Abstract
Diabetic retinopathy, a retinal vascular disease, is inhibited in animals treated with aminoguanidine, an inhibitor of inducible nitric-oxide synthase. This treatment also reduces retinal protein nitration, which is greater in diabetic rat retina than nondiabetic retina. As an approach to understanding the molecular mechanisms of diabetic retinopathy, we sought the identity of nitrotyrosine-containing proteins in retina from streptozotocin-induced diabetic rats and in a rat retinal Müller cell line grown in high glucose (25 mm). Anti-nitrotyrosine immunoprecipitation products from rat retina and Müller cells were analyzed by LC-MS/MS. Ten nitrated proteins in diabetic rat retina and three nitrated proteins in Müller cells grown in high glucose were identified; three additional nitrotyrosine-containing proteins were tentatively identified from diabetic retina. The identified nitrotyrosine-containing proteins participate in a variety of processes including glucose metabolism, signal transduction, and transcription/translation. Among the nitrated proteins were insulin-responsive glucose transporter type 4 (GLUT-4), which has been implicated previously in the pathogenesis of diabetes mellitus; exocyst complex component Exo70, which functions in insulin-stimulated glucose uptake of GLUT-4-containing vesicles; and fibroblast growth factor receptor 2, which influences retinal vascularization via fibroblast growth factor signaling. Nitration of tyrosine phosphorylation sites were identified in five proteins, including GLUT-4, exocyst complex component Exo70, protein-tyrosine phosphatase η, sensory neuron synuclein, and inositol trisphosphate receptor 3. Quantitation of nitration and phosphorylation at common tyrosine modification sites in GLUT-4 and protein-tyrosine phosphatase η from diabetic and nondiabetic animals suggests that nitration reduced tyrosine phosphorylation ∼2× in these proteins from diabetic retina. The present results provide new insights regarding tyrosine nitration and its potential role in the molecular mechanisms of diabetic retinopathy.
Retinopathy is one of the most common microvascular complications of diabetes mellitus and a leading cause of blindness. Although hyperglycemia is a risk factor in both type 1 and type 2 diabetes, the molecular mechanisms of the disease remain unclear (1, 2). Previous studies demonstrate that diabetic retinopathy involves inflammatory processes (3–5), that nitric oxide (NO)1 and NO-related reactive nitrogen species are important inflammatory mediators (2, 6), and that increased production of superoxide and NO occurs in diabetic retina (7). A possible role for NO in the pathogenesis of diabetic retinopathy has been suggested by findings that inducible nitric-oxide synthase (iNOS) has a regulatory effect on retinal oxygenation in experimental diabetes (8) and that the development of diabetic retinopathy in animals can be slowed by aminoguanidine (9), an inhibitor of iNOS, or by deficiency of iNOS (10). Furthermore higher nitrotyrosine levels have been observed in retina from diabetic rats than from nondiabetic animals and in rat retinal Müller cells cultured in high glucose (25 mm) relative to the cells grown in low glucose (5 nm) (11).
Nitrotyrosine is a low abundance oxidative protein modification occurring in about 1 in 106 tyrosines (12, 13). It is formed from the reaction of free or protein-bound tyrosine with reactive nitrogen species such as free radical nitrogen dioxide (14) and peroxynitrite (15). Tyrosine nitration has been associated with several inflammatory and neurodegenerative diseases including, for example, retinal ischemia, lung infection, Parkinson disease, Alzheimer disease, and Huntington disease (16). Previously we showed that tyrosine nitration in rat retina was modulated by light (17) and that nitration impairs aldolase A activity (18). Little is known about the targets of tyrosine nitration in retina and the role nitration may play in the pathophysiology of diabetic retinopathy. As an approach to understanding the pathogenic mechanisms of diabetic retinopathy, we sought the identity of proteins that are nitrated in diabetic rat retina and in rat retinal Müller cells grown in high glucose. Nitrotyrosine-containing proteins were selected by immunoprecipitation, and tyrosine nitration sites were identified by tandem mass spectrometry. The relative amount of nitration and phosphorylation at a common tyrosine in two proteins from diabetic and nondiabetic retina was compared by peptide mass mapping. The physiological roles of the identified proteins and the localization of nitration sites within structural and functional domains provide new insights into possible molecular mechanisms associated with this blinding disease.
EXPERIMENTAL PROCEDURES
Animal Procedures—
Male Sprague-Dawley rats (225–250 g) were randomly assigned to become diabetic or remain as nondiabetic (11). Diabetes was induced by intraperitoneal injection of a freshly prepared solution of streptozotocin in citrate buffer (pH 4.5) at 55 mg/kg of body weight. Insulin was given as needed to achieve slow weight gain without preventing hyperglycemia and glucosuria (0–2 units of insulin subcutaneously, 0–3 times/week). Diabetic rats were insulin-deficient but not grossly catabolic. All animals had free access to food and water and were maintained under a 14-h on/10-h off light cycle for 8–10 weeks. Food consumption and body weight were measured weekly. Just before the animals were sacrificed, serum glucose levels were measured using glucose oxidase-based methods, and glycated hemoglobin (an estimate of the average level of hyperglycemia over the previous 2 months) was measured by affinity chromatography (Glyc-Affin, Pierce) (9). Treatment of animals conformed to the Association for Research in Vision and Ophthalmology Resolution on Treatment of Animals in Research and to institutional animal care guidelines.
