidative stimulus and as a deglutathionylase when oxida-tive stress subsides, thereby stabilizing proteins during oxidativestress and preventing their irreversible modification to sulfonicacids. Therefore, downregulation of Grx1 leads to increased levelsof PrSSG and PrSSPr. It is being increasingly recognized thatreversible oxidation of thiols is an important regulatory mechan-ism akin to phosphorylation [48]. Indeed, this study shows thatsustained oxidative modification of thiol groups can result inaberrant dephosphorylation and/or increased proteasomal degra-dation as seen in the context of Akt1.

ShRNA-mediated Grx1 downregulation, but not depletion ofGSH, led to loss of pAkt1 and its kinase activity, indicating thatGrx1 is critical for maintaining kinase activity of Akt1. Although ithas been shown that overexpression of Grx1 protects againstH2O2-induced cell death by maintaining Akt1 kinase activity[41], we show for the first time that Grx1 is critical for main-tenance of constitutive Akt1 activity, and decreasing its expressionby 40% is sufficient to cause a significant loss of Akt1 kinaseactivity by over 25%. This effect was amplified when cells wereexposed to cadmium, a pan inhibitor of protein thiol disulfideoxidoreductases [36,37], wherein we saw 66% loss of pAkt1(Fig. 2C). To understand the mechanism underlying the effect ofGrx1 downregulation on Akt1 function, we mutated cysteineresidues at 296 and 310 flanking the Thr308 phosphorylation sitein the Akt1 kinase domain.

Mutation of Cys296 and 310 to Ser did not affect Akt1 kinaseactivity per se, and it was similar to that of WT Akt1 (Fig. 3).However, unlike the recombinant WT Akt1, downregulation ofGrx1 did not lower either C296–310S pAkt1 levels or its kinaseactivity (Figs. 4 and 5), indicating that although cysteineresidues 296 and 310 per se are not critical for Akt1 phosphor-ylation and its downstream kinase activity, oxidation of thesecysteines leads to loss of Akt1 phosphorylation and hence itsactivity. To test this, we derivatized protein thiols using thealkylating agent AMS and immunoprobed to assay reduced WTand C296–310S recombinant Akt1. We found that, whereasthere was a loss of reduced WT Akt1, there was no change inthe levels of reduced C296–310S Akt1 (Fig. 6A and B). This wasvalidated by immunoblotting for Akt1 after SDS–PAGE undernonreducing conditions (Fig. 6C and D), wherein we observedhigher molecular weight immunostained bands at higher inten-sities in WT Akt1 but not in C296–310S mutant, implying thatWT Akt1, but not C296–310S Akt1, is excessively oxidized aftercadmium treatment. Our results indicate that cysteines at 296and 310 are oxidatively modified upon Grx1 downregulationand may represent unique redox regulatory sites that are criticalfor maintaining Akt1 activity when Grx1 is functionally com-promised. It needs to be determined if the cysteines in Akt1 alsoplay similar roles.

In an earlier study, using heart-derived H9c2 cells, Murata et al.[41] had shown that H2O2 treatment leads to loss of Akt1, which is

Fig. 8. Cadmium-induced loss of Akt1 is prevented by proteasomal inhibitor. (A) Cells were treated with dimethyl sulfoxide or MG132 (10 mM) for 7 h and the inhibition ofproteasome activity was confirmed by significant increase in β-catenin and Akt1 protein levels. (B) Exposure of cells to cadmium acetate (200 mM) after pretreatment withMG132 prevented cadmium-induced Akt1 loss. (C) Pretreatment of cells with NAC also abolished cadmium-induced Akt1 loss. N2A cells overexpressing (D) WT or (E) C296–310S Akt1 were treated with saline or cadmium acetate (200 mM) for 3 h and Akt1 levels were measured. Whereas exposure to cadmium resulted in significant loss of WTAkt1, C296–310S Akt1 levels remained unaffected, indicating that the cysteine double mutant of Akt1 was resistant to cadmium-induced degradation. Representative blotsand quantification are shown. Data are presented as mean 7 SEM (n ¼ 3–8 individual experiments). *p o 0.05 and #p o 0.05, significant difference between treated groupand respective control and between cells treated with cadmium and with cadmium in the presence of MG132, respectively.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128 125

