| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1From the Doheny Eye Institute, and the 2Department of Ophthalmology, University of Southern California, Keck School of Medicine, Los Angeles, California; and the 3Beckman Research Institute of the City of Hope, Duarte, California.
 |
Abstract |
|---|
 Top Abstract MethodsResultsDiscussionReferences |
|---|
METHODS. Tyrosine nitration in proteins was detected by UV/Vis (visible) absorption and Western blot analysis. The identity of the nitrated protein was obtained by liquid chromatography-tandem mass spectrometry. The release of cytochrome c was assessed in whole retinal extract and in isolated mitochondria. The protein nitration in the inflamed retina was also localized by immunohistochemistry.
RESULTS. Before the leukocyte infiltration in the early phase of EAU, the mitochondria-originated peroxynitrite initiated the inflammatory insult by specifically nitrating three mitochondrial proteins. In vitro nitration of the control retina by peroxynitrite donor resulted in nonspecific nitration of all major retinal proteins. After nitration, cytochrome c was displaced from its original binding site in the respiratory chain. Further, the nitration appeared to commence in the early phase of inflammation, on postimmunization day 5, long before the peak of inflammation on day 14. Immunohistochemically, tyrosine-nitrated proteins were localized exclusively in the photoreceptor inner segments, which are known to be densely populated with mitochondria.
CONCLUSIONS. These data indicate that mitochondrial proteins are the prime targets of inactivation by the mitochondrial peroxynitrite and that photoreceptor mitochondria initiate the subsequent irreversible retinal damage in experimental uveitis.
Although the initial events leading to uveitis in humans are not always clear, the eventual loss of vision has always been ascribed to the photoreceptor damage caused by amplification of the inflammatory processes. In EAU, the disease process starts with events such as local antigen presentation, homing of activated T-cells to the retina, and expression of adhesion molecules by vascular endothelium. However, the most direct cause of the retinal damage is the various cytotoxic agents and free radicals that are released by the infiltrating macrophages and polymorphonuclear leukocytes.1 2 3 4 5 6 In vitro and in vivo studies in EAU suggest that activated macrophages are the major producers of free radicals (Zhang J, et al. IOVS 1993;34:ARVO Abstract 1000).7 These reactive species can amplify the local inflammatory processes and cause photoreceptor cell damage. Superoxide (O2–) and nitric oxide (NO·) are among the most important primary species generated by the macrophages. We have demonstrated the simultaneous production of O2– and NO· in EAU. Nitric oxide production, in particular, has been shown to increase twofold in the inflamed retina compared with the control retina (Zhang J, et al. IOVS 1993;34:ARVO Abstract 1000).7 Others have shown that tissue damage in EAU correlates with peroxynitrite (ONOO–) formation in infiltrating monocytes/macrophages within the outer retina.8 Further, at the peak of inflammation (postimmunization day 14), the oxidative damage inflicted by these reactive species was concentrated in the photoreceptors, as indicated by the localization of hydroperoxide-derived cellular carbonyls,9 10 11 due to an unusually high concentration of docosahexaenoic acid (22:6) in the photoreceptor outer segments. Cellular protein modification by tyrosine (Tyr) nitration was also found to occur at the same time, mainly in the photoreceptor layer, with only minor lesions in the retinal blood vessels.9 10 11 Contrary to these earlier observations and the dogma that tissue damage is initiated by activated macrophages, the present study revealed that retinal nitration damage occurred earlier (on day 5 after S-antigen injection), before there was any histologic or immunohistochemical evidence of macrophage infiltration.
In addition to blood-borne inflammatory cells, retinal microglia have recently been shown to be capable of exhibiting phagocytic and pathogenic functions similar to those of macrophages, and the effect of the microglia manifests before the arrival of the macrophages. Between postimmunization days 10 and 11, retinal microglia are activated and release pathogenic factors, such as ONOO– and tumor necrosis factor-
(TNF-).12 Using fluorescent dye, 4Di-10ASP, and anti-rat CD11b antibody to localize microglia,12 we found that in EAU, the retinal protein nitration also occurred before activation and migration of retinal microglia. Therefore, it appears that there is an alternative mechanism of retinal damage in the development of EAU, and this mechanism appears to operate apart from the effects of macrophages and microglia, especially in the release of reactive nitrogen species (RNS) and reactive oxygen species (ROS). These early events that set the destructive pathway in EAU have not been elucidated.
