The Proteasome: A Target of Oxidative Damage in Cultured Human Retina Pigment Epithelial Cells

Abstract

purpose. Dysfunction of the ubiquitin-proteasome pathway (UPP) is associated with several age-related degenerative diseases. The objective of this study was to investigate the effect of oxidative stress on the UPP in cultured human retina pigment epithelial cells.

methods. To mimic physiological oxidative stress, ARPE-19 cells were exposed to continuously generated H₂O₂ or A2E-mediated photooxidation. Proteasome activity was monitored using fluorogenic peptides as substrates. The ubiquitin conjugation activity and activities of E1 and E2 were determined by the thiolester assays. Levels of ubiquitin and ubiquitin conjugates were determined by Western blotting.

results. Exposure of ARPE-19 cells to 40 to 50 μM H₂O₂ for 4 hours resulted in a 30% to 50% reduction in all three peptidase activities of the proteasome. Similarly, exposure of A2E-loaded ARPE-19 cells to blue light resulted in a 40% to 60% reduction in proteasome activity. Loading of A2E or exposure to blue light alone had little effect on proteasome activity. In contrast, exposure of ARPE-19 to low levels of H₂O₂ (10 μM) stimulated ubiquitin conjugation activity. Loading of A2E, with or without exposure to blue light, upregulated the levels of ubiquitin-activating enzyme and increased conjugation activity. Exposure to H₂O₂ or A2E-mediated photooxidation also resulted in a twofold to threefold increase in levels of endogenous ubiquitin conjugates.

conclusions. These data show that the proteasome in ARPE-19 is susceptible to oxidative inactivation, whereas activities of the ubiquitin-conjugating enzymes are more resistant to oxidative stress. Oxidative inactivation of the proteasome appears to be one of the mechanisms underlying stress-induced accumulation of ubiquitin conjugates in the cells.

The ubiquitin-proteasome pathway (UPP) is the major nonlysosomal proteolytic pathway within cells. ^{1 2 3} In this pathway, proteins destined for degradation must be conjugated with a ubiquitin chain to be recognized and degraded by a large protease complex called the proteasome. The proteasome complex consists of a 20S proteolytic core and typically two regulatory 19S caps. The 19S cap recruits ubiquitin conjugates to the proteasome, then it cleaves ubiquitin moieties from the substrate, unfolds the polypeptide, and feeds it through the narrow channel of the proteolytic chamber of the 20S core. ⁴ Ubiquitin conjugation, or ubiquitination, is a highly ordered process in which a ubiquitin-activating enzyme (E1) activates and transfers ubiquitin to a ubiquitin-conjugating enzyme (E2), which then acts in concert with one of a large family of ubiquitin protein ligases (E3) to transfer ubiquitin to a lysine residue on the target substrate. ^{2 5} In most cases, multiple ubiquitins are conjugated to the initial ubiquitin moiety to form polyubiquitin chains. A chain of at least four ubiquitin moieties is often required for substrate recognition by the 26S proteasome complex. ^{6 7 8}  

The UPP is an important protein quality control system ^{9 10 11} that selectively degrades mutant, misfolded, or damaged proteins. ^{12 13 14} Timely removal of abnormal or damaged proteins by the UPP is essential for the cells to withstand and recover from various environmental stresses. ^{15 16} However, the UPP itself is also a target of such stresses. All three classes of ubiquitination enzymes (E1, E2, E3) have a cysteine in their active sites; therefore, the activities of these enzymes are subject to redox regulation. ^{17 18} In addition, other types of modifications, such as S-nitrosylation, can inactivate these enzymes. ¹⁹ Reactive oxygen species and reactive lipid peroxidation products, such as 4-hydroxynonenal (HNE), also impair the proteasome. ^{10 20 21 22 23 24}  

Similar to other types of cells, retina pigment epithelial (RPE) cells have an active UPP. ^{16 17 18 25 26} In a cell-free system, the RPE cell supernatant is capable of degrading a variety of substrates for the UPP, such as histone 2A, oxidized RNase, transducin, and β-lactoglobulin. ^{25 26} In contrast to many other types of cells, RPE cells have limited levels of free ubiquitin. Thus, adding exogenous ubiquitin to the RPE supernatant generated significantly higher levels of ubiquitin conjugates and also enhanced proteasome-dependent proteolysis. ^{16 25 26}  

Chronic exposure to light and high metabolic rate constantly generate reactive oxygen species in the retina. In addition, A2E, a major fluorophore of lipofuscin, acts as a photosensitizer in RPE to generate intracellular reactive oxygen species. ^{27 28} Therefore, the RPE is under continuous oxidative stress. The RPE, including ARPE-19 cells, developed an active defense system to cope with oxidative stress. For example, ARPE-19 cells have higher antioxidant capacity and are more resistant to oxidative damage than other types of cells. ^{29 30 31}  