rescued by Grx1 overexpression. However, they did not see anymobility shift indicative of AMS alkylation in C297–311S Akt1 andinterpreted it as formation of an intramolecular disulfide bondbetween Cys297 and Cys311 [41]. However, in N2A cells under

normal conditions, as well as after Grx1 knockdown, we do seethe presence of reduced cysteines in C296–310S Akt1 as assessedby AMS derivatization, which remains unchanged under bothconditions. A more recent study [49] has shown that platelet-derived growth factor-induced oxidative stress inhibits Akt2activity through oxidative modification of Cys124, which liesbetween the N-terminal pleckstrin homology domain and thekinase catalytic domain. Cys124 is, however, not conserved inAkt1. Whereas Wani et al. [49] observed that recombinant Akt1activity was not altered upon H2O2 treatment in a cell-freesystem, in N2A cells Grx1 downregulation leads to loss of Akt1kinase activity, indicating that in intact cells Akt1 is susceptible tooxidative modification resulting in loss of its activity, which is notseen for C296–310S Akt1.

Our earlier study demonstrated that oxidative modification ofcysteine residues in Akt1 increases its association with phospha-tase PP2A, leading to its dephosphorylation and hence loss ofkinase activity, in vivo in brain after MPTP administration to mice[35]. However, association of Akt1 with PP2A in N2A cells wasobscure, possibly because of low levels of PP2A in these cells(Fig. 7A). The paucity of association was seen both under consti-tutive conditions and after Grx1 downregulation (data not shown).However, when immunoprecipitation was carried out in presenceof mouse brain homogenate, recombinant WT Akt1 coimmuno-precipitated with PP2A (Fig. 7A) and this interaction was enhancedwhen cells were pretreated with cadmium (Fig. 7B). However, incells expressing C296–310S Akt1, cadmium treatment did notenhance this association (Fig. 7C), indicating that oxidative mod-ification of cysteines 296 and 310 contributes to changes ininteraction of Akt1 with PP2A and, hence, its phosphorylationstatus and kinase activity. Interestingly, pretreatment with NAC, aROS scavenger, could abolish cadmium-induced increased associa-tion of WT Akt1 with PP2A (Fig. 7D), providing further proof thatredox state of the critical thiols in Akt1 is indeed a determinant ofits association with PP2A.

Another interesting observation of this study is the decrease intotal Akt1 levels after Grx1 downregulation by shRNA or cadmiumexposure (Figs. 2 and 8B). Treatment with the proteasomal inhibitorMG132 prevented this loss, indicating that cadmium treatmentinduces Akt1 degradation through proteasomes (Fig. 8B). Althoughprevious reports had suggested that Akt1 is degraded by proteasome[44,45], understanding of the mechanism had been lacking. Interest-ingly, cadmium-induced loss of Akt1 protein was seen only inendogenous and recombinant WT Akt1 and not in the C296–310SAkt1, indicating that the loss was related to redox status of cysteineresidues at 296 and 310 (Fig. 8D and E). Hence, oxidative state ofcysteines at 296 and 310 might play a critical role in targeting Akt1 tothe proteasome.

In conclusion, our study demonstrates that oxidative modifica-tion of Cys296 and Cys310 in Akt1 leads to loss of its phosphor-ylation status and downstream kinase activity possibly throughincreased interaction of Akt1 with PP2A, which is not seen whenthe cysteine residues are mutated, indicating the importance ofreduced cysteine moieties 296 and 310 in maintaining kinaseactivity of Akt1. Further, loss of Akt1 kinase activity in N2A cells isalso related to increased proteasomal degradation of oxidizedAkt1, which is abolished when cysteine residues are mutated.Although mutation of the two critical cysteines does not interferewith kinase activity of Akt1 directly, they are important formaintenance of pAkt1 and optimum turnover of Akt1, and conse-quently for cell survival (Fig. 9). Preservation of protein thiol statusis thus critical for maintaining the function of redox-sensitivemolecules, such as Akt1. This has important implications inneurodegenerative conditions wherein maintenance of Akt1 levelsand activity is critical for cell survival, especially during oxidativestress.