At physiological pH, ONOO– formed in vivo can directly nitrate phenolic rings to form 3-nitrotyrosin (nitroTyr) from Tyr residues. The nitration is decreased by reductants such as ascorbic acid, ubiquinol, and uric acid, thus confirming the involvement of ONOO–.13 In recent years, although other metabolites of NO· have emerged as biological oxidants—for example, NO– can be converted to NO2+, which can serve as a nitrating agent14 —it is generally agreed that ONOO– is still the most plausible entity for effecting biological nitration and oxidation. Peroxynitrite has been implicated in the pathogenesis of a series of diseases, including acute and chronic inflammatory processes, sepsis, ischemia-reperfusion, and a variety of neurodegenerative and retinal disorders.15 16
Recently, the presence of nitric oxide synthase (NOS) has been shown in mammalian mitochondria.17 18 Mitochondrial (mt)NOS is located in the inner membrane and is constitutively expressed and modulated by Ca2+.17 Thus, with an abundance of substrate,18 NO· is produced in the mitochondria. Mitochondria are also a copious source of O2–, which is generated at the sites of complexes I and III of the electron transport chain.19 20 In tissues and in mitochondria, ONOO– is known to form from a facile reaction of O2– and NO· concomitantly generated in close proximity. Both O2– and NO· have low reactivity, and neither combines with tissue components to an appreciable extent.21 These facts are likely to suggest that mitochondria are being continuously challenged by ONOO– formed within the organelles themselves. In the past, nitration of Mn superoxide dismutase (SOD), a mitochondrial SOD, has been detected in animal and human tissues exposed to elevated fluxes of NO· during chronic and acute inflammation.22 23 In addition to MnSOD, other nitrated mitochondrial proteins have been detected in animals undergoing inflammatory processes; these proteins include mitochondrial aconitase, the voltage-dependent anion channel, mitochondrial adenosine triphosphatase (ATPase), and cytochrome c (cyto c).23 24
The photoreceptor cell layer, situated between the retinal pigment epithelium and the outer plexiform layer in adult mammalian retina, is specialized for absorption of light and initiation of the neuronal visual process. The photoreceptor outer segments concentrate on light–dark adaptation mechanisms, and the inner segments concentrate on energy production and protein synthesis.25 Photoreceptor cells are also known to have the highest rates of glycolysis and respiration among all retinal cells.26 For these reasons, inner segments of photoreceptor are densely packed with mitochondria.27 28
The present study is designed to determine the primary nitration target(s) of ONOO– and to detect the onset of this posttranslational modification in EAU. There are numerous proteins in the retina that complement its complex visual functions. Three of these proteins, all of which are essential for mitochondrial energetics and metabolism functions, were found to be selective prime targets of ONOO– nitration. Moreover, the protein nitration was found to commence early in the inflammatory process, long before the entry of inflammatory cells known to release O2– and NO· in the retina. In experimental uveitis, the initial insult that starts spiral degenerative processes in the photoreceptors has not been defined in the past.
|
Methods |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
Generation and Assays for In Vitro and In Vivo Nitrated Proteins
The kinetics of ONOO– generation from 3-morpholinosydnonimine (SIN-1; Sigma-Aldrich, St. Louis, MO) was determined using the previously reported method.29 Briefly, oxidation of dihydrorhodamine-123 (Molecular Probes, Eugene, OR) to rhodamine by 1 mM SIN-1 in 50 mM phosphate buffer (pH 7.4) was performed at 37°C, and production of rhodamine was monitored at 500 nm (
max = 78,000/M/cm). In vitro nitrated bovine serum albumin (BSA; Sigma-Aldrich) was used as a model protein to establish the chemical parameters for all assays of in vitro and in vivo nitrated retinal proteins. To prepare the nitrated BSA, 5.2 mg BSA was reacted with 1.3 mL 6.3 mM SIN-1 in 50 mM phosphate buffer (pH 7.6) at 37°C. The formation of nitroTyr was observed for up to 5 hours of incubation by spectrophotometric measurement and Western blot analysis. All UV/Vis (visible) absorption of nitroTyr was determined with a spectrophotometer (UV-160; Shimadzu, Kyoto, Japan), using a molar extinction coefficient of 4400/M/cm.30 Similarly, 3.2 mg rat cyto c (Sigma-Aldrich) was reacted with 0.84 mL of 50 mM phosphate buffer (pH 7.6), containing 7.0 mM SIN-1 for 6 hours at 37°C. At the end of incubation, the UV/Vis spectrum was measured to assess the extent of nitration. Authentic rat cyto c itself absorbs at 407 nm from the heme prosthetic group contained in the molecule. In vitro nitrated rat cyto c was used as a reference for in vivo nitrated retinal proteins.
