|Home | About | Journals | Submit | Contact Us | Français|
α-Synuclein is a major component of Lewy bodies, proteinacious inclusions which are a major hallmark of Parkinson’s disease (PD). Lewy bodies contain high levels of nitrated tyrosine residues as determined by antibodies specific for 3-nitrotyrosine (3NT) and via mass spectrometry (MS). We have developed a multiple reaction monitoring (MRM) mass spectrometry method to sensitively quantitate the 3NT levels of specific α-synuclein tyrosine residues. We found a nine-fold increase (relative to controls) in levels of 3NT at Tyr-39 of α-synuclein in an inducible transgenic cellular model of Parkinson’s disease in which monoamine oxidase B (MAO-B) is overexpressed and which emulates several features of PD. Increased nitration of Tyr-39 on endogenous α-synuclein via elevations in MAO-B levels could be abrogated by the addition of deprenyl, a specific MAO-B inhibitor. The increased levels of 3NT was selective for Tyr-39 as no significant increases in 3NT levels were detected at other tyrosine residues present in the protein (Tyr-125, Tyr-133, and Tyr-136). This is the first report of increased 3NT levels of a specific tyrosine in a PD model and the first use of MRM mass spectrometry to quantify changes in 3NT modifications at specific sites within a target protein.
The nature and function of α-synuclein and its role in PD have been heavily studied since the discovery of a dominant mutation in the α-synuclein gene in a number of Greek and Sicilian families with a familial form of the disorder1. α-Synuclein was subsequently identified to be a major component of Lewy bodies in sporadic cases of the disease2. The exact function of α-synuclein, the mechanism by which it aggregates and becomes a major component of Lewy bodies is a matter of intense investigation.
Qualitatively, Lewy bodies appear to contain high levels of nitrated tyrosine residues as indicated by antibodies specific for 3NT3. Use of these antibodies has shown that α-synuclein is a particularly sensitive target for nitrating agents and that nitrated α-synuclein is present in the Lewy bodies of postmortem tissues not only from PD patients but also other neurodegenerative synucleinopathies4. Nitrated α-synuclein has also been detected in the substantia nigra (SN) and ventral midbrain of mice treated with the PD-inducing agent 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine (MPTP)5. Recombinant α-synuclein treated with nitrating agents (peroxynitrite/CO2 myeloperoxidase/ H2O2/nitrite) generates highly stable nitrated α-synuclein oligos and aggregates as a consequence of oxidation6-9 and suggests that nitration is sufficient for α-synuclein aggregation.
Levels of the neurotransmitter dopamine were found to be selectively decreased in the corpus striatum and SN of PD patients10, and subsequently dopamine deficiency was found to be associated with the loss of dopaminergic cells within the SN11. Of the catacholamine neurotransmitters, dopamine is the most chemically predisposed towards oxidation12. Oxidative stress produced by dopamine metabolism may be exacerbated in PD patients in part due to increases in MAO-B activity which has been shown to increase with age and is associated with the disease itself13, 14. MAO-B is primarily expressed within astrocytes which themselves have high levels of protective antioxidants to detoxify reactive oxygen species; however H2O2 produced via MAO-B has a high membrane permeability and may diffuse to adjacent dopaminergic neurons15, 16 which are more vulnerable to oxidative stress17-19 leading to dopamine specific degeneration, a hallmark of PD. Our laboratory has established a doxycycline (dox)-inducible dopaminergic PC12 cell line which stably expresses human MAO-B at elevated levels comparable to those observed during normal aging and in PD. The elevated expression of MAO-B in PC12 cells was found to recapitulate other aspects of PD including mitochondrial complex I (CI) deficiency20 which is also associated with PD21-23. Administration of the CI inhibitors rotenone and MPTP result in dopaminergic neuronal loss similar to what is observed in PD patients24, 25. It is of interest to note that MAO-B is the enzyme responsible for the conversion of MPTP to its protoxic form 1-methyl-4-phenylpyridinium (MPP+) which is subsequently transported into dopaminergic neurons resulting in CI inhibition and ensuing dopaminergic toxicity26, 27.
