Synthetic Triterpenoids Attenuate Cytotoxic Retinal Injury: Cross-talk between Nrf2 and PI3K/AKT Signaling through Inhibition of the Lipid Phosphatase PTEN

Abstract

Purpose.: Evidence implicating oxidative stress in the pathogenesis of age-related macular degeneration suggests that antioxidant therapy could play a role in preventing its progression. The aim of this study was to determine whether derivatives of the triterpenoid (TP) 2-cyano-3,12-dioxooleana-1,9-dien-28-oic acid (CDDO; CDDO-imidazolide [-Im], CDDO-ethylamide [-EA], and CDDO-trifluoroethylamide [-TFEA]) confer cytoprotection from oxidative- and photooxidative-induced cellular damage and to explore the molecular mechanisms of this cytoprotection.

Methods.: Retinal pigment epithelial and retinal photoreceptor cell lines were treated with TP derivatives. Induction of Nrf2 signaling was measured by reporter assay. Cytoprotection was quantified by MTT assay. To determine whether TPs confer in vivo cytoprotection, BALB/c mice were pretreated with CDDO-TFEA, and retinal degeneration was induced by light exposure. To explore the association of TPs with PTEN, a biotinylated derivative of CDDO (CDDO-Bt) was used.

Results.: Treatment with CDDO-Im–, -TFEA–, or -EA–induced Nrf2 signaling and TP pretreatment protected retinal cell lines from oxidant-induced cell death. The antioxidant and cytoprotective potential of these compounds was then examined in vivo. Treatment of BALB/c mice with CDDO-TFEA induced the Nrf2-regulated transcripts glcl and trx1 in retinal tissue and was protective from photooxidative retinal damage. Treatment with CDDO-Im leads to phosphorylation of AKT. CDDO-Bt directly binds cysteine 124 within PTEN′s active site and inhibits PTEN′s lipid phosphatase activity in vitro. Thus the stimulation of AKT activity is mediated by TP inhibition of PTEN activity.

Conclusions.: These studies highlight the potential of TPs in retinal cytoprotection and implicate PTEN inhibition as a target in cytoprotection.

Age-related macular degeneration (AMD) is the leading cause of irreversible blindness among the elderly in developed countries. ^1,2 AMD is characterized by progressive degeneration of the area of central vision, the macula, and can potentially lead to devastating vision loss. In the United States approximately 1.75 million people have the advanced form of this disease ³ ; this number will grow as the population ages. 

A growing body of evidence suggests that cumulative oxidative injury occurring with aging contributes to the progression of AMD. ⁴ The Age-Related Eye Disease Study showed that therapy with an antioxidant cocktail containing β-carotene, vitamin C, and vitamin E prevented the progression of intermediate, nonneovascular AMD. ⁵ This finding added to multiple animal studies that demonstrated protection from retinal damage by antioxidant supplementation. ⁶ Additional experimental evidence of a role for oxidative stress in AMD pathogenesis is provided by several murine models of AMD that closely link oxidative stress and pathologic findings suggestive of AMD. ^7–10 Visual impairment in advanced AMD is associated with photoreceptor loss in the macula that is preceded by underlying retinal pigment epithelial (RPE) cell layer dysfunction and death. The retinal pigment epithelium is a single layer of cuboidal cells that performs diverse functions in the maintenance of retinal and photoreceptor health. ¹¹ Because of its proximity to a highly oxygenated blood supply and direct exposure to light, the retinal pigment epithelium's microenvironment contains a high burden of reactive oxygen species (ROS). Exposure of RPE cells to elevated levels of ROS inhibits the ability of these cells to perform vital functions in the maintenance of retinal health ¹² and can lead to programmed cell death. ^13,14  

The synthetic triterpenoid derivatives of oleanolic acid, including 2-cyano-3,12-dioxooleana-1,9-dien-28-oic acid (CDDO), CDDO-imidazolide (CDDO-Im), CDDO-ethylamide (CDDO-EA), CDDO-trifluoroethylamide (CDDO-TFEA) (Fig. 1), are a promising new class of agents for cytoprotection from oxidative injury. ¹⁵ Synthetic triterpenoids act by activating the Nrf2 pathway to induce the transcription of phase 2 detoxifying genes and antioxidant genes, ^16,17 including NAD(P)H: quinone oxidoreductase (Nqo1), glutamate cysteine ligase, catalytic subunit (Gclc), and heme oxygenase 1 (Hmox1). Expression of these proteins then combats the harmful effects of oxidative injury by eliminating the oxidative insult and detoxifying the end products of oxidative injury. Targeting Nrf2 signaling has proven effective for cytoprotection in multiple models of retinal degeneration. ^18,19  

Phosphatidylinositol 3-kinase (PI3K)/AKT pathway signaling plays an important role in modulating Nrf2-signaling. This pathway has been most extensively characterized in regulating apoptosis and cellular proliferation. We demonstrated previously that CDDO-Im treatment leads to PI3K/AKT activation and that PI3K/AKT activity is necessary for CDDO-Im–mediated stabilization of Nrf2. Moreover, inhibition of PI3K/AKT signaling blocks CDDO-Im–mediated induction of Hmox 1 transcription. ²⁰ Although the importance of PI3K/AKT activity in triterpenoid-mediated activation of Nrf2 signaling has been documented, the mechanism(s) of PI3K/AKT activation have not yet been clarified. 