Eighteen eyes were used for the nitration of retinal proteins in vitro. The experiments were performed in triplicate, and six retinas were combined as one experiment. The whole retinas were made permeable to SIN-1 by cutting them into pieces before exposing them to a dose of 8 mM SIN-1 in 1.5 mL of 50 mM phosphate buffer (pH 7.6), containing 1 mM phenylmethylsulfonyl fluoride (Sigma-Aldrich) and 0.2 U/mL aprotinin (Sigma-Aldrich) for 6 hours at 37°C. The cumulative dose for these experiments was 32 mM. Retinas from nine animals incubated without SIN-1 were used as the control. After incubation, the retinas were homogenized for 10 seconds and sonicated (30 seconds with a 5-second pulse on and 1-second pulse off; 70% duty cycle) using a digital sonifier (Branson Ultrasonics, Danbury, CT). The protein extracts were centrifuged at 14,000g for 1 hour before their use in the assays. Protein concentration was determined by protein assay (Bio-Rad Laboratories, Hercules, CA).
EAU was induced in 60 Lewis rats using bovine S-antigen. Twelve animals each were killed on postimmunization days 0, 5, 10, 12, and 14, with day 0 being the nonimmunized control and day 14, the peak of inflammation. In each period, experiments were performed in triplicate. Nine animals were used for UV detection and Western blot analysis, and three animals for hematoxylin and eosin (H&E) staining, to evaluate the inflammatory infiltrates and morphologic changes. In each period, six retinas were combined as one determination. To extract the proteins from inflamed tissue, the retinas were removed at the desired time point after immunization and were homogenized, sonicated, and centrifuged as were the in vitro nitrated retinal samples. Both in vitro and in vivo nitrated retinal proteins were assayed by UV/Vis absorption, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE), and Western blot analysis.
Electrophoretic and Western Blot Analyses
All unmodified and Tyr-nitrated proteins were separated by electrophoresis on SDS-PAGE (12% or 15%) and visualized with Coomassie blue. For Western blot analyses, the electrophoresed proteins were transferred to polyvinylidene fluoride (PVDF) membrane (Millipore, Bedford, MA). Nonspecific binding sites were blocked with 0.1% Tween-20 in 50 mM Tris-buffered saline (TTBS; pH 7.4) throughout the staining procedure. The PVDF membrane was then probed with preimmune serum or polyclonal anti-nitroTyr (1:200; Upstate Biotechnology, Lake Placid, NY) in TTBS for 2 hours. After washing, the membrane was incubated with biotin conjugated goat anti-rabbit IgG (1:5000; Dako, Carpinteria, CA). After enhancement with a complex of peroxidase-conjugated biotin and avidin (ABC kit; Vector, Burlingame, CA), visualization was performed with freshly prepared 3,3'-diaminobenzidine/NiCl2 reagent. For some runs, luminol-enhanced chemiluminescence (ECL; Amersham Pharmacia Biotech, Piscataway, NJ) was also used for visualization. The nitroTyr antibody was also preincubated with 10 mM 3-nitroTyr (Sigma-Aldrich), as a control for nitroTyr specificity. After this treatment, the immunoblot staining was totally abolished. Densitometry of Western blot bands was performed on a commercial system (Gel Doc EQ; Bio-Rad Laboratories).