Müller Cell Culture Conditions—
The transformed rat retinal Müller cell line rMC-1 was obtained as a gift from Dr. V. R. Sarthy, Northwestern University Medical School, Chicago, IL (19). The rMC-1 cell line expresses both glial fibrillary acidic protein and cellular retinaldehyde-binding protein, markers for retinal Müller cells. Cells were initially cultured and passaged in Dulbecco's modified Eagle's medium containing 5 mm glucose and 10% fetal bovine serum as described previously (11). Experiments were performed using 50% confluent cells in which the proliferation rate had been slowed down by reducing the concentration of fetal calf serum in media to 2%. rMC-1 cells grown under these conditions were incubated in either 5 or 25 mm glucose with the media changed every other day for up to 5 days. Cells were harvested by treatment with 0.5% trypsin and 0.02% EDTA.
Retina and Müller Cell Sample Preparation—
Rats were euthanized in CO2 atmosphere, and retinas were immediately isolated using a modified Winkler technique (20) such that the vitreal body was removed before the retina was excised. Isolated retinas were rinsed in phosphate-buffered saline and frozen at −80 °C until analysis. Three retinas from three diabetic rats were combined, and three retinas from three nondiabetic rats were combined for immunoprecipitation with anti-nitrotyrosine antibody. The two pooled retina samples were homogenized separately in Pierce mammalian protein extraction reagent (M-PER) and centrifuged at 13, 000 × g for 30 min, and the protein content of the soluble fraction was estimated by a modified Bradford assay (21). Rat retinal rMC-1 cells were grown in 10-cm Petri dishes in the presence of either 5 or 25 mm glucose, harvested from four dishes each, and processed as described above for nitrotyrosine immunoprecipitation.
For analyses of insulin-responsive glucose transporter type 4 (GLUT-4) and protein-tyrosine phosphatase η (PTP-η), retina pairs from seven diabetic and seven nondiabetic rats were homogenized in 2% SDS and 100 mm triethylammonium bicarbonate (100 μl), incubated for 45 min at room temperature with 0.2 unit of DNase I, and centrifuged at 20,000 × g for 20 min, and the supernatant was collected. Each extraction was repeated three times, and the protein content of the soluble fractions was estimated with the Pierce BCA protein assay. For peptide mass mapping of GLUT-4 and PTP-η, each retinal extract (10 μg) was fractionated by SDS-PAGE (10% acrylamide), and the gel was stained with colloidal Coomassie Blue (GelCode Blue, Pierce). For immunoprecipitation with anti-GLUT-4 or anti-PTP-η antibodies, the retinal extracts were pooled (100 μg/animal × 7), protein was precipitated with acetone to lower the SDS concentration, pellets were suspended in Pierce M-PER protein extraction buffer, and soluble protein was recovered by centrifugation.
Immunoprecipitation—
Nitrotyrosine immunoprecipitations were performed essentially as described previously (22). Briefly rabbit anti-nitrotyrosine polyclonal antibody (90 μg; Chemicon International, Temecula, CA) was coupled with immunopure immobilized protein G beads (400 μl; Pierce) for 2 h with gentle shaking in 0.14 m NaCl, 0.008 m sodium phosphate, 0.002 m potassium phosphate, and 0.01 m KCl at pH 7.4 and then washed to remove unbound antibody (Pierce immunoprecipitation kit, product 45225). The bound antibody was then covalently cross-linked to protein G with gentle shaking (1 h) in disuccinimidyl suberate (final concentration, 0.0025%), washed briefly with a low pH buffer (pH 2.8) to remove non-cross-linked antibody, and re-equilibrated into the above binding buffer. The immobilized anti-nitrotyrosine antibody (∼90 μg) was incubated at 4 °C overnight with 500 μg of retina extract or with 1500 μg of Müller cell extract in the above binding buffer (500 μl) and then washed three times with the binding buffer (400 μl), and nitrated proteins were eluted with 62.5 mm Tris-HCl and 2% SDS at pH 7.0 (60 μl, 60 °C for 20 min, three times). The eluate (∼180 μl) was vacuum-concentrated to ∼50 μl and applied to SDS-PAGE.