Fig. 9. Mutation of the critical cysteines in Akt1 prevents cadmium-mediated celldeath. N2A cells overexpressing WT Akt1 or C296–310S Akt1 were treated withsaline or cadmium acetate (200 mM) for 3 h. Cell viability was assessed by TUNELassay. TUNEL-positive cells were counted and are expressed as percentage of totalDAPI-stained nuclei. Representative images and quantification are shown. Over-expression of C296–310S Akt1 significantly attenuated cadmium-induced cytotoxi-city compared to WT Akt1 as seen by reduction in the number of TUNEL-positivecells. Data are presented as mean 7 SD (n ¼ 4 independent experiments).*p o 0.001 and #p o 0.001, significant difference with respect to vehicle-treatedcontrols and between cells treated with cadmium acetate, respectively.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128126

Authors’ contributions

F.A., P.N., L.D., and V.R. designed the experiments. F.A., P.N., andL.D. performed the experiments and F.A., P.N., and V.R. interpretedthe results and wrote the paper.

Acknowledgments

This work was supported by grants from the Departments ofScience & Technology and Biotechnology, Government of India.We thank Eisha Shaw and Ajit Ray for their help with someexperiments and in the preparation of the manuscript.

References

[1] Dudek, H.; Datta, S. R.; Franke, T. F.; Birnbaum, M. J.; Yao, R.; Cooper, G. M.;Segal, R. A.; Kaplan, D. R.; Greenberg, M. E. Regulation of neuronal survival bythe serine–threonine protein kinase Akt. Science 275:661–665; 1997.

[2] Van der Heide, L. P.; Ramakers, G. M.; Smidt, M. P. Insulin signaling in thecentral nervous system: learning to survive. Prog. Neurobiol. 79:205–221;2006.

[3] Ghasemi, R.; Haeri, A.; Dargahi, L.; Mohamed, Z.; Ahmadiani, A. Insulin in thebrain: sources, localization and functions. Mol. Neurobiol. 47:145–171; 2013.

[4] Datta, S. R.; Brunet, A.; Greenberg, M. E. Cellular survival: a play in three Akts.Genes Dev. 13:2905–2927; 1999.

[5] Downward, J. PI3-kinase, Akt and cell survival. Semin. Cell. Dev. Biol. 15:177–182; 2004.

[6] Alessi, D. R.; James, S. R.; Downes, C. P.; Holmes, A. B.; Gaffney, P. R.; Reese, C.B.; Cohen, P. Characterization of a 3-phosphoinositide-dependent proteinkinase which phosphorylates and activates protein kinase Balpha. Curr. Biol.7:261–269; 1997.

[7] Ivetac, I.; Gurung, R.; Hakim, S.; Horan, K. A.; Sheffield, D. A.; Binge, L. C.;Majerus, P. W.; Tiganis, T.; Mitchell, C. A. Regulation of PI(3)K/Akt signallingand cellular transformation by inositol polyphosphate 4-phosphatase-1. EMBORep. 10:487–493; 2009.

[8] Sarbassov, D. D.; Guertin, D. A.; Ali, S. M.; Sabatini, D. M. Phosphorylation andregulation of Akt/PKB by the rictor–mTOR complex. Science 307:1098–1101;2005.

[9] Andjelkovic, M.; Jakubowicz, T.; Cron, P.; Ming, X.; Han, J.; Hemmings, B. A.Activation and phosphorylation of a pleckstrin homology domain containingprotein kinase (RAC-PK/PKB) promoted by serum and protein phosphataseinhibitors. Proc. Natl. Acad. Sci. USA 93:5699–5704; 1996.