Immunohistochemical Localization of Tyrosine-Nitrated Proteins in the Retina
Frozen sections were obtained from six inflamed rat eyes enucleated on postimmunization day 5. Six nonimmunized rat eyes were used as the control. Frozen sections were fixed in acetone at 4°C for 8 minutes, and the nonantigenic sites were blocked with 2% goat serum and 2% BSA for 20 minutes. The sections were incubated with a polyclonal nitroTyr antibody (1:100; Upstate Biotechnology) overnight and then with the secondary antibody, goat anti-rabbit IgG biotin (1:200; Dako), for 2 hours. Endogenous peroxidase activity was blocked with 0.6% hydrogen peroxide in methanol for 20 minutes, and the signal was enhanced with the avidin-biotin complex kit (ABC kit; Vector) before visualization with 3-amino-9-ethylcarbazole (AEC). Three control procedures were used to ascertain the specificity of the primary antibody binding: (1) primary antibody was replaced by PBS; (2) primary antibody was blocked by reacting with 5 mM commercial nitroTyr (Sigma-Aldrich) before staining; or (3) NitroTyr in the retina was reduced to aminoTyr in situ by reacting with 10 mM sodium hydrosulfite (Sigma-Aldrich) at pH 9 to 10 for 1 hour at room temperature. All three treatments totally abolished the nitroTyr-positive staining.
Detection of Cytochrome c Release
Experimental uveitis was induced in 18 Lewis rats using bovine S-antigen, and 9 rats each were killed on postimmunization days 5 and 10. An additional nine nonimmunized animals were used as the control. For both experimental and control animals, six eyes were combined for one determination, and assays were performed in triplicate. A previously described method was used to obtain the cytosolic and mitochondrial fractions.31 32 Briefly, retinas were homogenized (Dounce homogenizer; Bellco Glass Co, Vineland, NJ) in buffer containing 280 mM sucrose, 5 mM HEPES, 1 mM EGTA, 0.05% BSA, and protease inhibitors. The homogenates were centrifuged at 1000g to remove the debris, mitochondria were pelleted at 10,000g; and the supernatants were centrifuged at 100,000g to obtain the cytosolic fraction. To detect the dissociation/displacement of cyto c binding from the electron transport chain assembly, the intact mitochondria in phosphate buffer were sonicated briefly (for 20 seconds; with 4 seconds pulse on and 1 second pulse off; 70% duty cycle), before centrifuging as just described. The presence of cyto c in the supernatant was determined by Western blot (15% gel) probed with monoclonal anti-rat cyto c antibody (BD PharMingen, San Diego, CA) and liquid chromatography/tandem mass spectrometry (LC-MS/MS).
Mass Spectrometry Studies
For the mass spectrometric analysis, the 32-kDa and 29-kDa gel bands matching the nitrated bands were excised separately from the Coomassie-blue–stained gel; in-gel trypsin digestion was performed, and tryptic fragments were analyzed by LC-MS/MS.33 A portion of the digested peptide mixture was also analyzed with a custom-built capillary liquid chromatography system34 interfaced directly to an ion-trap mass spectrometer (Thermo-Finnigan LCQ; Finnigan, San Jose, CA) as described previously.35 Full mass range spectra, high-resolution zoom scan spectra, and fragment ion (MS/MS) spectra were collected with the automated, data-dependent acquisition functions of the data system. Protein identification was made by comparing the collected MS/MS spectra to the OWL protein sequence database using the Sequest database search program.36
|
Results |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
The Tyr residues were nitrated at a much higher level in BSA than were the small amounts of nitroTyr formed in SIN-1-reacted retina and in inflamed retina nitrated in vivo. Because the molar absorption coefficient for nitroTyr is only 4400/M/cm,38 39 the maximum obtainable intensity for a 354-nm, pH 7 band for the SIN-1-reacted retinal proteins is small. In the in vitro nitrated retinal proteins, the absorption for nitroTyr residues (Fig. 1A) represents a composite of absorptions from three components: Tyr-nitrated proteins (Fig. 1A ; spectrum 4), end absorption of unmodified retinal proteins (Fig. 1A ; spectrum 1), and degradation products of SIN-1 after incubation (Fig. 1A ; spectrum 2). The sum of spectra 1 and 2 is represented by spectrum 3. Subtraction of spectrum 4 from 3 revealed an absorption peak with a maximum near 360 nm, indicative of nitroTyr chromophore (Fig. 1A) . This absorption does not shift to 425 nm at pH 11, because there is insufficient dissociation of phenolic hydrogens in the complex mixture of proteins.38 39