The aim of these studies was to examine whether triterpenoids (TPs) confer cytoprotection from oxidative and photooxidative retinal stress and to explore the molecular mechanisms of this cytoprotection. Because previous studies have implicated PI3K/AKT signaling in the activation of cytoprotective pathways by TPs, we explored a possible mechanism of TP-induction of AKT phosphorylation, a marker of PI3K/AKT activity. Collectively, these studies demonstrate the efficacy of synthetic TPs in cytoprotection from oxidative stress in vitro and photooxidative retinal damage in vivo and identify PTEN inhibition as a mechanism of TP-mediated activation of PI3K/AKT signaling. 

Materials and Methods

Reagents and Plasmids

The TPs CDDO, CDDO-Im, CDDO-EA, CDDO-TFEA, CDDO-Bt, and other TP derivatives (see Fig. 1 for structures) have been described previously. ^20–22 Stock solutions were made in dimethyl sulfoxide (DMSO; 0.01 M), and aliquots were frozen at −20°C. Sources of reagents were as follows: anti–hmox 1, anti–AKT, and anti–phospho-AKT antibodies from Cell Signaling Technology (Beverly, MA); anti–HA, anti–β-actin, and anti–α-tubulin antibodies from Santa Cruz Biotechnology (Santa Cruz, CA); human recombinant PTEN from R&D Systems (Minneapolis, MN); 2′, 7′-dichlorofluorescein diacetate (H₂DCFDA) from Molecular Probes (Eugene, OR); tris(2-carboxyethyl)phosphine (TCEP) from Thermo Fisher Scientific Inc. (Rockford, IL); high-capacity beads (DynaBeads MyOne Streptavidin T1) from Invitrogen (Carlsbad, CA). LY294002, n-ethylmaleimide (NEM), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), dithiothreitol (DTT), and all other compounds were from Sigma-Aldrich Chemical Co. (St. Louis, MO). The Nqo1-ARE-luciferase reporter plasmid was described previously. ²³ The HA-PTEN and HA-PTEN-C124S plasmids were described previously ²⁴ and obtained from Addgene (Cambridge, MA). 

Figure 1.

 

Structures of oleanolic acid, CDDO, and CDDO analogues.

Cell Culture

ARPE-19 retinal epithelial cells (American Type Culture Collection, Manassas, VA) were maintained in DMEM/F12 medium supplemented with 10% fetal bovine serum (FBS), 15 mM HEPES, and 2.5 mM sodium pyruvate. The 661W photoreceptor cell line was generously provided by Muayyad Al-Ubaidi (Department of Cell Biology, University of Oklahoma Health Sciences Center, Oklahoma City, OK) and was maintained in DMEM supplemented with 10% FBS. All tissue culture media, sera, and supplements were from Invitrogen except for sodium pyruvate (Sigma-Aldrich Chemical Co.). 

Cell Treatment and Viability Assay

ARPE-19 cells were plated in 24-well plates (40,000/well) or in 96-well plates (8000/well) and allowed to incubate overnight. Medium was discarded and replaced with medium containing dilutions of CDDO-Im, CDDO-TFEA, or CDDO-EA. After 1.5 to 24 hours, cells were treated with tert-butyl hydroperoxide (tBHP) overnight, and MTT viability assays were performed as described previously. ²⁵ Viability is expressed as “fractional survival” based on the ratio of the absorbance at 595-nm-treated cells compared with controls. 

Reporter Assays

ARPE-19 cells were plated in 24-well plates (40,000/well) and transfected with equimolar concentrations of Nqo1-ARE-luc and the CMX-β-gal expression vector (Fugene 6; Roche Applied Science). Twelve hours after transfection, cells were treated with TPs for an additional 24 hours. Cells were then lysed in 100 μL reporter lysis buffer (Promega, Madison, WI), luciferase activity was measured, and activity was normalized to β-galactosidase (β-gal) activity. The levels of transfected DNA were kept constant in all transfections, and all experiments were repeated at least three times. 