Immunoprecipitations with rabbit anti-GLUT-4 polyclonal antibodies (70 μg; Santa Cruz Biotechnology Inc., catalogue number sc-7938) and with goat anti-PTP-η polyclonal antibodies (70 μg; Santa Cruz Biotechnology Inc., catalogue number sc-13801) were performed using pooled protein samples prepared from diabetic and nondiabetic retina as described above. Antibodies were coupled to protein G beads and incubated overnight with the retinal proteins, and bound proteins were eluted as described above.
SDS-PAGE and Western Analysis—
Nitrotyrosine immunoprecipitation products (50%, v/v) were fractionated by 12% acrylamide SDS-PAGE and detected with colloidal Coomassie Blue (GelCode Blue, Pierce), and gel slices were excised from the top to the bottom of the lane for tryptic digestion and mass spectrometric protein analysis. Nitrotyrosine immunoprecipitation products (50%, v/v) of each sample were also subjected to Western analysis (17) using 12% acrylamide SDS-PAGE, polyvinylidene fluoride membrane (Millipore), mouse anti-nitrotyrosine monoclonal antibody (Millipore), and chemiluminescence detection (GE Healthcare). Western analysis with anti-GLUT-4 and anti-PTP-η antibodies utilized the same methods but with 10% acrylamide gels.
Protein Identification by LC-MS/MS—
Following SDS-PAGE, immunoprecipitation products were subjected to in situ tryptic digestion, and peptides were extracted, reduced in 100 mm DTT, and alkylated with iodoacetamide (23). The dried tryptic peptide mixture was dissolved in 15 μl of 0.1% formic acid and 5% acetonitrile for mass spectrometric analysis. Tryptic peptide mixtures were analyzed by on-line LC-MS/MS on a QTOF2 mass spectrometer (Waters) using a Cap LC system (Waters), a 0.3 × 3-mm trapping column (C18 PepMap 100, LC Packings), a reverse phase separating column (75 μm × 5 cm, Vydac C18), and a flow rate of 250 nl/min (24). Gradient LC separation was achieved with aqueous formic acid/acetonitrile solvents. The QTOF2 mass spectrometer was operated in standard MS/MS switching mode with the three most intense ions in each survey scan subjected to MS/MS analysis. Instrument operation and data acquisition utilized MassLynx 4.1 software (Waters). Initial protein identifications from MS/MS data utilized the Mascot search engine (Matrix Science, version 2.1.03) and the Swiss Protein database (release 51.1, November 17, 2006). The Swiss Protein database search parameters included all mammals (∼47,000 total sequence entries), two missed tryptic cleavage sites allowed, precursor ion mass tolerance of 1.2 Da, fragment ion mass tolerance of 0.8 Da, and protein modifications for Tyr nitration, Asn and Gln deamidation, Met oxidation, and Cys carbamidomethylation. Select gel slice MS/MS data sets were also analyzed by the same search engine and parameters against all mammals in the National Center for Biotechnology Information (NCBI) database (release 20070818, August 18, 2007, ∼669,000 total sequence entries). A minimum Mascot ion score of 25 was used for accepting all peptide MS/MS spectra except for nitrotyrosine-containing peptides, which were examined manually as described previously (22). The uniqueness of the nitrated protein identifications was evaluated by BLAST search of the NCBI database (release cited above) for rat sequences (∼70,000 total entries) and mouse sequences (∼139,000 total entries). BLAST searches were also performed using rat sequences (∼6300 total entries) and mouse sequences (∼13,600 total entries) in the Swiss Protein database (release 54.1, August 21, 2007).
Protein Identification by Peptide Mass Mapping—
Coomassie Blue-stained SDS-PAGE bands from retinal extracts that aligned with immunoreactive bands for GLUT-4 at ∼55 kDa or with immunoreactive bands for PTP-η at >100 kDa were excised and digested with trypsin. Extracted peptides from like samples were combined and analyzed by LC-MS/MS on a QTOF2 mass spectrometer as described above. The target proteins were not detected by MS/MS; therefore the following peptide mass mapping strategy was utilized. All theoretical tryptic peptides from GLUT-4 and PTP-η were generated in silico with PeptideMass using the following parameters: [M + H]+, monoisotopic, one missed cleavage, mass cutoff of 500 Da, and all known post-translational modifications. Each theoretical tryptic peptide was matched manually to the singly and doubly charged precursor ions that were generated by LC-MS to obtain the observed tryptic peptide masses [M + H]+. The observed tryptic peptide masses ([M + H]+) were then searched for protein identification using the Mascot peptide mass fingerprint search engine (Matrix Science, version 2.1.03) and the Swiss Protein database (release 54.6x, December 14, 2007). The Swiss Protein database search parameters included all mammals (∼57,000 total sequence entries), one missed tryptic cleavage site allowed, singly charged monoisotopic ion, peptide mass tolerance of 100 ppm, and protein modifications for Tyr nitration and Tyr phosphorylation. The relative amounts of nitration or phosphorylation in select peptides from GLUT-4 and PTP-η were estimated based on the ion intensity of the modified and unmodified peptides essentially as described previously for nitration (17) and glycosylation (25).