[10] Datta, S. R.; Dudek, H.; Tao, X.; Masters, S.; Fu, H.; Gotoh, Y.; Greenberg, M. E.Akt phosphorylation of BAD couples survival signals to the cell-intrinsic deathmachinery. Cell 91:231–241; 1997.

[11] Noshita, N.; Lewen, A.; Sugawara, T.; Chan, P. H. Evidence of phosphorylationof Akt and neuronal survival after transient focal cerebral ischemia in mice. J.Cereb. Blood Flow Metab. 21:1442–1450; 2001.

[12] Cardone, M. H.; Roy, N.; Stennicke, H. R.; Salvesen, G. S.; Franke, T. F.;Stanbridge, E.; Frisch, S.; Reed, J. C. Regulation of cell death proteasecaspase-9 by phosphorylation. Science 282:1318–1321; 1998.

[13] Cross, D. A.; Alessi, D. R.; Cohen, P.; Andjelkovich, M.; Hemmings, B. A.Inhibition of glycogen synthase kinase-3 by insulin mediated by proteinkinase B. Nature 378:785–789; 1995.

[14] Case, N.; Thomas, J.; Sen, B.; Styner, M.; Xie, Z.; Galior, K.; Rubin, J. Mechanicalregulation of glycogen synthase kinase 3 (GSK3) in mesenchymal stem cells isdependent on Akt protein serine 473 phosphorylation via mTORC2 protein.J. Biol. Chem. 286:39450–39456; 2011.

[15] Brunet, A.; Bonni, A.; Zigmond, M. J.; Lin, M. Z.; Juo, P.; Hu, L. S.; Anderson, M.J.; Arden, K. C.; Blenis, J.; Greenberg, M. E. Akt promotes cell survival byphosphorylating and inhibiting a forkhead transcription factor. Cell 96:857–-868; 1999.

[16] Dijkers, P. F.; Medema, R. H.; Lammers, J. W.; Koenderman, L.; Coffer, P. J.Expression of the pro-apoptotic Bcl-2 family member Bim is regulated by theforkhead transcription factor FKHR-L1. Curr. Biol. 10:1201–1204; 2000.

[17] Gilley, J.; Coffer, P. J.; Ham, J. FOXO transcription factors directly activate bimgene expression and promote apoptosis in sympathetic neurons. J. Cell Biol.162:613–622; 2003.

[18] Du, K.; Montminy, M. CREB is a regulatory target for the protein kinase Akt/PKB. J. Biol. Chem. 273:32377–32379; 1998.

[19] Simao, F.; Matte, A.; Pagnussat, A. S.; Netto, C. A.; Salbego, C. G. Resveratrolprevents CA1 neurons against ischemic injury by parallel modulation of bothGSK-3β and CREB through PI3-K/Akt pathways. Eur. J. Neurosci. 36:2899–2905;2012.

[20] D’Astous, M.; Mickley, K. R.; Dluzen, D. E.; Di Paolo, T. Differential protectiveproperties of estradiol and tamoxifen against methamphetamine-inducednigrostriatal dopaminergic toxicity in mice. Neuroendocrinology 82:111–120;2005.

[21] D’Astous, M.; Mendez, P.; Morissette, M.; Garcia-Segura, L. M.; Di Paolo, T.Implication of the phosphatidylinositol-3 kinase/protein kinase B signalling

pathway in the neuroprotective effect of estradiol in the striatum of 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine mice. Mol. Pharmacol. 69:1492–1498;2006.

[22] Quesada, A.; Lee, B. Y.; Micevych, P. E. PI3 kinase/Akt activation mediatesestrogen and IGF-1 nigral DA neuronal neuroprotection against a unilateral ratmodel of Parkinson's disease. Dev. Neurobiol. 68:632–644; 2008.

[23] Urata, Y.; Ihara, Y.; Murata, H.; Goto, S.; Koji, T.; Yodoi, J.; Inoue, S.; Kondo, T.17Beta-estradiol protects against oxidative stress-induced cell death throughthe glutathione/glutaredoxin-dependent redox regulation of Akt in myocar-diac H9c2 cells. J. Biol. Chem. 281:13092–13102; 2006.