The presence of an active ubiquitin proteasome system, together with constant exposure to oxidative stress, makes RPE cells an ideal model in which to assess the effect of oxidative stress on components of the ubiquitin-proteasome pathway. Previous studies showed that mild oxidative stress increases levels of ubiquitin conjugates in many types of cells and tissues, ^{13 15 32 33 34 35 36 37 38 39 40} whereas extensive oxidative stress reduces the levels of ubiquitin conjugates. ^{17 18} We demonstrated that the decline in ubiquitin conjugates on extensive oxidative stress results from the reversible inhibition of ubiquitin conjugation enzymes. ^{17 18} However, the mechanism that underlies the increase in levels of ubiquitin conjugates in response to mild oxidative stress was not fully understood. To investigate the mechanism by which mild oxidative stress increases the levels of ubiquitin conjugates in cells, we compared the susceptibilities of different components of the UPP to oxidative stress using ARPE-19 cells as a model system. The data indicate that the proteasome is more susceptible to H₂O₂-induced or photooxidation-induced inactivation than the ubiquitin-conjugating enzymes. The preferential inactivation of the proteasome by oxidative stress indicates a mechanism for the commonly observed accumulation of endogenous ubiquitin conjugates in response to oxidative stress in various cell types and tissues. 

Materials and Methods

Materials

N-ethyl-maleimide was obtained from Aldrich Chemical Co. (Milwaukee, WI). 4-(2-Aminoethyl)-benzenesulfonylfluoride hydrochloride was obtained from Calbiochem-Novabiochem Corp (La Jolla, CA). HNE was purchased from Cayman Chemical (Ann Arbor, MI). ¹²⁵I (NaI) was purchased from PerkinElmer (Boston, MA). SDS-PAGE reagents were from Bio-Rad (Hercules, CA). Dulbecco modified Eagle medium (DMEM) was purchased from Mediatech, Inc. (Herndon, VA) or Invitrogen Corporation (Carlsbad, CA). Fetal calf serum, nonessential amino acid solution, and antibiotics for cell cultures were purchased from Invitrogen Corporation. Antibodies to ubiquitin and ubiquitin-activating enzymes were produced in rabbits in this laboratory, as previously described. ^{34 41} Rabbit polyclonal antibody to HNE-modified proteins was provided by Luke Szweda (Oklahoma Medical Research Foundation). ⁴² The antibody to dinitrophenyl was purchased from Dako Cytomation (Carpinteria, CA). Horseradish peroxidase (HRP)-conjugated anti–rabbit secondary antibody was from Jackson ImmunoResearch (West Grove, PA). A2E was synthesized as described. ⁴³ All other chemicals were obtained from Sigma and were of the highest purity available. 

Exposure to H₂O₂

Confluent ARPE-19 cells were incubated in serum, pyruvate, and phenol red-free DMEM containing 4.5 g/L glucose in the presence or absence of 15 or 40 mU/mL glucose oxidase. Levels of H₂O₂ in the medium were monitored by a colorimetric method. ¹³  

Exposure to A2E and Blue Light

ARPE-19 cells known to be devoid of endogenous lipofuscin ²⁷ were grown to confluence and cultured in DMEM with 10% heat-inactivated fetal calf serum and 0.1 mM nonessential amino acid solution with or without 10 μM A2E for 2 weeks. The medium with fresh A2E was changed twice a week. As we demonstrated in previous studies, ARPE-19 cells accumulate A2E in the lysosomal compartment in a dose- and time-dependent manner. ^{27 44} When incubated with 10 μM A2E for 2 weeks, ARPE-19 cells accumulate optimal levels of intracellular A2E (approximately 10 nmol/10⁵ cells) for photooxidative stress on blue light exposure. ^{27 45 46} For blue light exposure, cell cultures were transferred to PBS with calcium, magnesium, and glucose and were exposed to 430 nm light delivered from a tungsten halogen source (430 ± 20 nm; 15 minutes; 2.62 mW/cm²). The cells were then incubated for an additional 6 hours in DMEM with 1% fetal calf serum and were harvested by scraping on ice. Controls included cultures that had neither accumulated A2E nor been exposed to blue light, cell cultures that accumulated A2E only, or cells that were exposed to blue light only. All the control cells were treated in the same manner as the cells that were exposed to A2E and blue light. 