Bioinformatics Analyses—
Protein structural and functional domain analysis was performed with ScanProsite, Motifscan, InterProScan, ProDom, and Pfam. Analysis of possible metabolic networks involving the identified nitrotyrosine-containing proteins was performed with Ingenuity Pathways Analysis software (Ingenuity® Systems).
RESULTS
Characteristics of the Diabetic Rats—
The streptozotocin-induced diabetic rats used in this study were hyperglycemic with mean serum glucose and glycated hemoglobin concentrations (301 ± 45 mg/dl and 9.0 ± 1.4%, respectively; n = 8) that were significantly greater than observed in nondiabetic rats (57 ± 7 mg/dl and 3.3 ± 0.4%, respectively; n = 8). The diabetic rats also failed to gain weight at a normal rate; mean body weights at 8–10 weeks of the study for diabetic rats and nondiabetic animals were 234 ± 55 and 448 ± 78 g, respectively (n = 8 per category).
SDS-PAGE and Western Analysis following Nitrotyrosine Immunoprecipitation—
Anti-nitrotyrosine immunoprecipitation products from rat retina and rMC-1 Müller cells were analyzed by SDS-PAGE and Western blot. Coomassie Blue staining was significantly more intense for the immunoprecipitation products from diabetic rat retina than from nondiabetic rat retina (Fig. 1A) and somewhat more intense for rMC-1 cells grown in 25 mm glucose relative to cells grown in 5 mm glucose (Fig. 1B). Nitrotyrosine immunoreactivity was also greater in immunoblots of the immunoprecipitation products from diabetic rat retina than from the nondiabetic rat retina (Fig. 1A). Little difference was apparent in nitrotyrosine immunoreactivity of the immunoprecipitation products from the rMC-1 cells grown in either in 25 or 5 mm glucose (Fig. 1B).
SDS-PAGE and Western analysis of nitrotyrosine immunoprecipitation products. A, nitrotyrosine immunoprecipitation products from diabetic rat retina. B, nitrotyrosine immunoprecipitation products from Müller cells grown in either 5 or 25 mm glucose. D, diabetic; ND, nondiabetic; Ab, anti-nitrotyrosine antibody (1 μg); BSAnY, nitrated BSA standard (∼0.1 μg) (17); LMW, low molecular weight markers, which include BSA (∼1 μg) at ∼75 kDa not chemically modified with tetranitromethane (17).
Nitrotyrosine-containing Proteins—
Following SDS-PAGE and tryptic digestion, immunoprecipitation product peptides were subjected to LC-MS/MS. Tandem mass spectra provided the sequence of 13 nitrotyrosine-containing peptides from diabetic rat retina and three nitrotyrosine-containing peptides from rat retinal Müller cells grown in high glucose (Table I), one of which contained two nitrated tyrosines. Each nitrotyrosine-containing peptide sequence was matched to a protein by Mascot analysis of all mammalian sequences in the Swiss Protein database and confirmed by BLAST analysis of rat and mouse sequences in both the Swiss Protein and NCBI protein databases. The expectation values (E-values) derived from the BLAST analyses of the rat and mouse sequences in the NCBI database are shown in Table I and support the statistical significance of the majority of alignments and protein identifications (E-values < 0.046). E-values reflect the number of higher scoring alignments expected to occur by chance. BLAST analyses of the larger mouse database yielded lower significance alignments (E-values between 0.07 and 0.18) for inositol 1,4,5-trisphosphate receptor type 3, transcription termination factor I-interacting protein 5, and APOBEC1-stimulating protein. To further support these three alignments, the MS/MS data from the relevant gel slices were also analyzed by Mascot search against all 669,000 mammalian sequences in the NCBI database; this search identified the same three proteins from the peptide sequences shown in Table I.
Nitrotyrosine-containing proteins identified by LC-MS/MS
Each nitrated peptide sequence was analyzed by BLAST search of the Swiss-Prot and NCBI databases, rat and mouse species, and found to match exactly and uniquely to the indicated protein sequence. E-values obtained from the BLAST search reflect the number of higher scoring alignments expected to occur by chance. At the time of these analyses, the NCBI database contained the greater number of entries, namely 70,000 rat sequences and 139,000 mouse sequences. Y*, nitrated tyrosine; C#, carbamidomethylated Cys; N@, deamidated Asn; Q@, deamidated Gln; M$, oxidized Met.