|
|
The EAU eyes were obtained from the animals on postimmunization days 0, 5, 10, 12, and 14. Day 0 denotes nonimmunized control animals. The total extracted proteins were fractionated on SDS-PAGE, using both 12% and 15% gels. In the EAU retina, SDS-PAGE (12% gel) revealed 10 major protein bands at 120, 110, 77, 68, 52, 45, 41, 32, 29, and 16 kDa (Fig. 2B , lane 2). This profile was similar in the nonimmunized control animals (Fig. 2B , lane 1). Western blot analysis of EAU samples were then run in parallel with nitrated BSA. An intensely positive nitroTyr band was present in the +SIN-1 sample (Fig. 2C , lane 5), but such positive staining was absent in the –SIN-1 sample (Fig. 2C , lane 6). The blots of EAU retinas indicated three relatively intense Tyr-nitrated protein bands, located at 32, 29, and 16 kDa (Fig. 2C , lanes 2 and 3). Other retinal proteins were nitrated to a much smaller extent and were only slightly visible in the background (Fig. 2C) . Moreover, these three bands appeared early in the inflammation, on postimmunization days 5 and 10, long before the peak of inflammation (day 14). In the retinas of control animals in which no EAU was induced, no Tyr nitration was detectable in any of the proteins (Fig. 2C , lane 1). Similarly, the replacement of primary antibody with preimmune serum resulted in no positive staining for nitroTyr (Fig. 2C , lane 4).
The 32-kDa (Fig. 2C , band A) and 29-kDa (Fig. 2C , band B) bands were excised separately from an electrophoresed gel and subjected to in-gel trypsin digestion. The tryptic fragments were analyzed by LC-MS/MS. The Sequest database search revealed that band A is a mitochondrial import stimulation factor (MSF; 14-3-3-; molecular mass, 29,085 Da; Fig. 3A ), and band B is rat phosphoglycerate mutase (PGM; molecular mass, 28,478 Da; Fig. 3B ). In both MSF and PGM, six peptides were identified to match the known sequences. The data therefore unambiguously identified these proteins. The fact that these proteins correspond to Tyr-nitrated protein in the Western blot is indicated by two lines of evidence: (1) Concentrations of both Western blot bands A and B (Fig. 2C) indicated that these bands are likely to be the major proteins in that particular area, and (2) other contaminating proteins are all those of low-abundance, as revealed by the LC-MS/MS data. According to Western blot (15% gel) results, the 14-kDa band from EAU days 5 and 10 (Fig. 2C , band C) was identified as cyto c. This sample was also blotted in parallel with both rat cyto c antibody (Fig. 4 , lanes 1 and 2) and nitroTyr antibody (Fig. 4 , lanes 3 and 4). Under the denaturing conditions of SDS-PAGE, cyto c aggregates to form cyto c dimer (28 kDa) and trimer (42 kDa, band A),41 and the trimerization product, in particular, is also visible (Fig. 4) . Therefore, the nitrated 14-kDa band in Figure 2C (band C), and Figure 4 (lanes 3 and 4, band C) was confirmed to be cyto c. Identity of the cyto c band was also confirmed by LC-MS/MS, using cyto c from both whole retina and isolated mitochondria. In both cases, the coverage for MS/MS spectra was extensive, indicating a high concentration of cyto c in the gel band. Sequential studies covering postimmunization days 0, 5, 10, 12, and 14 revealed that three Tyr-nitrated proteins, including MSF, PGM, and cyto c were at near maximum intensities on days 5 and 10 and then leveled off gradually from day 10 to the peak of inflammation on day 14 (Fig. 5) . From days 5 to 12, the intensities of cyto c are the highest, followed by PGM and MSF (Figs. 5B 5C) . During the period from day 0 to day 10, the retinal morphology was well preserved. Day 12 signified the onset of disease, and the entrance of inflammatory cells was visible. At the peak of inflammation, on day 14, a massive infiltration of inflammatory cells and the loss of the photoreceptor cell layer occurred (Fig. 5A) .1 11