De Novo Ubiquitin Conjugation and Thiolester Assay

To determine the ability to form ubiquitin conjugates, cells were lysed in 50 mM Tris-HCl, pH 7.6, containing 1 mM dithiothreitol (DTT). Conjugation activity was determined using endogenous enzymes and substrates with exogenous ¹²⁵I-ubiquitin. Briefly, the assay was carried out in a final volume of 25 μL, containing (final concentrations) approximately 10 mg/mL cell supernatant, 50 mM Tris buffer, pH 7.6, 2 mM adenosine triphosphate, 1 mM DTT, 5 mM MgCl₂, and 4 μM ¹²⁵I-labeled ubiquitin. The mixture was incubated at 37°C for 20 minutes, and the reaction was stopped by the addition of 25 μL 2× SDS-PAGE loading buffer. To determine the thiolesters of E1 and E2, the reaction was stopped by the addition of 25 μL 2× SDS-PAGE loading buffer with or without 50 mM DTT. After boiling at 100°C for 3 minutes, aliquots of the mixture were resolved by SDS-PAGE. The de novo–formed ubiquitin conjugates and E1∼ubiquitin (E1∼Ub) or E2∼ubiquitin thiolesters were visualized by autoradiography. In contrast to ubiquitin conjugates, ubiquitin thiolesters were labile to reducing reagents, and bands that disappeared in the presence of 50 mM DTT were E1∼Ub or E2∼Ub thiolesters. The molecular weights of these E1∼Ub and E2∼Ub thiolesters were established using purified E1 and E2 as in our previous work. ^{34 41}  

Detection of Endogenous Ubiquitin Conjugates, E1, HNE-Modified Proteins, and Carbonyl-Containing Proteins

Levels of endogenous ubiquitin conjugates, E1, HNE-modified proteins, and carbonyl-containing proteins in cell lysates were determined by Western blotting, as described previously. ^{14 32 41} Briefly, proteins were resolved by SDS-PAGE on 12% gels and transferred to nitrocellulose. The blots were probed with antibodies to ubiquitin, E1, or HNE-modified proteins. To determine levels of carbonyl-containing proteins, the supernatants were first reacted with dinitrophenylhydrazine, as previously described, ¹³ and the levels of dinitrophenylhydrazine-derived proteins were determined by Western blotting using an antibody to dinitrophenyl. The specifically bound antibody was detected by chemiluminescence (Super Signal kit; Pierce, Rockford, IL) after incubation with HRP-conjugated anti–rabbit secondary antibodies. 

Proteasome Activity Assay

ARPE-19 cells were lysed in 25 mM Tris-HCl buffer, pH 7.6. All three peptidase activities of the proteasome were determined using fluorogenic peptides, as described. ⁴⁷ Succinyl-Leu-Leu-Val-Tyr-amidomethylcoumarin (LLVY-AMC) was used for the chymotrypsin-like activity, N-t-butyloxycarbonyl-Leu-Ser-Thr-Arg-amidomethylcoumarin (LSTR-AMC) was used for the trypsin-like activity, and benzyloxycarbonyl-Leu-Leu-Glu-amidomethylcoumarin (LLE-AMC) was used for the peptidylglutamyl-peptide hydrolase activity. The mixture, containing 20 μg cell supernatant in 25 mM Tris-HCl, pH 7.6, was incubated at 37°C with the appropriate concentrations of peptide substrate (LLVY-AMC at 25 μM, LLE-AMC and LSTR-AMC at 40 μM) in a buffer containing 50 mM Tris-HCl, pH 8, 100 mM NaCl, 5 mM EDTA, 1 mM EGTA, 3 mM NaN₃, and 0.04% 3-[(3-cholamidopropyl)dimethylammonio]-1-propanesulfonate. The final volume of the assay was 200 μL. Enzymatic kinetics was measured with a temperature-controlled microplate fluorometric reader. Excitation/emission wavelengths were 380/440 nm. Proteasome activity was defined as the peptidase activity in the cell extracts that was inhibited by 20 μM N-Cbz-Leu-Leu-leucinal (MG132), a potent proteasome inhibitor. ⁴⁸  

Statistical Analysis

Data are expressed as mean ± SD. The data were statistically compared using one-way ANOVA with Tukey post hoc test or two-way ANOVA with Tukey post hoc test (see data in 1 2 3 4 Fig. 5D ). All analyses were performed using statistical software (Systat, version 11; Systat Software, Inc., San Jose, CA). P < 0.05 was considered statistically significant. 

Results

Continuous Generation of H₂O₂ by Glucose/Glucose Oxidase in the Presence of ARPE-19 Cells