Detection of Reactive Oxygen Species

Cells were treated with CDDO-Im for 24 hours and incubated with 10 μM H₂DCFDA for 30 minutes. Cells were then challenged with 750 μM tBHP for 15 minutes, and mean fluorescence intensity of 10,000 cells was analyzed by flow cytometry using a 480-nm excitation wavelength and a 525-nm emission wavelength. Experiments were repeated three times in triplicate, and results of a representative experiment are shown. 

Gene Expression Analysis

Total RNA was isolated from ARPE-19 cells and retinal tissue using reagent (Trizol; Invitrogen) according to the manufacturer's instructions, and cDNA was synthesized using M-MLV reverse transcriptase (Invitrogen). Gene expression analysis for each sample was performed (iQ SYBR Green Supermix; Bio-Rad Laboratories, Hercules, CA). The following primers were used: mouse nqo1 (forward 5′-aatgggccagtacaatcagg-3′ and reverse 5′-ccagccctaaggatctctcc-3′), mouse glcl (forward 5′-gtctcaagaacatcgcctcc-3′ and reverse 5′-ctgcacatctaccacgcagt-3′), and human nqo1 (forward 5′-gcatagaggtccgactccac-3′ and reverse 5′-ggactgcaccagagccat-3′). Gene expression measurements were normalized to the endogenous reference gene β-actin (forward 5′-ccttgcacatgccggag-3′ and reverse 5′-gcacagagcctcgcctt-3′). 

Isolation of PTEN-Bound CDDO-Bt and Western Blot Analysis

For the CDDO-Bt precipitation experiments, untransfected ARPE-19 cells or cells transfected with either HA-PTEN or HA-PTEN-C124S were treated with 1 μM biotinylated TP for 30 minutes and lysed in 100 mM Tris-HCl, pH 7.4, 1% Triton X-100, 1 mM phenylmethylsulfonyl fluoride, 10 μM leupeptin, and 5 μg/mL aprotinin. Total protein (0.5 mg) was incubated with 30 μL high-capacity beads (DynaBeads MyOne Streptavidin T1; Invitrogen) for 1 hour, pelleted, and washed four times with 1× PBS, 0.4% Tween 20. Samples were resuspended in 40 μL Laemmli loading buffer, boiled for 5 minutes to remove bound proteins from the beads, and subjected to Western blot analysis. The in vitro interaction of PTEN and CDDO- Bt was characterized using full-length recombinant PTEN (R&D Systems). 

In Vivo Experiments

Animal experimentation was conducted in accordance with protocols approved by the Institutional Animal Care and Use Committee of Dartmouth Medical School and according to the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Light intensity within the cages in our colony room was 60 to 100 lux, and mice were kept in a 12-hour (7 AM-7 PM) light-dark cycle. For pharmacodynamic studies, 6-week-old female BALB/c mice (Jackson Laboratory) were fed powdered rodent chow (5002; PMI Feeds, St. Louis, MO) or powdered diet containing CDDO-TFEA (200 mg/kg diet) for 7 days. Eyes were enucleated, retinas were isolated, and retinal RNA was isolated. Gene expression analysis was performed in triplicate (n = 4 for CDDO-TFEA–treated group, and n = 4 for vehicle group). The procedure for exposure to light was described previously. ¹⁸ All light exposure started at 10 AM. Five- to 6-week-old BALB/c mice were fed control or diet containing CDDO-TFEA for 7 days. After 2 days of treatment, mice were dark adapted for 16 hours, pupils were dilated with 1% atropine eyedrops 30 minutes before exposure to light, and mice were exposed to 8000 lux of white fluorescent light for 1.5 hours in cages with reflective interiors. Mice were then euthanatized, and eyes were enucleated, prepared for histology as described previously, ¹⁹ and stained with hematoxylin-eosin. Two sections from each eye were analyzed. Digitized images were obtained, and the ONL thickness of eight defined locations—250 μm, 500 μm, 750 μm, and 1000 μM superior and inferior from the optic nerve head—were measured. These measurements were then averaged. 

In Vitro PTEN Activity Assay

Active, purified, full-length PTEN (R&D Systems) was diluted in enzyme buffer (25 mM Tris, pH 8.0, 100 mM NaCl, 10 mM TCEP, 0.03% BRIJ35, 2 mg/mL BSA) and coincubated with either vehicle or TP for 5 minutes at 37°C before the addition of phosphatidylinositol-3,4,5-triphosphate (PIP3; Cayman Chemical, Ann Arbor, MI). The final phosphatase reaction consisted of 15 ng PTEN, 60 μM PIP3 substrate, and various concentrations of TP or DMSO control. After 8-minute incubation at 37°C, the reaction was terminated by the addition of a phosphate detection reagent (Biomol Green; BioMol Research Laboratories, Inc., Plymouth Meeting, PA). After incubation for 20 minutes at room temperature, absorbance at 595 nm was determined. Phosphatase activity is expressed as “fractional activity” based on the ratio of the absorbance at 595 nm of treated reactions compared with the controls. 