A representative MS/MS spectrum is shown in Fig. 2, and the data interpretation used to identify the nitrotyrosine-containing 13-amino acid peptide from GLUT-4 is described here in detail. A doubly charged ion ([M + 2H]2+, m/z = 791.30) in the MS spectrum (scan number = 73) of the peptide that eluted at retention time 14.266 min was selected after Mascot analysis. The singly charged b and y product ions (y1, y3, y4, y10, y11, b1, b4, b8, b9, b10, and b11) are labeled in the MS/MS spectrum with the corresponding amino acid sequence VIEQSY*N@ATWLGR shown where Y* is nitrotyrosine and N@ is deamidated Asn. This peptide sequence matched exactly to residues 51–63 of the protein, and nitrotyrosine was assigned to Tyr56 (Table I). The loss of NH3 from the singly charged b8 and b10 ions was also detected as labeled in the MS/MS spectrum. The E-values generated from the BLAST searches of the determined peptide sequence support the protein identification. Similar approaches were used to interpret the 12 other MS/MS spectra from diabetic rat retina shown in supplemental Fig. 1 and the three MS/MS spectra from Müller cells grown in high glucose shown in supplemental Fig. 2.
Representative MS/MS spectrum of a nitrotyrosine-containing peptide from diabetic rat retina. The MS/MS spectrum of a nitrated peptide from GLUT-4 is shown (MS/MS spectrum Mascot ion score is 13; Y*, nitrated tyrosine; N@, deamidated asparagine).
Nitration at Tyrosine Phosphorylation Sites in GLUT-4 and PTP-η—
Nitration of tyrosine was identified at phosphorylation sites in several proteins, including GLUT-4 peptide 51VIEQSYNATWLGR63 and PTP-η peptide 1032NVYAIVMLTK1041. To investigate whether tyrosine nitration possibly competes with phosphorylation in these peptides, GLUT-4 and PTP-η from diabetic and nondiabetic retina (from seven rats each) were excised from SDS-PAGE (supplemental Fig. 3), and tryptic peptides were analyzed by LC-MS/MS. Neither protein was detected by MS/MS due to low abundance, but peptide mass mapping provided 19–25% sequence coverage with average errors of 26–34 ppm and E-values of e−10 to e−11, confirming the presence of the target proteins in the excised gel bands (supplemental Table 1). For the above two peptides, singly, doubly, and triply charged ions were quantified by ion intensities for the nitrated, phosphorylated, and unmodified species, all of which were found in both diabetic and nondiabetic retina (supplemental Table 2). Relative amounts are summarized in Fig. 3 and presented in detail in supplemental Table 2. Similar amounts of unmodified peptide were detected in diabetic and nondiabetic retina and averaged ∼47% for the GLUT-4 peptide and ∼61% for the PTP-η peptide. For both peptides, diabetic retina contained relatively more of the nitrated (24–42%) than phosphorylated (10–13%) species. In nondiabetic retina, the amount of phosphorylated peptides (20–26%) was ∼2× higher relative to diabetic retina (10–13%). Notably for PTP-η in nondiabetic retina, the amount of phosphorylated peptide (26%) was significantly greater than the nitrated species (15%).
Relative amounts of nitration and phosphorylation at same as common tyrosine modification sites in GLUT-4 and PTP-η. GLUT-4 peptide 51VIEQSYNATWLGR63 and PTP-η peptide 1032NVYAIVMLTK1041 are each nitrated at a tyrosine phosphorylation site (underlined) in rat retina. Tryptic peptides from rat retinal GLUT-4 and PTP-η from seven diabetic and seven nondiabetic animals were analyzed by LC-MS (supplemental Fig. 3 and supplemental Table 1). The relative amounts of the tyrosine nitrated (nY), tyrosine phosphorylated (pY), or unmodified peptides are shown based on the sum peptide ion intensities for singly, doubly, and triply charged peptide species (see supplemental Table 2).
DISCUSSION
Diabetic retinopathy is a major complication of diabetes mellitus, a disease affecting ∼200 million people worldwide, including about 20 million in the United States (1). The prevalence of diabetic retinopathy increases with the duration of diabetes; nearly 100% with type 1 and over 60% of those with type 2 diabetes have some retinopathy after 20 years (1). This retinopathy is a poorly understood, progressive complication of diabetes involving the retinal vasculature. In early stages of the disease, microaneurysms form in the retina that can allow serum to leak into the retina, causing edema and decreased vision. In the later stages, known as proliferative diabetic retinopathy, retinal and optic nerve head neovascularization occurs, generating hemorrhaging, and associated fibrovascular tissue may contract, leading to scarring and retinal detachment. Without timely diagnosis and treatment, blindness may ensue. Previous studies suggest that inflammatory processes involving nitric oxide mediators and the generation of nitrotyrosine-containing proteins may play a role in the disease (2, 9, 11, 26). Diabetes also appears to contribute to the development of other ocular disorders such as glaucoma (2). Notably aminoguanidine inhibition of iNOS not only slows the progression of diabetic retinopathy in rats, it also protects rat retina from injury from ischemia (27) and glaucoma (28). To probe the role tyrosine nitration may play in diabetic retinopathy, we sought the identity of nitrotyrosine-containing proteins in diabetic rat retina and in rat retinal Müller cells cultured in the presence of high glucose.