The high metabolic rate and high oxygen consumption of the retinal tissue continuously generate reactive oxygen species. In addition, ocular inflammation increases the production of H₂O₂ and other reactive oxygen species. As much as 50 μM H₂O₂ can be detected in the vitreous of human eyes. ^{49 50 51} To mimic the chronic oxidative stress, we used a glucose/glucose oxidase system to continuously generate relatively constant levels of H₂O₂ in the presence of ARPE-19 cells. Data in Figure 1indicate that RPE cells are able to scavenge H₂O₂ from the medium. In the absence of cells, the levels of H₂O₂ in the medium increased in a dose- and time- dependent manner on the addition of glucose oxidase (Fig. 1A) . In the presence of ARPE-19 cells, levels of H₂O₂ in the medium were lower than in the absence of cells. The H₂O₂ levels reached 20 μM by 1 hour after the addition of 15 mU/mL glucose oxidase but decreased to 3 μM by 3 hours (Fig 1A) , indicating that the cells increased their capability to scavenge H₂O₂ in adaptation to mild oxidative stress. In contrast, the addition of 40 mU/mL glucose oxidase maintained the levels of H₂O₂ in the medium at a relatively constant concentration of 40 to 50 μM (Fig. 1A) . To determine whether the release of catalase or other enzymes into the medium by ARPE-19 cells contributed to the decomposition of H₂O₂, we measured H₂O₂ accumulation in the RPE-conditioned medium. The medium that contained 40 mU glucose oxidase was incubated in the presence of ARPE-19 cells for 2 hours and then was incubated in the absence of ARPE-19 cells for another 4 hours. H₂O₂ levels were monitored during the period of incubation. We found that H₂O₂ accumulated in the RPE-conditioned medium at a rate comparable to that of the nonconditioned medium (Fig. 1A) . These data indicated that the H₂O₂ was mainly decomposed inside the cells. 

Photooxidation-Induced Protein Modification by HNE

In vivo, there was an age-dependent accumulation of lipofuscin, which in turn acted as a photosensitizer in RPE, leading to an increased production of reactive oxygen species in response to short-wavelength light. ^{52 53 54 55} To mimic in vivo photooxidation, ARPE-19 cells were loaded with A2E, a major fluorophore and a photo-sensitizing component of lipofuscin, and were subsequently exposed to blue light. In control ARPE-19 cells, two bands were recognized by the antibody to HNE-modified proteins (Fig. 1B) . Several new HNE-reactive bands were detected after A2E and blue light exposure (Fig. 1B , arrows). Most HNE-positive proteins were of high mass, which barely entered the resolving gel (Fig. 1B) . In contrast, exposure to A2E or blue light alone did not increase the levels of HNE-modified proteins in the cells (Fig. 1B) . These data are consistent with our previous work ⁴⁵ and confirmed that A2E is capable of triggering photooxidation in ARPE-19 cells. 

Accumulation of Ubiquitin Conjugates on Oxidative Stress

Previous work indicates that the level of endogenous ubiquitin conjugates is a sensitive marker of oxidative stress. ^{32 34 37 38 39 40 56 57} To study the effect of physiologically relevant oxidative stress on the UPP, we first determined the levels of endogenous ubiquitin conjugates in RPE cells. As reported in other cell types, most ubiquitin conjugates detected in ARPE-19 cells were of high mass (Fig. 2) . Levels of endogenous ubiquitin conjugates increased approximately 2.5-fold when the cells were exposed to an average of 10 μM H₂O₂ for 4 hours (Fig. 2A) . Exposure to 50 μM H₂O₂ increased the levels of endogenous ubiquitin conjugates by approximately 3.2-fold. A2E-mediated photooxidation also resulted in a 3.5-fold increase in levels of endogenous ubiquitin conjugates in the cells (Fig. 2B) . Exposure to blue light alone had little effect on levels of endogenous ubiquitin conjugates, whereas accumulation of A2E alone resulted in a 1.7-fold increase in levels of endogenous conjugates (Fig. 2B) . The dramatic accumulation of endogenous ubiquitin conjugates when the RPE was exposed to physiologically relevant levels of oxidative stress indicated that the ability of the cells to process ubiquitin-conjugates was impaired by oxidative stress. 

Stimulation of Ubiquitin Conjugating Activity by Physiologically Relevant Oxidative Stress

The steady state levels of endogenous ubiquitin conjugates are the net balance between the rate of ubiquitin conjugation and the rate of degradation of the conjugates by the proteasome or deubiquitination by isopeptidase. To search for the mechanism whereby physiologically relevant oxidative stress increases the levels of endogenous ubiquitin conjugates, we evaluated the effect of oxidative stress on ubiquitin-conjugating activity. As shown in Figure 3A(left), exposure to 10 μM H₂O₂ resulted in a 2.5-fold increase in ubiquitin conjugating activity (as indicated by de novo–formed ubiquitin conjugates). Exposure to higher levels of H₂O₂ (50 μM) also resulted in a 1.8-fold increase the conjugating activity. Thiolester assays showed that the level of E1∼Ub thiolester did not change after exposure to oxidative insult (Figs. 3A[right], 3D), indicating that E1 activity was not affected by these levels of oxidative stress. Consistent with multiple E2s in mammalian cells, thiolester assays detected at least four different active E2s in ARPE-19 cells. The levels of all these E2∼Ub thiolesters increased approximately 60% on exposure to 10 μM H₂O₂. When exposed to 50 μM H₂O₂ for 4 hours, the levels of these E2∼Ub thiolesters returned to the levels found in control cells. These results indicate that low (nontoxic), but not high (toxic), levels of oxidative stress stimulate some E2s of the ubiquitin-conjugating machinery. 

