Quantitative phosphoproteomic analysis of prion-infected neuronal cells

Prion diseases or transmissible spongiform encephalopathies (TSEs) are fatal diseases associated with the conversion of the cellular prion protein (PrPC) to the abnormal prion protein (PrPSc). Since the molecular mechanisms in pathogenesis are widely unclear, we analyzed the global phospho-proteome and detected a differential pattern of tyrosine- and threonine phosphorylated proteins in PrPSc-replicating and pentosan polysulfate (PPS)-rescued N2a cells in two-dimensional gel electrophoresis. To quantify phosphorylated proteins, we performed a SILAC (stable isotope labeling by amino acids in cell culture) analysis and identified 105 proteins, which showed a regulated phosphorylation upon PrPSc infection. Among those proteins, we validated the dephosphorylation of stathmin and Cdc2 and the induced phosphorylation of cofilin in PrPSc-infected N2a cells in Western blot analyses. Our analysis showed for the first time a differentially regulated phospho-proteome in PrPSc infection, which could contribute to the establishment of novel protein markers and to the development of novel therapeutic intervention strategies in targeting prion-associated disease.

Transmissible spongiform encephalopathies (TSEs) are fatal neurodegenerative diseases occurring in many different host species including humans, which develop e.g. Creutzfeld Jacob disease (sCJD) [1]. The development of TSEs is associated with the self-propagating conversion of the normal host cellular prion protein (PrP C ) into the abnormal protease-resistant isoform (PrP Sc or PrP res ) in an autocatalytic manner [2]. PrP Sc plays a key role as an infectious agent in certain degenerative diseases of the central nervous system [3].
The cellular functions of PrP C and PrP Sc still remain enigmatic. The cellular prion protein can be variably glycosylated at two N-glycosylation sites and is C-terminally attached to the cell surface by a glycosyl phosphatidylinositol (GPI) anchor. GPI-anchored proteins are found in lipid rafts, highly cholesterol-and glycolipid-enriched membrane domains associated with a large number of signaling molecules such as G-protein-coupled receptors and protein kinases suggesting that signaling transduction pathways might play a role in TSEs [4]. Hence, previous publications described a functional role of PrP C as a signaling molecule with major findings indicating that PrP C interacts with and activates Src family kinases [5][6][7]. Increased levels of active Src kinases in scrapieinfected cells then led to the activation of downstream signal transduction pathways [8]. Recently, activation of the JAK-STAT signaling pathway in astrocytes of scrapie-infected brains was observed underlining that signal transduction pathways may play pivotal roles in prion pathogenesis [9]. Interestingly, it was demonstrated that inhibition of the non-receptor tyrosine kinase c-Abl strongly activates the lysosomal degradation of PrP Sc [10]. These data indicate that specific interference with cellular signaling pathways could represent a novel strategy in treatment of TSEs.
We have performed a quantitative analysis of the phospho-proteome to obtain a global insight into deregulated signal transduction pathways in scrapie-infected neuronal cells. We analyzed tyrosine-and threonine-phosphorylated proteins in the murine neuroblastoma cell line N2a58/22L, which were infected with the PrP Sc strain 22L [11]. We have treated N2a58/22L cells with pentosan polysulfate (PPS), a known inhibitor of 22L PrP Sc replication in N2a cells [12], resulting in the PrP Sc -rescued cell line N2a58# which served as an uninfected control. Successful rescue from PrP Sc was demonstrated in the colony assay as reflected by the absence of proteinase K (PK)-resistant PrP Sc in N2a58# cells after PPS treatment ( Figure 1A). PrP Sc replication and the effect of PPS-treatment were further studied in an immunoblot. After PK digestion, PrP Sc replication was only observed in N2a58/22L cells ( Figure 1B, lanes 2 and 4). Compared to 22L-infected N2a58/22L cells, PPS-treated N2a58# cells showed a different glycosylation profile as expected for PrP C [13][14][15]. The glycosylation pattern of PrP C in N2a58# cells displayed high amounts of di-and mono-glycosylated PrP C , whereas in N2a58/22L cells predominantly mono-and non-glycosylated PrP Sc was detected ( Figure 1B, lanes 1  and 3). Altogether, PPS treatment of N2a58/22L cells successfully abolished PrP Sc formation in N2a58# cells, which served as a non-infected control cell line in our study.
To analyze differentially phosphorylated proteins in N2a58/22L cells in comparison to N2a58# cells, we separated equal protein amounts by two-dimensional gel electrophoresis. Gels were stained with Coomassie Blue to demonstrate equal protein amounts in N2a58/22L and N2a58# cells ( Figure 1C, left panels). In parallel, gels were blotted onto membranes and incubated with phospho-specific antibodies to detect tyrosine-( Figure 1C, middle panels) or threonine-phosphorylated proteins ( Figure 1C, right panels). Interestingly, considerable differences in phosphorylation patterns were observed ( Figure 1C, asterisks), while other phosphorylated proteins were not changed in N2a58/22L and N2a58# cells ( Figure 1C). These data imply differentially regulated phosphoproteins in response to 22L infection of neuronal cells.