The low abundance of certain protein post-translation modifications (PTMs) such as 3NT combined with the small but consequential changes contribute to the difficulty of identifying and quantifying PTMs present at a specific position of a target protein. To conclusively identify and quantify α-synuclein PTMs in this study, we have first analyzed in vitro nitrated human recombinant α-synuclein by MS/MS to identify 3NT modified α-synuclein peptides and to determine the optimal targets for Multiple Reaction Monitoring (MRM) MS based quantitation. MRM is the most sensitive and quantitative mass spectrometry approach available28, 29 and has the advantage that it can independently measure each tyrosine in α-synuclein. This sensitive MS based technique allowed us to examine the affects of increased MAO-B expression in a PC12 cell model on 3NT modification of α-synuclein, emulating what occurs during normal aging and PD and which may contribute to α-synuclein aggregation.
Creation and maintenance of dox-inducible MAO-B PC12 cells was previously described by our laboratory20. MAO-B was induced by treating cells with 40 μg/mL of dox for 24 hours; PC12 cells were neuronally differentiated by using 50 ng/mL nerve growth factor (NGF) for 2 days prior to addition of dox. As a positive control, MAO-B PC12 cells (no dox treatment) were treated with 150 μM H2O2 for 24 hours. To inhibit MAO-B activity, 10 μM of deprenyl was added to the medium just prior to dox treatment.
Treated or control cells were lysed with RIPA lysis buffer (Sigma) containing complete mini protease inhibitor tablets (Roche). Cell lysate protein concentrations were adjusted to 5 μg/μL, then pre-cleared with EZview protein A affinity gel (Sigma) for 4 hours at 4° C. The lysate was then incubated with anti-α-synuclein antibody (Syn 202, Santa Cruz Biotechnology) overnight at 4° C with continuous gentle mixing followed by incubation of the lysate with EZview protein A affinity gel (Sigma) overnight at 4° C with continuous mixing. The resin was then washed 3 times with 1 mL of RIPA buffer and eluted by boiling in SDS-PAGE (reducing) loading buffer, then separated by SDS-PAGE on a 10% gel (1 mm, Bis-Tris NuPAGE Invitrogen). The gel was stained with Imperial protein stain (Thermo Fisher Scientific) and the band corresponding to the α-synuclein protein excised, washed, and digested with 40 ng of Asp-N (Roche) or trypsin (Porcine Sequencing grade, Promega) for 16 hours at 37° C. No reduction/alkylation step was necessary as α-synuclein contains no cysteine residues.
Recombinant human α-synuclein (Sigma) was nitrated using a modified procedure described in Cassina et. al.30. Briefly, 250 ng of recombinant α-synuclein was treated with 40 mM tetranitromethane in 0.1 M Tris-HCl (pH 8.0) and 0.1 M KCl at room temperature for 30 minutes. The sample was then run on a SDS-PAGE gel (as described above), excised and digested as described above for immunoprecipitated α-synuclein.
To identify sites of PTMs, tetranitromethane-treated human recombinant α-synuclein and α-synuclein immunprecipited from rat PC12 cells were subjected to in-gel protease digestion with Asp-N and trypsin and the resulting peptides were analyzed by electrospray HPLC-MS/MS using a hybrid quadrupole time-of-flight mass spectrometer (QSTAR, Applied Biosystems MDS SCIEX, Concorde, CAN) as described previously31, 32. Detailed protocols are listed in the legend of Supporting Information Table S-1, titled QSTAR Elite parameters. All peptide assignments from the resulting mass spectrometric data were made using the bioinformatics search engine Mascot version 2.2.04 (Matrix Sciences, London, United Kingdom)33. A complete listing of the parameters used for Mascot database searches are listed in the figure legend of Supporting Information Table S-1A, Mascot Database Searches. To provide a quantitative comparison of the levels of tyrosine nitration, MRM analysis targeted the levels of several 3NT modified peptides and their unmodified counterparts. For these experiments, a 4000 QTRAP hybrid triple quadrupole/linear ion trap mass spectrometer (Applied Biosystems) was used as described previously34. Further details of the MRM methods can be found in the legend of Supporting Information Table S-2 under MRM parameters.