We further assessed the effect of A2E-mediated photooxidation on the ubiquitin-conjugating activity. We found that the accumulation of A2E, regardless of blue light exposure, enhanced the ability of cell lysates to form de novo ubiquitin conjugates (Figs. 3B 3E)by 1.8- to 2.6-fold. The enhanced ubiquitin-conjugating activity appeared to correlate with a 2.5-fold increase in the levels of E1∼Ub and E2∼Ub thiolesters (Figs. 3B[right], 3E). Western blotting assays showed that levels of E1, particularly E1A, ⁴¹ increased approximately eightfold on accumulation of A2E in the cells (Figs. 3C 3F) . Exposure to blue light alone had no detectable effect on the levels of E1 in the cells. These data indicated that the upregulation of the ubiquitin-conjugating system in response to mild oxidative stress may partially explain the accumulation of endogenous ubiquitin conjugates in RPE cells. 

Inactivation of the Proteasome by Physiologically Relevant Oxidative Stress

The upregulation of ubiquitin-conjugating activity in response to mild oxidative stress can only partially explain the oxidation-induced increase in levels of endogenous ubiquitin conjugates. To further explore the causes of the elevated levels of endogenous ubiquitin conjugates, we evaluated the effect of physiological oxidative stress on three peptidase activities of the proteasome. Whereas exposure to 10 μM H₂O₂ had no detectable effect on any peptidase activity of the proteasome, treatment with 50 μM H₂O₂ resulted in approximately 50% inhibition of chymotrypsin-like and trypsin-like activities. The peptidylglutamyl peptide hydrolase activity of the proteasome was inhibited by approximately 30% under these conditions (Fig. 4A) . The decrease in the peptidase activities of the proteasome was not caused by loss of cell viability because there was no significant loss of cell viability within this time (Fig. 4C) . To further test the effect of physiologically relevant oxidation on the activity of the proteasome, we evaluated the effect of A2E-mediated photooxidation on proteasome activity in ARPE-19 cells. We found that exposure to blue light or accumulation of A2E alone had little effect on proteasome activity (Fig. 4B) . In contrast, accumulation of A2E and exposure to blue light together resulted in a 40% to 60% decrease in the three peptidase activities of the proteasome. The chymotrypsin-like activity and trypsin-like activity of the proteasome were preferentially affected by photooxidation. The data indicate that, as in other cell types, the proteasome in RPE is susceptible to oxidative stress and can be inactivated by physiologically relevant oxidative stress. 

Effects of H₂O₂ HNE and Oxidized Proteins on Proteasome Activity in Supernatant of ARPE-19 Cells

H₂O₂- or photooxidation-induced inactivation of the proteasome may be mediated by the reactive oxygen species or by lipid peroxidation products, such as HNE. It has been shown that HNE can inhibit the proteasome. ^{22 58} Additionally, data in Figure 1Bshowed that cellular proteins were modified by HNE on A2E-mediated photooxidation. Oxidation-induced formation of protein aggregates could also inhibit the proteasome. It has been reported that protein aggregates inhibit the proteasome-mediated degradation of several typical substrates in intact cells ^{59 60} or in cell-free systems. ^{24 58 61} To determine the mechanism of oxidative inactivation of proteasome, we directly tested the effect of H₂O₂, HNE, and oxidized proteins on the chymotrypsin-like activity of the proteasome in the supernatants of ARPE-19 cells. We found that adding as much as 100 μM H₂O₂ to the supernatant had little effect on proteasome activity (Fig. 5A) . However, adding 500 μM or greater concentrations of H₂O₂ to the supernatant inhibited the proteasome in a dose-dependent manner (Fig. 5A) . Similarly, adding 500 μM or greater concentrations of HNE to the supernatant of ARPE-19 cells also inhibited the proteasome (Fig. 5B) . To determine the effect of oxidized proteins on proteasome activity, we prepared the proteasome-free supernatant of ARPE-19 cells by centrifugation at 100,000g for 5 hours. ^{16 62} A fraction of the proteasome-free supernatant was treated with 2 mM H₂O₂ at 37°C for 2 hours, followed by the addition of 10 U/mL catalase to eliminate residual H₂O₂. Another fraction of the proteasome-free supernatant was treated identically, but without H₂O₂. Protein oxidation was monitored by determining levels of protein carbonyls in the supernatant. ^{13 15} As shown in Figure 5C , treatment of the supernatant with 2 mM H₂O₂ resulted in a significant increase in protein carbonyls, an indicator of protein oxidation. We added the nonoxidized or oxidized proteasome-free fractions to the proteasome-enriched fraction of ARPE-19 cells and incubated at 37°C for 0, 0.5, 1, and 2 hours and then measured the chymotrypsin-like activity of the proteasome. We found that adding the nonoxidized proteasome-free fraction to the proteasome-enriched fraction had no effect on proteasome activity (Fig. 5D) . Adding oxidized proteasome-free fraction to the proteasome-enriched fraction stimulated the chymotrypsin-like activity of the proteasome by 20% to 40%, independently of time of incubation (Fig. 5D) . However, it remains unknown how the H₂O₂-treated proteasome-free fraction stimulated proteasome activity in the proteasome-enriched fraction. These data indicate that H₂O₂ and HNE, but not oxidized proteins, inhibit proteasome activity. 