Nitrotyrosine immunoaffinity precipitation was utilized to preferentially enrich for nitrated proteins, and SDS-PAGE fractionation was used to simplify the complexity of the immunoprecipitates for LC-MS/MS analysis. To select for low abundance nitrated peptides, ion scores in the Mascot database searches were disregarded, and 1256 candidate spectra were compiled, including 821 spectra from diabetic rat retina and 435 spectra from Müller cells cultured in 25 mm glucose. Manual inspection and interpretation of all the candidate MS/MS spectra led to the identification of 16 nitrotyrosine-containing peptides (Table I). These methods provided statistically significant identification of tyrosine nitration sites in 10 retinal proteins from in vivo retina and three proteins from cultured retinal Müller cells. The sequence of one of the nitrotyrosine-containing peptides is common to both γ-crystallin B and γ-crystallin C, and it is not clear whether both of these proteins are nitrated or only one. Three other potential nitrotyrosine-containing proteins were identified. Alignments are considered tentative for inositol 1,4,5-trisphosphate receptor type 3, transcription termination factor I-interacting protein 5, and APOBEC1-stimulating protein because of lower significance E-values from the BLAST searches of NCBI mouse sequences. However, all the MS/MS spectra provided reliable peptide sequence information, and the tentative protein assignments are considered in the following discussion of structure, function, and pathways.
As an approach to determining the significance of tyrosine nitration in diabetic retinopathy, metabolic network systems involving the nitrotyrosine-containing proteins were sought using bioinformatics methods. Pathway analysis suggested that 10 of the nitrated proteins could function in metabolic networks involving cellular growth and proliferation, connective tissue development and function, and/or musculoskeletal system development. Of these networks, cellular growth and proliferation is consistent with the neovascularization manifested in diabetic retinopathy. The 10 proteins highlighted in the cellular growth and proliferation pathway included GLUT-4, sensory neuron synuclein, exocyst complex component Exo70, potassium voltage-gated channel KQT 2, PTP-η, serine/threonine-protein kinase PLK1, inositol 1,4,5-trisphosphate receptor 3, transcription termination factor I-interacting protein 5, fibroblast growth factor receptor 2, and G protein-activated inward rectifier potassium channel 4.
Bioinformatics analysis of protein domains was also pursued for insights into the physiological significance of the identified tyrosine nitration sites. All of the identified nitration sites occur in structural or functional domains as shown schematically in Fig. 4. Of particular interest, five of the identified nitration sites appear to be targets of tyrosine phosphorylation. For example, nitrotyrosine (Tyr(N)352) was detected in the MIR 4 domain of inositol 1,4,5-trisphosphate receptor type 3 (Fig. 4A), a multipass membrane protein. The MIR domain refers to “mannosyltransferase, inositol 1,4,5-trisphosphate receptor, and ryanodine receptor” domain common to several signal transduction-associated proteins and functions in binding and catalysis. Tyr352 is a component of the tyrosine kinase phosphorylation motif RnagEkikY (residues 344–352; lowercase denotes variable residues) and therefore also a target of phosphorylation. In GLUT-4 (Fig. 4B), a 12-pass membrane protein, nitrotyrosine (Tyr(N)56) was detected in an extracellular domain adjacent to a glycosylation site. Tyr56 is a component of the tyrosine kinase phosphorylation motif KviEqsY (residues 50–56) and a target of phosphorylation. In PTP-η (Fig. 4C), a single pass membrane protein, nitrotyrosine (Tyr(N)1034) was detected in the catalytic tyrosine-protein phosphatase domain. Tyr1034 is a component of the tyrosine kinase phosphorylation motif RmvwEknvY (residues 1026–1034) and a target of phosphorylation. Nitration of sensory neuron synuclein (Fig. 4D) was detected at Tyr(N)39 in repeat 2 of an 11-amino acid tandem repeat. Tyr39 is a component of the tyrosine kinase phosphorylation motif KtkEgvmY (residues 32–39) and a target of phosphorylation. Nitration of exocyst complex component Exo70 (Fig. 4E) was identified at another phosphorylation target, namely Tyr409 within the tyrosine kinase phosphorylation motif KndpDkeY (residues 402–409).
Localization of tyrosine nitration sites in protein domains. The identified tyrosine nitration sites were bioinformatically localized to the indicated protein domains (A–P), which are described in the text. ▪, transmembrane;
, cytoplasmic;
, extracellular; □, random; gray box, domain;
, tyrosine phosphorylation motif. Transmem, transmembrane; PKC, protein kinase C; DDT, DNA-binding homeobox-containing proteins and the different transcription and chromatin-remodeling factors; MBD, methyl-CpG-binding domain; PHD, plant homeodomain.