Statistical Analysis

Results are mean ± SD. Probabilities were calculated by unpaired t-test. Statistical significance was set at P < 0.05. 

Results

CDDO-Im, -TFEA, and -EA Activate Nrf2 Signaling

We first examined the ability of the TP derivatives CDDO-Im, -TFEA, and -EA to activate Nrf2 signaling in ARPE-19 cells using the Nqo1-ARE-luc reporter plasmid. Treatment with CDDO-Im, -TFEA, or -EA led to dose-dependent increases in reporter activity, with CDDO-Im being the most potent inducer, followed by CDDO-TFEA and then -EA (Fig. 2A). To offer a more quantitative analysis of the transcriptional activation of Nrf2-regulated genes in ARPE-19 cells by CDDO-Im and -TFEA, real-time reverse transcription PCR was used to measure the relative transcript levels of Nqo1. These TPs increased levels of Nqo1 transcription in ARPE-19 cells by 10-fold (Fig. 2B). We have shown previously that CDDO-Im was a potent inducer of the Nrf2-regulated gene Hmox 1 in vitro and in vivo. ²⁰ Here we show that ARPE-19 treatment with 100 nM CDDO-Im, -TFEA, or -EA increased the expression of hmox 1 protein as early as 4 hours after treatment (Fig. 2C). ARPE-19 cells were then treated with 1 nM, 10 nM, or 100 nM dosages for 8 hours, and hmox 1 protein levels were measured (Supplementary Fig. S1). Once again these TPs were potent inducers of hmox 1 expression. Induction of hmox 1 was seen after treatment with as little as 10 nM CDDO-Im, whereas 100 nm of -TFEA and -EA were required to detect hmox 1 induction. 

Figure 2.

 

TP treatment induces Hmox 1 expression and Nrf2 activity in ARPE-19 cells. (A) ARPE-19 cells were treated with dilutions of CDDO-Im, -EA, or -TFEA for 24 hours, and Nrf2 activity was measured by transfection of the Nqo1-ARE-luc reporter plasmid (n = 3). (B) Cells were treated with 100 nM CDDO-Im or CDDO-TFEA for 24 to 48 hours. Nqo1 transcript levels were measured by real-time PCR and are expressed as fold-induction of vehicle-treated cells (n = 3). (C) Cells were treated with 100 nM of CDDO-Im, -TFEA, or -EA for 0, 1, 2, 4, 8, or 24 hours, hmox1 protein levels were determined by Western blot analysis, and α-tubulin levels were measured as a loading control (n = 3). (D) Cells were treated with CDDO-Im for 24 hours. H₂DCFDA was added for 15 minutes, and the cells were challenged with 750 μM tBHP for 15 minutes to induce ROS. Mean fluorescence intensity of 10,000 cells was detected by flow cytometry. CDDO-Im–pretreated cells had significantly reduced (*P < 0.05) ROS formation compared with untreated control (n = 3).

Activation of Nrf2 signaling induces the expression of antioxidant genes, thus reducing the oxidative burden of stressed cells. To test the functional significance of CDDO-Im—mediated Nrf2 activation in ARPE-19 cells, we measured the ability of CDDO-Im pretreatment to reduce ROS levels in tBHP-exposed cells. Cells were pretreated with various concentrations of CDDO-Im for 24 hours before exposure to H₂DCFDA and tBHP challenge. On oxidation, H₂DCFDA is converted to the fluorescent compound 2′,7′-dichlorofluorescein which can be detected by flow cytometry. Pretreatment with CDDO-Im decreased ROS levels in a dose-dependent manner. Maximum inhibition of 70% was seen in cells pretreated with 100 nm CDDO-Im (Fig. 2D). 