[24] Wu, Y.; Shang, Y.; Sun, S.; Liang, H.; Liu, R. Erythropoietin prevents PC12 cellsfrom 1-methyl-4-phenylpyridinium ion-induced apoptosis via the Akt/GSK-3beta/caspase-3 mediated signaling pathway. Apoptosis 12:365–375; 2007.

[25] Leak, R. K.; Liou, A. K.; Zigmond, M. J. Effect of sublethal 6-hydroxydopamineon the response to subsequent oxidative stress in dopaminergic cells:evidence for preconditioning. J. Neurochem. 99:1151–1163; 2006.

[26] Malagelada, C.; Jin, Z. H.; Jackson-Lewis, V.; Przedborski, S.; Greene, L. A.Rapamycin protects against neuron death in in vitro and in vivo models ofParkinson's disease. J. Neurosci. 30:1166–1175; 2010.

[27] Humbert, S.; Bryson, E. A.; Cordelieres, F. P.; Connors, N. C.; Datta, S. R.;Finkbeiner, S.; Greenberg, M. E.; Saudou, F. The IGF-1/Akt pathway isneuroprotective in Huntington's disease and involves Huntingtin phosphor-ylation by Akt. Dev. Cell 2:831–837; 2002.

[28] Griffin, R. J.; Moloney, A.; Kelliher, M.; Johnston, J. A.; Ravid, R.; Dockery, P.;O’Connor, R.; O’Neill, C. Activation of Akt/PKB, increased phosphorylation ofAkt substrates and loss and altered distribution of Akt and PTEN are featuresof Alzheimer's disease pathology. J. Neurochem. 93:105–117; 2005.

[29] Dewil, M.; Lambrechts, D.; Sciot, R.; Shaw, P. J.; Ince, P. G.; Robberecht, W.; Vanden Bosch, L. Vascular endothelial growth factor counteracts the loss ofphospho-Akt preceding motor neurone degeneration in amyotrophic lateralsclerosis. Neuropathol. Appl. Neurobiol. 33:499–509; 2007.

[30] Ries, V.; Henchcliffe, C.; Kareva, T.; Rzhetskaya, M.; Bland, R.; During, M. J.;Kholodilov, N.; Burke, R. E. Oncoprotein Akt/PKB induces trophic effects inmurine models of Parkinson's disease. Proc. Natl. Acad. Sci. USA 103:18757–18762; 2006.

[31] Timmons, S.; Coakley, M. F.; Moloney, A. M.; O’Neill, C. Akt signal transductiondysfunction in Parkinson's disease. Neurosci. Lett. 467:30–35; 2009.

[32] Malagelada, C.; Jin, Z. H.; Greene, L. A. RTP801 is induced in Parkinson'sdisease and mediates neuron death by inhibiting Akt phosphorylation/activa-tion. J. Neurosci. 28:14363–14371; 2008.

[33] Rodriguez-Blanco, J.; Martin, V.; Herrera, F.; Garcia-Santos, G.; Antolin, I.;Rodriguez, C. Intracellular signaling pathways involved in post-mitotic dopa-minergic PC12 cell death induced by 6-hydroxydopamine. J. Neurochem.107:127–140; 2008.

[34] Yang, Y.; Gehrke, S.; Haque, M. E.; Imai, Y.; Kosek, J.; Yang, L.; Beal, M. F.;Nishimura, I.; Wakamatsu, K.; Ito, S.; Takahashi, R.; Lu, B. Inactivation ofDrosophila DJ-1 leads to impairments of oxidative stress response andphosphatidylinositol 3-kinase/Akt signaling. Proc. Natl. Acad. Sci. USA102:13670–13675; 2005.