Due to the relatively low expression of α-synuclein in rodent PC12 cell cultures, nitrated recombinant human α-synuclein was first analyzed by mass spectrometry to provide a tandem mass spectral data set to use as a positive control to aid in the identification of nitrated tyrosine in α-synuclein immunoprecipitated from MAO-B PC12 cell lysates. The analytical strategy for identifying and quantifying PTMs present on peptides expressed at low levels is demonstrated on Figure 1. Although human α-synuclein is not identical to the homologues expressed in rodents, it is highly conserved; both rat and mouse sequences are 95% identical (133 out of 140 amino acids) to the human sequence. The human form also contains 4 of the 5 tyrosine residues (Tyr-39, Tyr-125, Tyr-133, and Tyr-136) found in the mouse and rat homologs, although both mouse and rat α-synuclein have an additional tyrosine at position 107 not present in humans. Given this similarity, nitrated human α-synuclein samples were digested with trypsin and Asp-N and the resulting peptides analyzed by HPLC MS/MS. Analysis of the spectral data showed that a near complete sequence map of α-synuclein had been obtained, covering 99% or 89% of the protein, respectively (Supporting Information Tables S-1A and S-1B). Included in these MS/MS data were a set of high quality spectra that encompassed one or more nitrated tyrosine residues; and these were examined in detail to develop unique and chemically specific criteria for identifying α-synuclein peptides containing 3-nitrotyrosine as described below (Figure 2A, 2B and Supporting Information).
Three of the four tyrosines of interest are located in the C-terminal domain of α-synuclein at positions 125, 133 and 136. Trypsin digestion produced a large, highly-charged 38 amino acid peptide spanning residues 103-140 (m/z 1076.4+4, M = 4301.5 Da) that contained all three of these C-terminal tyrosines. While we were able to unambiguously identify this large C-terminal peptide by MS/MS, its relatively large size precluded an unequivocal determination of the precise site(s) of 3NT modification. In contrast, digestion with Asp-N resulted in a set of peptides that were more amenable to PTM identification, as they generated peptides where each of these 3 tyrosines can be separately examined. Although Asp-N digestion can be prone to producing peptides with missed cleavages, these missed cleavage products appeared consistently from digest to digest. A complete list of peptide ions, their sequence assignments and corresponding Mascot scores obtained by searching against the Swiss-Prot database are listed in the Supporting Information section (Table S-1).
To better identify peptides containing 3-nitrotyrosine, we used a set of stringent criteria to avoid incorrect assignments. This criteria consisted of: (1) observation of both unmodified and 3NT modified peptides that showed similar fragmentation patterns, (2) peptides containing a 3NT modification showed a chromatographic shift to a later eluted time (~2-4 minutes) than the corresponding unmodified peptide due to the increased hydrophobicity, (3) mass measurements of the peptide molecular weight showed good mass accuracy (< 0.15 Da, Table S-1A in Supporting Information) for both the unmodified and 3NT modified peptide, and where the latter 3NT containing peptide showed a 45 Da mass shift (+NO2, -H; Δ45 Da), (4) MS/MS fragmentation spectra of the 3NT containing peptides required the presence of 3NT “flanking” y- or b- fragment ions, and (5) peptides containing a 3NT must show 3NT immonium ions at m/z 181.1, while peptides containing an unmodified tyrosine residue showed immonium ions at m/z 136.1. Using these criteria, four 3NT modified peptides (Tyr-39, Tyr-125, Tyr-133, and Tyr-136) were identified in human α-synuclein from their corresponding MS/MS spectra (Figure 2B, and Supporting Information data Figures S-1B, S-2B, and S-3B).