The possibility that nitration may compete with tyrosine phosphorylation and contribute to mechanisms of pathology in diabetic retinopathy prompted us to compare the relative amounts of nitrated and phosphorylated peptides in GLUT-4 and PTP-η from diabetic and nondiabetic rat retina. Our quantitative analyses of two peptides exhibiting nitration at tyrosine phosphorylation sites support more nitrated species and fewer phosphorylated species in diabetic retina compared with nondiabetic retina. Notably nitrated peptides were detected in both diabetic and nondiabetic retina, but the results are consistent with earlier Western analyses (11) showing higher levels of nitrotyrosine in diabetic retina. Although peptide mass mapping does not associate the modifications with specific residues, in combination with tyrosine phosphorylation motifs and our nitrotyrosine MS/MS data, the results suggest a relative reduction in tyrosine phosphorylation in GLUT-4 and PTP-η in the diabetic animals due to tyrosine nitration. GLUT-4 is the main glucose transporter activated by insulin in skeletal muscle and adipocytes. Based on immunodetection, others have reported that GLUT-4 is not expressed in human ocular tissues (29); however, our data supports GLUT-4 expression in rat retina by both immunodetection and peptide mass mapping. Nitration of GLUT-4 could impact glucose transport into retinal cells and play a role in diabetic retinopathy (30, 31). PTP-η converts protein tyrosine phosphate to protein tyrosine, contributing to a variety of cell signaling process. This phosphatase has recently been shown to modulate vascular endothelial growth factor signaling (32), further suggesting that nitration of a tyrosine phosphorylation site in the catalytic domain could impact retinal angiogenesis in diabetic retinopathy.
We identified tyrosine nitration sites in several other types of structural or functional domains (Fig. 4). Tyrosine nitration was localized to the FAD-dependent oxidoreductase domain of kynurenine 3-monooxygenase (Fig. 4F), a two-pass mitochondrial outer membrane protein. In potassium voltage-gated channel subfamily KQT member 2 (Fig. 4G), nitrotyrosine was detected at Tyr(N)226 in the cation channel region of this six-pass membrane protein near the S4 transmembrane voltage sensor region (residues 198–221). In G protein-activated inward rectifier potassium channel 4 (Fig. 4H), a two-pass membrane protein, nitrotyrosine was identified in the N-terminal cytoplasmic domain. Two adjacent nitrotyrosines were detected in the N-terminal cytoplasmic domain of sensory neuron sodium channel 2 (Fig. 4I), a 24-pass membrane protein. Nitration of fibroblast growth factor receptor 2 (Fig. 4J), a single pass membrane protein, was identified in the catalytic protein kinase domain adjacent to a tyrosine phosphorylation site. Tyrosine nitration was detected in Greek key 1 of γ-crystallin B and/or γ-crystallin C (Fig. 4K), homologous cytosolic proteins that contain four β/γ Greek key domains composed of four antiparallel β-strands. In BTB/POZ domain-containing protein 5 (Fig. 4L), nitrotyrosine was detected in Kelch protein interaction domain 1. The BTB domain (broad-complex, tramtrack, and bric a brac), also known as the POZ domain (poxvirus and zinc finger), is a homodimerization domain in proteins containing multiple zinc fingers of the C2H2 type. In placental prolactin-like protein K (Fig. 4M), a secreted protein, Tyr(N)214 was detected adjacent to a threonine phosphorylation site (Thr213) within the casein kinase II phosphorylation motif TylE (residues 213–216). Several other phosphorylation sites are localized within the C-terminal region. Tyrosine nitration in serine/threonine-protein kinase PLK1 (Fig. 4N) was identified at Tyr(N)481 immediately adjacent to protein interaction domain POLO box 1 (residues 417–480). In transcription termination factor I-interacting protein 5 (Fig. 4O), a nuclear protein, Tyr(N)1843 was detected in the C-terminal region near the protein interaction bromodomain (residues 1755–1825). Tyrosine nitration of APOBEC1-stimulating protein (Fig. 4P), which shuttles between the nucleus and cytoplasm, was detected in the RRM1 domain, an RNA recognition motif.