|
|
|
|
|
|
|
Discussion |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
In the present study, the occurrence of mitochondrial Tyr nitration preceded the arrival of macrophages, activation of retinal microglia, and oxidative modification of membrane lipids (Wu GS, et al. IOVS 2004;45:ARVO E-Abstract 744). This conclusion was based on the results of H&E staining for retinal structural alteration and common leukocyte antigen staining for inflammatory infiltrates. Further, the protein nitration also preceded the activation of resident retinal microglia. Quiescent retinal microglia reside normally at the nerve fiber, inner plexiform, and outer plexiform layers. In responding to the inflammatory stimuli, they migrate to the photoreceptor layer and release cytotoxic agents.12 Aside from nitration, the ONOO– generated in tissues also induces membrane lipid peroxidation.45 46 47 At the time of maximum protein nitration, at postimmunization day 5, lipid peroxidation products were still absent. Abundant lipid peroxidation-derived cellular carbonyls were visible, however, at day 14 (Wu GS, et al. IOVS 2004;45:ARVO E-Abstract 744).9 Therefore, in the early phase of EAU, the nitration processes are promoted over the lipid peroxidation in these protein targets.
Peroxynitrite has been implicated in the pathogenesis of several diseases. Although nitration may also result from other biochemical pathways that lead to 3-nitration of Tyr moieties in vivo, it is generally agreed that ONOO– is the most plausible species for biological nitration.48 In these systems, ONOO– generation requires NO· and O2–.16 The concomitant generation of these two agents at a localized site is known to result in the formation of ONOO– by a combination reaction at the rate of 6.7 x 109 M/s, three times faster than the rate of O2– dismutation by SOD.21 49 50 We and others have reported in the past that both O2– and NO· are among the most important primary oxidant species generated by macrophages in inflammatory diseases such as uveitis and that ONOO– is therefore the abundant oxidant under these conditions (Zhang J, et al. IOVS 1993;34:ARVO Abstract 1000).7 8 However, in the present study, nitration of mitochondrial proteins occurred before the infiltration of macrophages or of any other inflammatory cells (Fig. 5) , indicating that the generation of reactive species and formation of ONOO– occurred within the retinal cells and was not from macrophages.
Photoreceptor cells, which are responsible for all visual processes, have the highest rate of glycolysis and respiration, as revealed by the metabolic mapping of mammalian retina using H3-2-deoxyglucose autoradiography.26 Because of the high metabolic requirement, the inner segments in the photoreceptor cells are packed with mitochondria, with a density unseen in any other cells.27 28 Mitochondria are an important cellular source of O2–. It is estimated that 1% to 2% of the O2 consumed undergoes partial reduction, generating O2–.19 51 Two sites on the mitochondrial respiratory chain have also been found to achieve these conversions by the oxidation of flavin mononucleotide of NADH-dehydrogenase (complex I) and unstable ubisemiquinone (complex III).19 51 In recent years, mitochondrial production of NO· by mtNOS has been recognized.17 18 52 Mitochondrial NOS, which is located in the inner membrane and is modulated by Ca2+, controls mitochondrial respiration.17 Three isoforms of NOS are known to exist: the constitutive endothelial (eNOS), the neuronal (nNOS), and the inducible (iNOS) isoforms. mtNOS is a new isoform that has been shown to have eNOS-like activity. Nitric oxide produced by mtNOS and L-arginine is readily diffusible through cell membranes, whereas O2– is not. Therefore, it is conceivable that a charged combination product, such as ONOO–, with limited diffusion would be principally formed close to the site and in the same compartment as O2–, probably near the inner membrane.53 In the absence of macrophages, the actively respiring mitochondria in the inner segments of photoreceptor cells would be the early source of ONOO– to cause the nitration of cellular proteins at the proximity.