Triterpenoids Protect against Oxidative Stress–Induced Cell Death

Previous studies demonstrated that activation of Nrf2 signaling provides cytoprotection against photooxidative and oxidative stress. ^16,20 The oxidants hydrogen peroxide and tBHP are membrane-permeable agents that cause oxidative damage to proteins, depolarization of mitochondrial membrane potential, and cell death. These oxidizing agents have been widely used to assess the cytoprotective potential of antioxidants in a variety of cell lines. Here, we sought to evaluate the cytoprotective potential of synthetic TP derivatives in ARPE-19 cells and in the 661W photoreceptor cell line. ARPE-19 cells were pretreated with dilutions of CDDO-Im for 24 hours before exposure to tBHP for 12 hours and then cell viability was determined by MTT assay. CDDO-Im pretreated cells had a significant survival advantage over the control cells (Fig. 3A). Significant cytoprotection was observed after pretreatment with as little at 100 fM CDDO-Im before exposure to 162.5 μmol/L tBHP (Supplementary Fig. S2A). Cytoprotection was observed with 2 hours of CDDO-Im pretreatment before tBHP exposure (Supplementary Fig. S2B). Cells were then pretreated with CDDO-Im, CDDO-TFEA, or CDDO-EA before tBHP exposure (Fig. 3B), and IC₅₀ values were determined. The IC₅₀ values for CDDO-Im, -EA, and -TFEA were 9 nM, 190 nM, and 90 nM, respectively. Pretreatment of 661W cells with CDDO-Im or TFEA offered cytoprotection from tBHP-induced cell death as well (Fig. 3C). 

Figure 3.

 

TP pretreatment protects retinal cells against tBHP-induced cell death in vitro. (A) ARPE-19 cells were treated with dilutions of CDDO-Im for 24 hours before exposure to three concentrations of tBHP. Cell survival was determined by MTT assay and is expressed as “fractional survival”—the ratio of absorbance at 595 nm of treated cells compared with controls (n = 3). (B) ARPE-19 cells were treated with dilutions of CDDO-Im, -EA, or -TFEA for 24 hours before exposure to tBHP (325 μM) overnight (n = 3). (C) 661W cells were treated with dilutions of CDDO-Im or -TFEA for 24 hours before exposure to tBHP (2 mM) overnight (n = 3).

In Vivo Cytoprotection by CDDO-TFEA

We next extended these studies to an in vivo model. Oral administration of CDDO-Im, -EA, or -TFEA induced Nqo1 transcripts in multiple mouse tissues, including liver, small intestine, and cerebral cortex ²² ; however, TP effects on mouse retina had not been explored. Preliminary studies had shown higher levels of CDDO-TFEA in eye tissue compared with CDDO-Im or -EA after oral gavage of these compounds. We therefore used CDDO-TFEA for in vivo experiments rather than -Im or -EA. Given that cytoprotective activity requires Nrf2 activation and not necessarily the presence of the TP, we next performed pharmacodynamic studies. These studies showed that neural retinal levels of the Nrf2 regulated transcripts Nqo1 and Gclc were induced in CDDO-TFEA–fed BALB/c mice (200 mg/kg) compared with vehicle-fed controls (Fig. 4A). We then tested whether CDDO-TFEA had in vivo, retinal cytoprotective potency. Light exposure of BALB/c mice is an established model of in vivo photooxidative damage that has been used previously to evaluate potential retinal protective therapies. ^18,26,27 The toxic effects of light exposure can be quantified histologically by measuring the thickness of the of photoreceptor outer nuclear layer (ONL). In this model, pretreatment of BALB/c mice with CDDO-TFEA reduced light-induced retinal damage. Ninety-six hours after light exposure, the average ONL thickness at defined locations of CDDO-TFEA–treated mice (200 mg/kg) was significantly higher than that of control mice (12.9 ± 1.33 μm in control mice vs. 21.9 ± 6.76 μm in CDDO-TFEA–treated mice) (Fig. 4B). Taken together, these results show that treatment with CDDO-TFEA activates retinal Nrf2 signaling and attenuates light-induced retinal damage in mice. 

Figure 4.

 

CDDO-TFEA treatment induces Nrf2-regulated genes and attenuates light-induced retinal degeneration in BALB/c mice. (A) BALB/c mice (n = 4) were fed control diet or diet containing 200 mg/kg CDDO-TFEA for 7 days. Total retinal RNA was isolated, and RNA levels were determined by real-time PCR. CDDO-TFEA–treated mice had significantly increased (*P < 0.05) Nqo1 and Glcl transcripts compared with untreated control. (B) Quantification of ONL thickness in CDDO-TFEA–treated and untreated control mice after light-induced retinal damage. CDDO-TFEA–treated mice had significantly increased (*P < 0.05) ONL thickness compared with untreated controls.