[35] Durgadoss, L.; Nidadavolu, P.; Valli, R. K.; Saeed, U.; Mishra, M.; Seth, P.;Ravindranath, V. Redox modification of Akt in mouse midbrain mediated bythe dopaminergic neurotoxin, MPTP leads to downregulation of pAkt. FASEB J.26:1473–1483; 2012.

[36] Chrestensen, C. A.; Starke, D. W.; Mieyal, J. J. Acute cadmium exposureinactivates thioltransferase (glutaredoxin), inhibits intracellular reduction ofprotein–glutathionyl-mixed disulfides, and initiates apoptosis. J. Biol. Chem.275:26556–26565; 2000.

[37] Hansen, J. M.; Zhang, H.; Jones, D. P. Differential oxidation of thioredoxin-1,thioredoxin-2, and glutathione by metal ions. Free Radic. Biol. Med. 40:138–145; 2006.

[38] Saeed, U.; Ray, A.; Valli, R. K.; Kumar, A. M.; Ravindranath, V. DJ-1 loss byglutaredoxin but not glutathione depletion triggers Daxx translocation andcell death. Antioxid. Redox Signaling 13:127–144; 2010.

[39] Tietze, F. Enzymic method for quantitative determination of nanogramamounts of total and oxidized glutathione: applications to mammalian bloodand other tissues. Anal. Biochem. 27:502–522; 1969.

[40] Schmittgen, T. D.; Livak, K. J. Analyzing real-time PCR data by the comparativeC(T) method. Nat. Protoc. 3:1101–1108; 2008.

[41] Murata, H.; Ihara, Y.; Nakamura, H.; Yodoi, J.; Sumikawa, K.; Kondo, T.Glutaredoxin exerts an antiapoptotic effect by regulating the redox state ofAkt. J. Biol. Chem. 278:50226–50233; 2003.

[42] Oh, S. H.; Lim, S. C. A rapid and transient ROS generation by cadmium triggersapoptosis via caspase-dependent pathway in HepG2 cells and this is inhibitedthrough N-acetylcysteine-mediated catalase upregulation. Toxicol. Appl. Phar-macol. 212:212–223; 2006.

[43] Armstrong, J. S.; Steinauer, K. K.; Hornung, B.; Irish, J. M.; Lecane, P.; Birrell, G. W.;Peehl, D. M.; Knox, S. J. Role of glutathione depletion and reactive oxygen speciesgeneration in apoptotic signaling in a human B lymphoma cell line. Cell DeathDiffer. 9:252–263; 2002.

[44] Basso, A. D.; Solit, D. B.; Chiosis, G.; Giri, B.; Tsichlis, P.; Rosen, N. Akt forms anintracellular complex with heat shock protein 90 (Hsp90) and Cdc37 and isdestabilized by inhibitors of Hsp90 function. J. Biol. Chem. 277:39858–39866;2002.

[45] Adachi, M.; Katsumura, K. R.; Fujii, K.; Kobayashi, S.; Aoki, H.; Matsuzaki, M.Proteasome-dependent decrease in Akt by growth factors in vascular smoothmuscle cells. FEBS Lett. 554:77–78; 2003.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128 127

[46] Kim, N. G.; Xu, C.; Gumbiner, B. M. Identification of targets of the Wntpathway destruction complex in addition to beta-catenin. Proc. Natl. Acad. Sci.USA 106:5165–5170; 2009.

[47] Ghezzi, P. Protein glutathionylation in health and disease. Biochim. Biophys.Acta 1830:3165–3172; 2013.

[48] Go, Y. M.; Jones, D. P. Thiol/disulfide redox states in signaling and sensing. Crit.Rev. Biochem. Mol. Biol. 48:173–181; 2013.

[49] Wani, R.; Qian, J.; Yin, L.; Bechtold, E.; King, S. B.; Poole, L. B.; Paek, E.; Tsang, A. W.;Furdui, C. M. Isoform-specific regulation of Akt by PDGF-induced reactive oxygenspecies. Proc. Natl. Acad. Sci. USA 108:10550–10555; 2011.

F. Ahmad et al. / Free Radical Biology and Medicine 74 (2014) 118–128128