Cytochrome c is a member of the mitochondrial respiratory chain assembly embedded in the inner membrane compartment. The mitochondrial respiratory chain consists of four complexes (I through IV), coenzyme Q, and cyto c. Electrons flow down the chain to complex IV, where they reduce O2 to H2O. Cyto c, which is situated between complexes III and IV, is an electron carrier in this electron transport process. Unlike other respiratory chain complexes, cyto c faces intermembrane space rather than matrix.54 Therefore, nitration of cyto c without nitration of complexes I through IV might also indicate that the gradient of ONOO– produced in the mitochondria could be concentrated in the intermembrane space rather than in the matrix.
Mitochondrial DNA is a double-stranded, 16,569-bp circular molecule. It contains 37 genes coding for two rRNAs, 22 tRNAs, and 13 polypeptides. The mtDNA-encoded polypeptides are all subunits of enzyme complexes of the oxidative phosphorylation system.55 Therefore, most of the proteins required for the mitochondrial function are encoded by nuclear genes, synthesized by cytoplasmic ribosomes and imported to mitochondria after translation.56 This import process must route the cytosolic proteins to their correct submitochondrial compartments, and those destined for the innermost compartment are transported across two membranes. In recent years, it has been realized that there are cytosolic protein factors that chaperon and target cytoplasmic precursor proteins to mitochondrial membrane receptors. The MSF serves these functions.57 58 It catalyzes both the depolymerization and the unfolding of in vitro synthesized preprotein and binds to the preprotein and keeps it in an import-competent conformation. The resulting MSF-preprotein complex is then bound to a subcomplex of the mitochondrial outer membrane protein translocation machinery, and the preprotein is transferred after ATP-dependent dissociation of MSF.58 Although MSF originates as a cytosolic factor, during the chaperone process, it sits on the mitochondrial membrane receptors to transfer preproteins; therefore, MSF is exposed to the ONOO– generated within mitochondria. Phosphoglycerate mutase catalyzes the interconversion of 2- and 3-phosphoglycerate in the glycolytic/gluconeogenic pathways. These reactions are essential components in the metabolism of glucose and/or 2,3-bisphosphoglycerate in all cells.59 Although this protein does not reside intramitochondrially, it is an essential enzyme in glycolysis, one of the major reactions in mitochondrial metabolism. Tyrosine nitration leads to dysfunction of nitrated proteins, as has been shown or suggested in the case of SOD, cytoskeletal actin, neuronal Tyr hydroxylase, cytochrome P450, and prostacyclin synthase.60
In the present study, release of cyto c was detected in the early phase of EAU, on postimmunization day 5. When intact mitochondria and cytosol were separated, the release of cyto c into the cytosolic fraction was not observed. However, when mild sonication was applied to disrupt the outer mitochondrial membranes, substantial cyto c was detected intermitochondrially. In the respiratory chain, cyto c is bound to complex III and cytochrome oxidase by electrostatic interaction and is therefore stable to sonication but sensitive to most detergents.44 In the present study, no detergent was used in processing retina and mitochondria. In the previous reports, when cyto c was released from apoptotic or permeabilized mitochondria, it was often found that cyto c was already dissociated from the electron transport chain before pathologic membrane rupture.42 43 Therefore, it appears that the release of cyto c requires two simultaneous impairments: (1) rupture or permeabilization of mitochondrial outer membranes and (2) detachment of cyto c from the respiratory chain complex. In this study, the integrity of mitochondrial outer membranes was still intact on day 5, but cyto c was already displaced from its normal binding site in the respiratory chain due to Tyr nitration in the molecule. In the present study, the specific cyto c Tyr residue nitrated in vitro was found in Tyr48 (Fig. 7) . In vitro, nitration of horse heart cyto c by chemically generated ONOO– has been shown to cause nitration of a critical heme-vicinal Tyr67and Tyr48; the process promoted a conformational change, displacing the Met-80 heme ligand.15 The exact mechanism for the displacement of nitrated cyto c from the respiratory chain is not known. It is plausible that the bulky nitro group attached to the Tyr at the critical site causes unwanted steric hindrance for the proper binding to the inner membrane to occur.