PI3K/AKT Signaling Is Necessary for CDDO-Im–Mediated AKT Activation

We have shown previously that CDDO-Im treatment of leukemic U937 cells induced transient phosphorylation of AKT and that inhibition of PI3K/AKT signaling blocked CDDO-Im–mediated induction of hmox 1 transcription, ²⁰ thus highlighting communication between the AKT and Nrf2 signaling pathways. These observations were extended to the ARPE-19 cell line. Treatment of ARPE-19 cells with CDDO-Im induced AKT phosphorylation, with peak levels of phosphorylated AKT accumulating 30 minutes after treatment (Fig. 5A). Inhibition of PI3K/AKT signaling with the PI3K inhibitor LY294002 blocked CDDO-Im induction of AKT phosphorylation (Fig. 5B) and hmox 1 expression (Supplementary Fig. S3A). To explore the functional significance of AKT signaling on TP-mediated cytoprotection, ARPE-19 cells were exposed to LY294002 before CDDO-Im pretreatment and tBHP exposure. Interestingly, treatment of ARPE-19 cells with LY294002 before CDDO-Im exposure inhibited the cytoprotective effects of CDDO-Im (Supplementary Fig. S3B). In the absence of LY294002 pretreatment, exposure to CDDO-Im led to a 34% ± 5% increase in cellular survival over baseline survival after tBHP treatment. After LY294002 pretreatment, there was only a 10% ± 5% increase in cellular survival over baseline. Importantly, under these treatment conditions, baseline cell survival was not affected by LY294002 treatment. These results suggest that intact AKT signaling is required for TP-mediated cytoprotection. 

Figure 5.

 

TP treatment induces AKT phosphorylation in a PI3K-dependent manner. (A) ARPE-19 cells were treated with 100 nm CDDO-Im for 0, 15, 30, 60, or 120 minutes (n = 3). (B) Cells were treated with 15 μM LY294002 or vehicle control for 1 hour before treatment with 100 nm CDDO-IM for 30 minutes (n = 3). Whole cell lysates were prepared. pAKT and total AKT levels were determined by Western blot analysis. Long exposure (LE) and short exposure (SE) of the same pAKT blot were included to demonstrate both the induction of pAKT by CDDO-Im (SE) and the failure of CDDO-Im to induce pAKT after LY294002 pretreatment (LE).

CDDO-Bt Interacts with Cysteine 124 of PTEN

We next sought to determine the molecular mechanism of TP-mediated PI3K/AKT activation. Regulation of AKT phosphorylation occurs through the formation and degradation of the second messenger phosphatidylinositol (3,4,5)-triphosphate (PIP3), which are regulated by PI3K and PTEN, respectively. ^3–5 Synthetic TP potency depends on the presence of activated Michael reaction (enone) functions ¹⁶ that facilitate interactions with reactive cysteines within target proteins such as KEAP1 and IκB kinase β (IKK). Protein tyrosine phosphatases such as PTEN contain low pKa cysteine residues within their active sites that are potential targets for Michael addition by TPs. It is well established that loss of PTEN lipid phosphatase activity leads to accumulation of PIP3 and activation of PI3K/AKT signaling. ²⁸ Therefore, we hypothesized that TPs could activate PI3K/AKT signaling through inhibition of PTEN phosphatase activity. 

To determine whether TPs interact with PTEN, we used a biotinylated analogue of CDDO, CDDO-Bt (Fig. 1). Although the TP portion of the molecule maintains growth inhibitory activity similar to that of synthetic TP derivatives, the biotin functional group of CDDO-Bt can be selectively precipitated with avidin-coated beads to identify TP-interacting proteins. This analogue binds to a critical cysteine residue (Cys179) within the activation loop of IKK. ²¹ To determine whether CDDO-Bt had cytoprotective properties, we treated cells with CDDO-Bt and measured Nqo1-ARE-luc activity and CDDO-Bt-mediated cytoprotection (Supplementary Fig. S4A). The concentrations required to induce Nqo1-ARE-luc activity and to confer cytoprotection of ARPE-19 cells from tBHP-induced cell death were higher than those required for CDDO-Im (Supplementary Fig. S4B). However, these observations correlate with the observed differences in potency between CDDO-Im and CDDO-Bt in inhibition of cellular proliferation and inhibition of IκBα degradation. To determine whether CDDO-Bt binds to PTEN, lysates from cells treated with CDDO-Bt were precipitated with avidin-coated high-capacity beads (DynaBeads; Invitrogen) before Western blot analysis with anti–PTEN antibodies. We found that CDDO-Bt binds to endogenous PTEN and, importantly, that PTEN was not observed in precipitates from cells treated with DMSO (Fig. 6A). In addition, in vitro binding of purified, recombinant, full-length PTEN was inhibited by pretreatment with NEM, suggesting that reactive cysteine residues are required for CDDO-Bt binding to PTEN (Supplementary Fig. S4C). 

Figure 6.

 

CDDO-Bt interacts with cysteine 124 of PTEN. (A) ARPE-19 cells were treated with CDDO-Bt for 1 hour. CDDO-Bt-protein complexes were precipitated from cell lysates with high-capacity beads, and PTEN was detected by Western blot analysis (n = 4). (B) ARPE-19 were transfected transiently with HA-PTEN or HA-PTEN-C124S before treatment with CDDO-Bt and precipitation of CDDO-Bt-protein complexes. The presence of PTEN in total protein lysates and after CDDO-Bt pull-down (PD) was detected by Western blot analysis with anti–HA or PTEN antibody (n = 4).