Discussion

The UPP plays important roles in a vast number of cellular functions, including protein quality control and signal transduction. ³ A functional UPP is required for the cells to cope with various stresses, including heavy metals, ^{63 64} amino acid analogs, and oxidation. ^{15 16} However, an extensive oxidative insult is also likely to damage or impair the function of critical components of the UPP. ^{17 18 19 20 21 22 23 65} RPE cells in vivo are exposed to chronic oxidative stress from high oxygen consumption and lipofuscin-sensitized photooxidation. In this work we evaluated the effects of physiologically relevant oxidative stressors on the UPP in cultured human RPE cells. We found that the UPP can withstand 10 μM H₂O₂ without experiencing major functional damage. In fact, this level of H₂O₂ even stimulated the ubiquitin-conjugating activity (Fig. 3A) . However, prolonged exposure to 50 μM H₂O₂ inhibited all three peptidase activities of the proteasome. Furthermore, A2E-mediated photooxidation inhibited proteasome activity. However, exposure to the same oxidizing agents for the same dosage did not inhibit ubiquitin-conjugating activity. 

To determine the potential mechanism by which oxidative stress inactivates the proteasome, we directly tested the effect of H₂O₂ and HNE on the proteasome activity in RPE supernatants. We found that H₂O₂ and HNE are capable of inactivating the chymotrypsin-like activity of the proteasome (Fig. 5) . However, relatively high levels (>100 μM) of H₂O₂ and HNE were required to inactivate the proteasome in the cell supernatants. The resistance to a single bolus of H₂O₂ or HNE may be related to the rapid detoxification of H₂O₂ or HNE by the active antioxidant system in these cells, such as high levels of glutathione. Consistent with this idea, we also found that the RPE cells were resistant to a single bolus of H₂O₂. Exposure of ARPE-19 cells to a single bolus of 200 μM H₂O₂ for 1 hour only decreased proteasome activity by less than 20% (data not shown). Data in Figure 1also indicate that a significant fraction of H₂O₂ in the medium was decomposed in the presence of ARPE-19 cells. To test whether the proteasome in ARPE-19 cells could be inhibited by oxidized proteins, we oxidized the proteasome-free supernatant of ARPE-19 cells with 2 mM H₂O₂ for 2 hours and then tested its effects on proteasome activity in the proteasome-enriched fraction of ARPE-19 cells. We found that adding the H₂O₂-treated proteasome-free fraction stimulated, rather than inhibited, the chymotrypsin-like activity of the proteasome. These data suggest that reactive oxygen species or lipid peroxidation products such as HNE play a major role in the inactivation of the proteasome on physiologically relevant oxidative stress. 

Consistent with the observation that the proteasome in ARPE-19 cells is susceptible to oxidative stress, oxidative inactivation of the proteasome has been reported in many cell types ^{66 67} Modifications and inactivation of the proteasome by HNE were detected in other types of cells and tissues. ^{22 68} Oxidative inactivation of the proteasome was also observed on ischemia-reperfusion injury, ^{47 69} UVA or UVB radiation, ⁶¹ and chronic exposure to high concentrations of oxygen. ²¹ In addition to reactive oxygen species and lipid peroxidation products, protein aggregates, including oxidized or HNE-modified proteins, may also inhibit the proteasome directly or indirectly. ^{24 58 59 60 61 70} However, data presented here indicated that the proteasome was inhibited by reactive oxygen species or lipid peroxidation products, but not by oxidized proteins under these mild oxidative conditions (Fig. 5D) . However, our data do not rule out the possibility that proteins that are extensively oxidized will inhibit the proteasome. Compared with most other studies, ^{24 58 61} we used relatively mild and physiologically relevant oxidative conditions. 

Increased levels of endogenous ubiquitin conjugates have been reported in many types of cells and tissues in response to oxidative stress. ^{13 15 32 33 34 35 36 37 38 39 40} Because oxidized proteins are preferred substrates for the UPP, the increase in substrate availability may be one of the mechanisms for the accumulation of ubiquitin conjugates. ³⁴ However, this work indicates that another mechanism, oxidative inactivation of the proteasome, could also contribute to the accumulation of ubiquitin conjugates in response to oxidative stress. 