The employment of these assignment criteria for 3NT-containing peptides is illustrated in Figure 2 for the α-synuclein peptide E-35-GVLYVGSKTK-45 (containing Tyr-39). The chromatographic shift in retention time is depicted in the total ion chromatogram in Figure 3 where the nitrated peptide eluted ~2 minutes after the corresponding unmodified peptide. The MS/MS spectra for the unmodified peptide and the peptide containing an 3NT modifcation at Tyr-39 are shown in Figure 2A and 2B, respectively, and clearly demonstrate the presence of either a tyrosine (m/z 136.1) or nitrotyrosine immonium ion (m/z 181.1) in the low mass region. In regards to sequence ions, the y7 ion containing the tyrosine residue was observed to be 163 Da larger than the y6 ion in the unmodified peptide, while the 3NT residue results in a y7 ion that is 208 Da larger than the y6 ion in the MS/MS spectra of the 3NT containing peptide (note that fragment ions that do not contain the tyrosine residue, such as the y6 ion at 619.4 m/z, have identical masses). The masses of the unmodified and modified peptides deduced from their respective triply charged ions at m/z 394.2 and 409.2, also show the expected 45 Da shift. In all, this and the other MS/MS spectra of nitrated and unmodified tyrosine-containing peptides (see Supporting Information), were particularly valuable for validating the nitrated peptides obtained in the rat PD cell model as described below.
α-Synuclein was immunopurified from PC12 cells treated with dox to induce MAO-B overexpression, dox + deprenyl (an MAO-B inhibitor), or H2O2 versus control cells (no dox). The immunopreciptated fractions were then subjected to SDS-PAGE separation, and the band(s) corresponding to α-synuclein were digested by trypsin or Asp-N, and analyzed by HPLC MS/MS. Sequence coverage obtained from PC12 cell lysates after immunoprecipitation and mass spectrometric analysis of the gel-separated proteins was reasonable, i.e., 51% and 57% for Asp-N and trypsin, respectively, and well within our expectations given the significantly lower amount of material available (see Supporting Information Table S-1C and S-1D). Nonetheless, this coverage was considerably lower than what we obtained from our in vitro experiments using recombinant human α-synuclein and not sufficient to accurately quantify sites of nitration. Therefore, to both identify and quantify these potential sites of nitration in the PC12 cell model, we developed a highly targeted and sensitive multiple reaction monitoring (MRM) approach.
These MRM experiments were developed and validated first using the human recombinant proteins (tetranitromethane-treated and untreated) and then used to assess relative changes in tyrosine nitration status amongst the four conditions used in the PC12 model. The tyrosine-containing peptides used for MRM quantitation were E-35-GVLYVGSKTK-45, E-126-MPSEEGYQ-134, D-135-YEPEA-140 for human and rat samples, since their sequence were identical in these regions. However, as the rat and human peptides contained Tyr-125 varied at two residues, D-115-MPVDPDNEAY-125 was used for human and D-115-MPVDPSSEAY-125 was used for rat samples. For Tyr-107 (present only in rat α-synuclein) the peptide D-98-QMGKGEEGYPQEGILE-114 was used. A partial list of MRM transitions and other MRM parameters used is presented in Table 1, details of all MRM transitions are listed in Supporting Information Table S-2.
As determined by MRM analysis, approximately 1% of Tyr-39 residues on α-synuclein obtained from PC12 cells contain a 3NT modification under control conditions. A nine-fold increase (as compared to controls) in levels of 3NT at Tyr-39 was detected by MRM in the inducible transgenic cellular (PC12) model of Parkinson’s disease in which MAO-B was overexpressed (Figure 4). The increased nitration of Tyr-39 could be abrogated by the addition of deprenyl, a specific MAO-B inhibitor, reversing 3NT levels to approximately control levels (1.3 fold change). This suggests that MAO-B activity is a critical component for the observed increase in 3NT levels, presumably due to increased hydrogen peroxide and indirect nitrative stress created as a by-product of MAO-B metabolism of dopamine35. This conclusion is corroborated by a similar increase in 3NT levels (~6 fold increase) observed following treatment of control PC12 cells with 150 μM H2O2 for 24 hours.
The increase in levels of 3NT was selective for Tyr-39; no significant increases in 3NT levels were detected at other tyrosine residues present in α-synuclein (Tyr-125, Tyr-133, and Tyr-136). While our data also suggested 3NT levels of Tyr-136 increased after MAO-B over-expression the signal was near the limit of quantitation of the MRM assay and led to significant variability between biological replicates and did not reach statistical significance. No detectable levels of 3NT modification were observed on peptides containing either Tyr-125 or Tyr-133 from any of the treatments however the unmodified peptides were detected by MRM from α-synuclein immunopurified from MAO-B PC12 cells (data not shown). Neither the unmodified or the 3NT-modified peptide containing Tyr-107 was detected from immunopurified samples by MRM despite the unmodified peptide being detected by MS/MS analysis; further MRM transition optimization may be required.