PTEN contains a number of reactive cysteines including an active-site cysteine (C124), ²⁴ cysteine 71, which forms a disulfide bond with cysteine 124 on oxidation, ²⁹ and cysteine 212, which forms a disulfide bond with thioredoxin. ³⁰ To test whether CDDO-Bt was binding to the C124 of PTEN, we performed CDDO-Bt precipitation on lysates from cells transfected with wild-type PTEN (HA-PTEN) or PTEN with the C124 mutated to serine (HA-PTEN-C124S) and then probed with anti–HA antibody by Western blot. We found that CDDO-Bt binding to HA-PTEN-C124S was significantly diminished compared with binding to wild-type PTEN (Fig. 6B). 

Triterpenoids Inhibit In Vitro PTEN Phosphatase Activity

To further explore whether TPs inhibit PTEN lipid phosphatase activity, we performed an in vitro PTEN-lipid phosphatase activity assay. In this assay, recombinant, full-length, human PTEN is incubated with PIP3 substrate. PTEN lipid phosphatase activity converts PIP3 to phosphatidylinositol(4,5)-bisphosphate (PIP2) and free phosphate. Free phosphate can then be measured as a marker of PTEN activity. We found that preincubation of PTEN with CDDO-Im inhibited the lipid phosphatase activity of PTEN in a dose-dependent manner (Fig. 7A). Free thiols can interact with CDDO-Im by Michael addition, limit its bioavailability, and, therefore, limit its activity. ²¹ We hypothesized that preincubation of CDDO-Im with DTT would reduce CDDO-Im–mediated inhibition of PTEN activity. Preincubation of CDDO-Im with DTT did, in fact, diminish the inhibitory effect of CDDO-Im on PTEN activity in a dosage-dependent manner (Fig. 7B). Previous structure activity studies have identified key features of TP structure that confer antioxidant and anti-inflammatory activity. These features include introduction of the Michael addition, forming 1-en-3-one functionality on ring A and the addition of the electron-withdrawing nitrile group on C-2 of ring A. ¹⁶ Interestingly, these modifications enhanced TP inhibition of PTEN lipid phosphatase activity as well (Supplementary Fig. S5), suggesting that the structural features that enhance TP inhibition of PTEN activity parallel those that confer TP antioxidant and anti-inflammatory activity. 

Figure 7.

 

TP inhibition of PTEN lipid phosphatase activity. (A) PTEN and various dilutions of CDDO-Im were coincubated for 5 minutes before the addition of PIP3. Liberated phosphate was detected with a phosphate detection reagent, and absorbance was read at 595 nm (n = 4). PTEN activity was significantly reduced (*P < 0.05) with CDDO-Im treatment. (B) CDDO-Im and DTT were coincubated for 5 minutes before the addition of PTEN and PIP3 (n = 3).

Discussion

In this study, we demonstrated TP-mediated retinal cytoprotection from oxidative injury in vitro and light-induced injury in vivo. We show that CDDO-Im, -EA, and -TFEA are potent activators of Nrf2 signaling in vitro and that oral CDDO-TFEA treatment induced the expression of phase 2 enzymes in mouse retinas. CDDO-Im–mediated cytoprotection of retinal cells required PI3K/AKT signaling, and the lipid phosphatase activity of PTEN, a negative regulator of PI3K/AKT signaling, was inhibited by TP treatment. These results demonstrate the cytoprotection of retinal cells by TPs and suggest that this multifunctional class of drugs targets multiple pathways in the cytoprotection of cells. 

Activation of Nrf2 signaling and the consequent induction of phase 2 enzymes have emerged as a pivotal pathway for protection of tissues against endogenous and exogenous stress. There is growing interest in pharmacologic manipulation of this pathway to prevent pathologic tissue injury associated with chronic and acute cellular stress. Activation of Nrf2 signaling by various pharmacologic agents including sulforaphane, ³¹ curcumin, ^32,33 avicins, ³⁴ carnosol, ³⁵ resveratrol, ³⁶ and retinoic acid, ³⁷ protects retinal and other tissues from oxidative and cytotoxic injury. We have previously shown that TPs are potent inducers of Nrf2 signaling ^16,20 and that they provide in vivo protection from oxidative injury. ¹⁷ Here, we show for the first time that CDDO-TFEA treatment leads to in vivo induction of phase 2 enzymes in the retina and protects retinal photoreceptors from light-induced retinal damage. We then build on previous findings that implicate PI3K/AKT signaling in TP-mediated cytoprotection and Nrf2 signaling by demonstrating that PI3K/AKT signaling is necessary for CDDO-Im–mediated cytoprotection. In addition, we show that TPs target the lipid phosphatase activity of PTEN, a key negative regulator PI3K/AKT signaling, thus highlighting a mechanism of TP-mediated activation of PI3K/AKT signaling. 