The UPP plays an important role in protein quality control. Many forms of damaged proteins, including oxidized proteins, are degraded by the proteasome. ^{10 13 15 71 72 73 74} Oxidative stress may not only increase the generation of oxidized cellular proteins, it may impair the machinery that degrades oxidized proteins. Consistent with an age-related decline in antioxidant capacity in many tissues, proteasome activity in many tissues, including retina, also decreases with aging. ^{75 76 77} The age-related decline in proteasome activity may be responsible for the accumulation of damaged proteins in various tissues, particularly for the postmitotic cells. ^{67 74 78 79 80} Given that the accumulation of oxidized or otherwise damaged proteins is associated with many age-related diseases, protecting the proteasome from oxidative inactivation may be a valid strategy to prevent age-related accumulation of damaged proteins and the onset or progress of age-related diseases. ^{10 81 82}  

In addition to protein quality control, the UPP plays important roles in signal transduction. Various transcription factors, such as hypoxia-inducible factor (HIF), ^{83 84} p53 ^{85 86} and STAT-1, ^{87 88} are the substrates of the UPP. The activity of NF-κB is also regulated by proteasome-mediated degradation of the inhibitor of NF-κB (I-κB). ^{89 90 91 92} Our previous work demonstrated that inhibition of the UPP in ARPE-19 cells altered the signal transduction cascade. ⁹³ We have shown that inhibition of the proteasome resulted in accumulation of HIF-1α and increased the expression and secretion of VEGF. ⁹³  

Age-related macular degeneration (AMD) is the leading cause of blindness in industrialized countries. Elevated levels of VEGF in the eye are associated with the development of AMD, particularly the wet form. ^{94 95} A growing body of literature demonstrates that oxidative stress is one of the risk factors for AMD. ⁹⁶ Previous work also showed that oxidative stress, including A2E-mediated photooxidation, stimulates the expression and secretion of VEGF by RPE cells. ^{45 97 98 99} Our previous work showed that inhibition of the proteasome also increased the expression and secretion of VEGF by ARPE-19 cells. ⁹³ This work demonstrates that the proteasome is a target of oxidative stress. However, it remains to be determined whether oxidative inactivation of the proteasome is a mechanistic link between oxidative stress and enhanced expression and secretion of VEGF. 

ARPE-19 is an adult human RPE cell line that displays many differentiated properties of the retinal pigment epithelium in vivo. For example, ARPE-19 cells express RPE-specific proteins, such as RPE65 and cellular retinaldehyde-binding. ^{100 101 102} Other RPE-specific features of ARPE-19 cells include morphologic and functional polarization when plated on laminin-coated filters, formation of tight junction, high transepithelial resistance, expression and secretion of VEGF, and pigment epithelium-derived factor. ^{100 103} Therefore, the ARPE-19 cell line is the most commonly used RPE cell line in eye research. Recent reports indicate that ARPE-19 and primary human fetal RPE cells have similar sensitivity to oxidants and other types of cellular stressors. ^{101 104} Compounds that protect ARPE-19 cells from oxidative damage also protect primary human fetal RPE cells. ^{101 104} We should emphasize, however, that ARPE-19 cells are not identical with RPE in vivo. Further studies to confirm the effects of oxidative stress on the UPP in RPE under in vivo or ex vivo conditions are warranted. 

Taken together, the data indicate that the proteasome is a target of oxidative stress and that inhibition of the proteasome may account, at least in part, for the accumulation of ubiquitin conjugates on experimental oxidative insult and under pathologic conditions. Given the important roles of the UPP in various cellular functions, oxidative inactivation of the UPP is likely to have physiological consequences. For example, oxidative inactivation of the proteasome in RPE could be one of the mechanistic connections between oxidative stress and the pathogenesis of AMD. Therefore, protecting the proteasome from oxidative inactivation may be a valid strategy for the prevention of age-related diseases, including AMD. 

 Supported by National Institutes of Health Grants EY11717 (FS), EY13250 (AT), and EY12951 (JRS); United States Department of Agriculture Grant CRIS 1950-51000-060-01A (AT); and Portuguese Foundation for Science and Technology Grant POCI/SAU-OBS/57772/2004 (PP). AF is a recipient of a Fellowship from the Portuguese Foundation for Science and Technology (SFRH/BD/19039/2004).

 Submitted for publication December 5, 2007; revised January 25 and March 19, 2008; accepted June 4, 2008.

 Disclosure: X. Zhang, None; J. Zhou, None; A.F. Fernandes, None; J.R. Sparrow, None; P. Pereira, None; A. Taylor, None; F. Shang, None

 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.

 

Figure 1.

 

Exposure of ARPE19 cells to H₂O₂- or A2E-mediated photooxidation. (A) Glucose oxidase was added to 10 mL serum-, pyruvate-, and phenol red-free DMEM in the presence or absence of a monolayer of confluent ARPE-19 cells and incubated at 37°C. Aliquots of the medium were taken at the indicated times, and levels of H₂O₂ in the medium were determined colorimetrically. (B) The ARPE-19 cells were treated with blue light alone, A2E alone, or with both A2E and blue light. Levels of HNE-modified proteins in the cells were determined by Western blotting with an antibody specific for HNE-modified proteins. Arrows: HNE-modified proteins detected only when the cells were exposed simultaneously to A2E and blue light.