We have used an MS-based strategy to identify low abundant peptides with PTMs that involved the following: (1) identifying a set of well-validated peptide spectra using recombinant protein; (2), using these peptides to design very selective and specific MRM transitions; and (3), using well-vetted MRM transitions to exploit the sensitivity of MRM-MS to identify and quantify 3NT PTMs present on α-synuclein expressed at relatively low levels in PC12 cells prepared under four separate conditions. Here we report for the first time a collection of MS/MS spectra of α-synuclein peptides containing 3NT modifications. More importantly, utilizing the sensitivity and selectivity of MRM-MS, we demonstrated a selective 9-fold increase in 3NT levels at Tyr-39 of α-synuclein in an oxidative cellular model of Parkinson’s disease.
While mass spectrometry has been previously utilized to discover PTMs such as 3NT with some success36, 37, these studies were unable to accurately quantitate differences in PTM levels between samples. The importance of rigorously scrutinizing MS data was illustrated by a recent investigation by Laszlo Prokai and associates38 in which 3NT modified peptides reported by other groups were proven to be misassigned. Our criteria for identifying PTMs was particularly stringent due to this concern. MRM-MS has been previously utilized to measure the levels of free tyrosine and 3NT levels39, 40 however to our knowledge there are no published reports to date of the use of MRM-MS to quantify 3NT levels in a targeted fashion on a specific protein.
Aggregated α-synuclein present in Lewy bodies has been reported to be highly modified post-transcriptionally including both phosphorylation at Ser-12941 and nitration of tyrosine residues3, 4. However, neither the degree of nitration nor the identification of the specific tyrosines modified has been previously elucidated. α-Synuclein appears to be a particularly sensitive target for nitration by peroxynitrite5 and nitrated α-synuclein monomers and dimers have been shown to accelerate protein fibril formation42. Tyrosine residues within α-synuclein are also required for aggregation caused by oxidation43 and proteasomal degradation of α-synuclein has been shown to require nitration42. Nitration of Tyr-39 in particular has been shown to decrease binding of α-synuclein to synthetic vesicles42 which may lead to increased oligomerization; mutation of Tyr-39 to a cysteine residue results in increased formation of α-synuclein fibrils and neurotoxicity44. Protein conformation can impact on its accessibility to peroxynitrite and subsequent 3NT modification; when α-synuclein is in a “collapsed” formation, Tyr-39 is required for aggregation but not if denaturants are present45. It is therefore quite significant that we have found this same tyrosine residue to be selectively nitrated under conditions of oxidative MAO-B overexpression and that the observed nine-fold increase could be reversed with deprenyl, a selective MAO-B antagonist. While the exact source of the nitration is difficult to identify, it has been demonstrated previously that neuronal nitric oxide synthase (NOS) levels are increased in PC12 cells following inhibition of proteasomal activity46 and H2O2 treatment of human cardiac microvascular endothelial cells increases NOS levels47. Thus increased NOS levels via increased MAO-B and subsequent H202 production could result in elevated NO and subsequent α-synuclein nitration.
PTMs on α-synuclein resulting from increases in oxidative and/or nitrosative stress may enhance formation of aggregates through stabilization of α-synuclein complex structures and/or by proteasomal inhibition48. Under physiological conditions α-synuclein is a natively unfolded protein, this state of intrinsic disorder may predispose α-synuclein to oxidative and nitrative modifications49, 50. This sensitivity coupled with the environment of the substantia nigra, where increases in oxidative stress including age-related increases in MAO-B levels may result in oxidative/nitrative modifications to α-synuclein. This in turn may lead to aggregation of α-synuclein and decreased proteasomal activity which may contribute to dopaminergic cell death in this brain region. We hope that by unraveling the mysteries surrounding the specific mechanisms contributing to α-synuclein aggregation we will better understand the neurodegeneration observed in PD.
This work was supported by National Institutes of Health Grants; RL1 NS062415 (JKA), PL1 AG032118 (BWG) and Nathan Shock Center of Excellence P30 AG025708 (BWG).