The biological significance and relevance to diabetic retinopathy of the identified nitration events have yet to be determined; however, the functions of the other nitrotyrosine-containing proteins in Table I and Fig. 4 all appear to be susceptible to possible modulation by nitration. Nitration of γ-crystallin B and γ-crystallin C may enhance the chaperone function of these proteins as reported for oxidative modification of α-crystallin and HSP 70 (33, 34). Sensory neuron synuclein may be involved in modulating axonal architecture and neurofilament network integrity (35), and although its function remains unclear (36), nitration of the observed tyrosine phosphorylation site could have biological consequences. Exocyst complex component Exo70 functions in the docking of exocystic vesicles with specific sites on plasma membranes (37) and appears to play a role in insulin-stimulated glucose uptake of GLUT4-containing vesicles (38); nitration of the observed tyrosine phosphorylation site could alter its activity in diabetic retinopathy. BTB/POZ domain-containing proteins are implicated in a variety of biological processes including DNA binding, regulation of transcription, and organization of macromolecular structures like chromatin (39); the observed nitration in a Kelch protein interaction domain could impact function. The potassium voltage-gated channel subfamily KQT 2 is a multipass membrane protein that functions in neuronal excitability (40). Its current can be down-regulated by tyrosine kinase inhibitors, suggesting that the observed tyrosine nitration within the cation channel region may also influence channel activity. Placental prolactin-like protein K belongs to a large family of proteins related to the lactation-promoting hormone prolactin. Within this family, it shares sequence similarity with proliferin, which is a known regulator of angiogenesis (41). Kynurenine 3-monooxygenase, also known as kynurenine 3-hydroxylase, functions in tryptophan catabolism, converting kynurenine to 3-hydroxykynurenine, which can adduct protein-bound Lys, His, and Cys residues (42). The observed nitration of a functional domain could impact the activity of this enzyme. Serine/threonine-protein kinase PLK1 functions in regulating mitotic activity (43); the observed nitration adjacent to the POLO protein interaction domain could impact signal transduction pathways. Inositol 1,4,5-trisphosphate receptor 3 mediates the release of intracellular calcium (44); nitration of the observed tyrosine phosphorylation site in MIR domain 4 could modulate ligand interactions and signal transduction pathways. Transcription termination factor I-interacting protein 5 is a DNA-binding protein thought to be involved in transcriptional regulation (45); the observed nitration in the C terminus adjacent to the protein interaction bromodomain could modulate its activity. APOBEC1-stimulating protein is a subunit of the apolipoprotein B mRNA post-transcriptional editing complex (46); the observed nitration within RNA recognition motif 1 could affect protein function. Fibroblast growth factor receptor 2 (FGFR2) serves as a receptor for both acidic FGF and basic FGF. FGF signaling has been implicated in retinal vascularization (47), and the observed nitration of FGFR2 adjacent to a tyrosine phosphorylation site in the catalytic tyrosine kinase domain could be relevant to diabetic retinal neovascularization. Sensory neuron sodium channel 2 functions in generating and propagating action potentials where opening of the sodium channel may be dependent upon ligand binding rather than voltage (48); the two adjacent nitrotyrosine observed near the N terminus could influence ligand/protein interactions. G protein-activated inward rectifier potassium channel 4 forms a heteromultimeric pore that allows greater flow of potassium into than out of the cell (49). The observed nitration in the N-terminal cytoplasmic domain could influence the signaling properties of the channel.
In summary, the present study provides new insights for exploring the significance of tyrosine nitration in diabetic disease processes. Five of the tyrosine nitration sites we identified are also targets of phosphorylation. We quantified the amount of nitration and phosphorylation at two of these sites and found nitration to be greater in diabetic retina, whereas nondiabetic retina appears to contain relatively more phosphorylation. The results suggest that nitration reduced tyrosine phosphorylation at these sites in diabetic retina and implicate potential consequences to signal transduction processes. All of the identified tyrosine nitration sites occur in protein domains that may modulate other protein functions and cellular processes. Further investigation is required to determine the biological consequences of the identified tyrosine nitration events and their relevance to the pathogenic mechanisms of diabetic retinopathy.
Footnotes
-
Published, MCP Papers in Press, December 28, 2007, DOI 10.1074/mcp.M700417-MCP200
-
↵ 1 The abbreviations used are: NO, nitric oxide; BTB/POZ, broad-complex, tramtrack, bric a brac domain, also known as the poxvirus and zinc finger domain; E-value, expectation value; FGF, fibroblast growth factor; FGFR2, fibroblast growth factor receptor 2; GLUT-4, insulin-responsive glucose transporter type 4; HSP, heat shock protein; iNOS, inducible nitric-oxide synthase; MIR, mannosyltransferase, inositol 1,4,5-trisphosphate receptor, and ryanodine receptor domain; Tyr(N), nitrotyrosine; PTP-η, protein-tyrosine phosphatase η; rMC-1, retinal Müller cell line 1; BLAST, Basic Local Alignment Search Tool.
-
↵* This work was supported, in whole or in part, by National Institutes of Health Grants EY00300, EY11373, EY14239, and EY15638. This work was also supported by Ohio Biomedical Research Technology Transfer Grant 05-29, The Foundation Fighting Blindness; a Research to Prevent Blindness (RPB) Challenge Grant (to the Cole Eye Institute), an RPB Senior Investigator Award (to J. W. C.), a Steinbach Award (to J. W. C. and The Cleveland Clinic Foundation). The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
-
↵S The on-line version of this article (available at http://www.mcponline.org) contains supplemental material.
- Received August 31, 2007.
- Revision received December 26, 2007.
- © 2008 The American Society for Biochemistry and Molecular Biology
NIH Funded Research - Final Version Full Access