Modulation of PTEN activity plays a pivotal role in the regulation of PI3K/AKT signaling. ²⁸ It is well established that the inhibition of PTEN activity by mutation, gene knockout, or small inhibitory RNA-mediated knockdown enhances PI3K/AKT signaling. ³⁸ PTEN activity is regulated by diverse mechanisms. Oxidation of PTEN by physiologic as well as pharmacologic dosages of H₂O₂ leads to the formation an internal disulfide bond between cysteines 71 and 124, inhibition of PTEN activity, and enhanced PI3K/AKT signaling. ²⁹ The phase 2 enzyme Trx 1 inhibits PTEN activity through direct protein-protein interaction and causes enhanced PI3K/AKT signaling. ³⁰ Furthermore, a functional interaction between DJ-1, an enzyme associated with early-onset Parkinson's disease, and PTEN has been described. DJ-1 negatively regulates PTEN activity ³⁹ and alters PI3K/AKT signaling. Overexpression of DJ-1 causes enhanced PI3K/AKT signaling, and knockdown of DJ-1 diminishes PI3K/AKT signaling. Interestingly, DJ-1 function has been linked to stabilization of Nrf2, ⁴⁰ reduced levels of cellular oxidative injury, and resistance to various apoptotic stimuli in a PTEN-dependent fashion. ³⁹ Collectively, these studies show that inhibition of PTEN activity, whether through genetic or pharmacologic manipulation, or protein-protein interactions can lead to enhanced AKT signaling. We demonstrate in this report a TP binding to the active site cysteine (C124) of PTEN and inhibition of PTEN lipid phosphatase activity by submicromolar concentrations of CDDO-Im. In addition, we show that this inhibition requires the same Michael addition-forming functional groups that are essential in activation of Nrf2 signaling as well as TP-mediated anti-oxidant and anti-inflammatory activity. 

The results of this study highlight the pleiotropic and seemingly paradoxical actions of synthetic TPs. At low dosages TPs activate Nrf2 signaling and are cytoprotective and anti-inflammatory, but elevated dosages of TPs cause growth suppression and apoptosis. We have shown previously that treatment with 100 nm CDDO-Im protects cells from oxidative insult whereas treatment with >300 nm CDDO-Im leads to the formation of ROS. ²⁰ In this study, we demonstrate that PI3K/AKT signaling is necessary for TP-mediated cytoprotection and that TPs activate this signaling pathway in ARPE-19 cells. In contrast, previous studies have demonstrated that micromolar dosages of CDDO, CDDO-Im, and CDDO-methyl ester cause apoptosis and inhibit PI3K/AKT signaling. ^41–44 What underlies these seemingly paradoxical effects? One likely explanation stems from our knowledge of TP action. We and others have previously shown that TP activity requires Michael addition-forming functional groups and that TPs directly interact with STAT proteins, KEAP1, and IKK. ¹⁵ In this study, we add PTEN to the group of TP-interacting proteins. Preferential interaction of TPs with some cellular targets, such as KEAP1 and PTEN, at low concentrations could lead to the cytoprotective activity of TPs, whereas interactions with other targets at higher dosages could cause growth inhibition and apoptosis. Of particular concern in the long-term use of TP derivatives is the potential for chronic toxicity. Previous studies demonstrated tolerance of long-term in vivo TP treatment, ^45,46 and the potential for chronic ocular toxicity should be addressed. The identification of the network of TP-binding proteins, the discovery of relevant TP targets, and the elucidation of the molecular mechanisms of the pleiotropic effects of TP treatment are active areas of investigation in our laboratory. 

Supplementary Materials

Supplementary Figure S1 - (PDF) 

Supplementary Figure S2 - (PDF) 

Supplementary Figure S3 - (PDF) 

Supplementary Figure S4 - (PDF) 

Supplementary Figure S5 - (PDF) 

Footnotes

 Supported by grants from Reata Pharmaceuticals and the National Institutes of Health (Grant CA-78814).

Footnotes

 Disclosure: I. Pitha-Rowe, Reata Pharmaceuticals (F); K. Liby, Reata Pharmaceuticals (F); D. Royce, Reata Pharmaceuticals (F); M. Sporn, Reata Pharmaceuticals (F)

Footnotes

 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

The authors thank Mark Yore, Charlotte Williams, and Renee Risingsong for technical assistance and advice; and James T. Handa and Michael Zegans for helpful comments on this manuscript.