Figure 2.

 

Oxidative stress results in an accumulation of ubiquitin conjugates. (A) ARPE-19 cells were exposed to 10 or 50 μM H₂O₂ for 4 hours, and levels of endogenous ubiquitin conjugates were determined by Western blotting. (B) ARPE-19 cells were treated with blue light alone, A2E alone, or both A2E and blue light, as described. Levels of endogenous ubiquitin conjugates were determined by Western blotting. Levels of β-actin were used as protein-loading controls. (C) Densitometry quantification of conjugates in (A) and (B). Bar graphs represent the mean ± SD of three independent experiments. Statistical comparisons were performed using one-way ANOVA with Tukey post hoc test. P > 0.05, bars with identical letters. P < 0.05, bars with dissimilar letters.

Figure 3.

 

Physiologically relevant oxidative stress stimulates ubiquitin-conjugating activity in ARPE-19 cells. (A) ARPE19 cells were exposed to 10 and 50 μM H₂O₂ for 4 hours. Ubiquitin-conjugating activity, E1 activity, and E2 activities were determined by thiolester assay. (B) ARPE-19 cells were treated with blue light alone, A2E alone, or A2E together with blue light. Ubiquitin conjugation activity, E1 activity, and E2 activity were determined by thiolester assay. Ubiquitin conjugates were stable, whereas E1∼Ub or E2∼Ub thiolesters were labile to reducing reagents. The bands that disappeared in the presence of DTT were E1∼Ub (110 kDa) or E2∼Ub (20–40 kDa) thiolesters. (C) ARPE-19 cells were treated with blue light alone, A2E alone, or A2E together with blue light, and levels of E1 were determined by Western blotting. To clearly visualize E1A and E1B in the A2E-treated cells, two differently exposed blots were presented. (D–F) Densitometry quantification of (A–C), respectively. Bar graphs represent mean ± SD of three independent experiments. Data were statistically compared using one-way ANOVA with Tukey post hoc test. P > 0.05, bars with identical letters. P < 0.05, bars with dissimilar letters.

Figure 4.

 

Physiologically relevant oxidative stress inactivates the proteasome in ARPE-19 cells. (A) ARPE-19 cells were exposed to a continuously generated H₂O₂ for 4 hours. (B) ARPE-19 cells were exposed to blue light alone, accumulated A2E alone, or accumulated A2E and exposed to blue light, as indicated in Figure 1 . (C) Cells were harvested, and proteasome activities were determined using fluorogenic peptides as substrates. ARPE-19 cells were exposed to continuously generated H₂O₂ for 4 hours, and cell viability was determined with MTS assay. Proteasome activity and viability of cells not treated with H₂O₂ were arbitrarily defined as 100%, and the rest were expressed as relative activities after normalization with controls. Data were mean ± SD of three independent experiments and were statistically compared using one-way ANOVA with Tukey post hoc test. *P < 0.05 and **P < 0.01 compared with control.

Figure 5.

 

H₂O₂ and HNE, but not oxidized proteins, inhibit the proteasome in the supernatant of ARPE-19 cells. ARPE-19 cells were homogenized in lysis buffer (50 mM sodium phosphate buffer, pH 7.4) and centrifuged at 30,000g for 20 minutes. Resultant supernatants (1 mg protein/mL) were incubated with the indicated concentration of H₂O₂ (A) or HNE (B) for 1 hour at 37°C. Chymotrypsin-like activity of the proteasome in the supernatants was measured as described in Figure 4 . A fraction of ARPE-19 cell supernatant was centrifuged at 100,000g for 5 hours. The resultant supernatant was devoid of proteasome and was defined as proteasome-free supernatant. The pellet was resuspended with the lysis buffer and was defined as the proteasome-enriched fraction. The proteasome-free fraction was treated with or without 2 mM H₂O₂ at 37°C for 2 hours, followed by incubation with 10 U/mL catalase for 30 minutes to eliminate any unreacted H₂O₂. Protein oxidation was monitored by determining levels of protein carbonyls in the supernatant (C). Nonoxidized and oxidized proteasome-free supernatants were added to the proteasome-enriched fraction and were incubated for 0, 0.5, 1, and 2 hours at 37°C, and the chymotrypsin-like activity of the proteasome was determined as described (D). (A, B) Data were mean ± SD of three independent experiments. (C, D) Data were mean ± SD of three experiments using the same batch of RPE supernatant. (A, B) Data were statistically compared using one-way ANOVA with Tukey post hoc test. (D) Data were statistically compared using two-way ANOVA with Tukey post hoc test. P > 0.05, bars with identical letters. P < 0.05, bars with dissimilar letters.

The authors thank Jerry Dallal for his help with the statistical analysis.