When cyto c is released from its native location within the hierarchically arranged mitochondrial respiratory complexes III and IV, complex III remains mostly reduced, electron transport is interrupted, and more electrons become available for O2– formation.61 62 Therefore, the loss of cyto c further accelerates the production of O2–, suggesting that cyto c release is both a cause and a consequence of O2– formation in mitochondria. Taken together, these observations indicate that the initial round of insult induced by ONOO– in effecting nitration and dissociation of cyto c further results in more O2– production and its feedback to accelerate the destructive cycle.
In vivo nitration in pathologic conditions invariably nitrates only a small fraction of a particular protein. Therefore, the consequence of the mitochondrial nitration marks the beginning of an irreversible mitochondrial dysfunction. Gross cellular apoptosis or necrosis, however, is yet to come. In this laboratory, we performed TUNEL staining on the EAU retinas at multiple stages of inflammation and found that nuclear apoptotic change was absent at postimmunization day 5 (Rajendram R et al., unpublished observation, 2004). Further, it was recently recognized that there are two pools of cyto c in the mitochondria: one in the inner membrane and the other in the intercristal spaces connected to the inner membrane space only by relatively long thin tubules. Cytochrome c localized in these tubular structures may not be released into the cytosol in the absence of gross mitochondrial structural changes. This innermost pool of cyto c may be sufficient to maintain a portion of electron transport chain function for some time,41 and it may be why we do not see gross cellular changes at this early stage of inflammation, even though mitochondrial protein modification has definitively occurred.
In endotoxin-induced uveitis, an increase of iNOS has been noted in the ocular tissues before the inflammatory cell infiltration.63 64 In S-antigen-induced EAU, however, the initial signal leading to upregulation of mtNOS has not been dealt in the past. In an organ-specific autoimmune disease such as EAU, the CD4-positive T-cells are known to be present in the retina early in the inflammation. For example, after adaptive transfer of S-antigen specific T-cells, these T-cells were seen in the retina within 24 hours, although loss of retinal stratification was not observed until after 120 hours.65 The local antigen presentation to these S-antigen autoreactive T-cells can result in the generation of TNF- by the antigen-presenting cells. Tumor necrosis factor-, an inflammatory agonist, is known to upregulate NOS, and subsequently to produce ROS.66 67 68 TNF- can also increase mitochondrial Ca2+, a known stimulator of mitochondrial ROS.66 69 70 In this process, TNF- initially mobilizes Ca2+ from its endoplasmic storage to the mitochondria66 71 ; Ca2+ then triggers mtNOS activity.61 These CD4-positive cells also signal mast cells directly, activating their degranulation process to release more TNF-.72 73 Although mast cells are found in great quantity only in the choroid, histamine and TNF- released by these cells have been shown to infuse the photoreceptors.72 73
In summary, in the early phase of EAU before leukocyte infiltration, we found that the three major Tyr-nitrated retinal proteins are MSF, PGM, and cyto c, all of which are mitochondria-related proteins. Immunohistochemistry revealed nitroTyr to be exclusively localized in the inner segments of photoreceptor cells, a layer known to be densely populated with mitochondria. These findings provide evidence for a rather selective process that modifies specific proteins in vivo. These proteins were also nitrated long before the arrival of macrophages. It appears that in this stage of inflammation, mitochondria are the major source of ONOO–, and mitochondrial proteins are the prime target for inactivation by the mitochondrial ONOO–. Hence, for the first time, these findings implicated photoreceptor mitochondria as initiators of the proinflammatory response in EAU. Although these studies revealed nitration of mitochondria-related proteins, further evaluation, in particular, of DNA fragmentation and other apoptotic changes may demonstrate the protein nitration in the pathogenesis of the disease and possibly its therapeutic outcome.
|
Footnotes |
|---|
Submitted for publication December 27, 2004; revised February 8, 2005; accepted February 12, 2005.
Disclosure: G.S. Wu, None; T.D. Lee, None; R.E. Moore, None; N.A. Rao, 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.
Corresponding author: Narsing A. Rao, Doheny Eye Institute, 1450 San Pablo Street, DVRC Rm 211, Los Angeles, CA 90033-1088; nrao{at}usc.edu.
|
References |
|---|
|
TopAbstractMethodsResultsDiscussionReferences |
|---|