Generally, global detection of phosphorylated proteins is still challenging, as antisera often recognize phosphorylated residues dependent on the surrounding sequence. For a general detection of proteins post-translationally phosphorylated at those sites, we performed a SILAC analysis allowing the identification and relative quantification of differential phosphoprotein regulation. Therefore, N2a58# cells were grown in light isotope containing and N2a58/22L cells in heavy isotope containing Figure 1 Differentially phosphorylated proteins in PrP Sc -positive and -negative N2a cells. (A) PrP res -positive N2a58/22L cells were treated with pentosan polysulfate (PPS) to obtain PrP res -negative N2a58# cells. Successful PPS treatment was validated in a colony assay. Cells were grown to confluence on cover slips and directly lysed on nitrocellulose. Where indicated 20 μg/ml proteinase K (PK) was added followed by the detection of PrP expression using the 6H4 monoclonal antibody. In non-treated cells (-), PrP was detected in both, cured and infected N2a cells. Upon digestion with PK (+), PrP res was only observed in N2a58/22L cells. (B) Equal amounts of protein lysates were incubated with 20 μg/ml PK or left untreated. PrP was detected with the 8H4 monoclonal antibody showing the typical migration pattern of PrP and PrP res in infected and PPS-treated N2a58# cells. In parallel, lysates were incubated with PK to visualize PK-resistant PrP res in N2a58/22L. (C) 150 μg of N2a58# or prioninfected N2a58/22L cell lysates were separated by two-dimensional gel electrophoresis followed either by Coomassie staining or immunoblotting for detection of tyrosine-and threonine-phosphorylated proteins. Black asterisks indicate changed intensities of protein phosphorylation. medium. Equal amounts of protein lysates were mixed, separated by gel electrophoresis, trypsinized and followed by enrichment of phosphoproteins, which were then analyzed by mass spectrometry. We identified 109 different phosphoproteins of which 105 were also quantified (Tables 1 and 2). We observed 75 proteins with a ratio of identified peptides in N2a58/22L versus N2a58# cells ranging from 0.46 to 0.99 (Table 1). Conversely, 30 phosphoproteins showed a ratio between 1.01 and 1.79 (Table 2). We defined proteins exhibiting a ratio < 0.70 as dephosphorylated proteins and proteins with ratios between 0.70 and 1.40 as proteins, whose phosphorylation was not altered in 22L-infected N2a58/22L cells. Ratios > 1.40 were considered as proteins whose phosphorylation increased upon Scrapie infection.
Among quantified phosphoproteins, we then considered specific phosphosites in selected target proteins, such as Cdc2, stathmin, and cofilin as analyzed by massspectrometry (Table 3). An increase of cofilin S3 phosphorylation in N2a58/22L cells was suggested by a ratio 1.63, while the amount of the two tyrosine phosphorylation sites (Y15, Y160) in Cdc2 were decreased upon 22L infection. Stathmin phosphopeptides containing serine 38 were increased, whereas the amount of stathmin phosphopeptides harboring serine 25 in N2a58/22L cells was significantly lower (Table 3).
To validate the results obtained in the SILAC phosphoproteomic analysis we performed Western blots for cofilin 1, Cdc2, and stathmin using antibodies for the detection of specific phosphosites. As predicted by the SILAC analysis, cofilin 1 phosphorylation was significantly induced in Scrapie-infected N2a58/22L cells compared to PPS-treated N2a58# cells ( Figure 2, left panels). Cofilin represents a potent regulator of the actin filaments, which is controlled by phosphorylation of serine 3 mediated through the LIM-kinase 1 (LIMK-1) in vitro and in vivo [16]. These data support previous studies indicating a direct interaction of PrP Sc with cofilin [17]. Together with our finding that phosphorylation of cofilin is induced in PrP Sc -infected neuronal cells; the results indicate a significant role for the protein in neurodegeneration processes. Stathmin acts as an important regulatory protein of microtubule dynamics, which can be directly targeted by Cdc2 [18]. In our analysis, we showed that stathmin S38 phosphorylation was decreased ( Figure 2, middle panels), which correlates with the inactivation of Cdc2 in N2a58/22L cells ( Figure 2, right panels) implying that there is a functional interaction. Cdc2 is a crucial kinase in starting M phase events during the cell cycle progression and regulates important mitotic structure changes, including nuclear envelope breakdown and spindle assembly [19]. Dephosphorylation of stathmin S38 led to an inhibition of cells at G2/M phase, lack of spindle assembly, and growth inhibition [20,21]. Together with the finding that the prion gene is transcriptionally activated in the G1 phase in confluent and terminally differentiated cells [22], we assume that control of the cell cycle might be important in prion diseases.
Aberrant signal transduction pathways are implicated in many diseases. However, perturbations in phosphorylation-based signaling networks are typically studied in a hypothesis-driven approach. In this study, we performed the first global analysis of the phosphoproteome of scrapie-infected neuronal cells, since the knowledge of PrPdependent deregulation of the signalling network is poor. SILAC provides a powerful and accurate technique for relative proteome-wide quantification by massspectrometry. Its versatility has been demonstrated by a wide range of applications, especially for intracellular signal transduction pathways [23][24][25]. Since we applied SILAC for the quantitative detection of the phosphoproteome in scrapie-infected neuroblastoma cells, we found 105 different phosphoproteins. Among identified proteins, we validated the regulated phosphorylation of cofilin, stathmin and Cdc2 indicating that the identification of phosphoproteins in scrapie-infected neuronal cells by SILAC is reliable. Future work is necessary to determine whether the identified novel phosphoproteins are involved in prion diseases and if they probably represent sensitive and specific biomarkers for diagnosis or therapeutic intervention strategies.

Colony assay
The colony assay was performed as previously described with minor modifications [26]. In brief, cells were grown on glass cover slips to confluence using a 24 well plate. The cell layer was soaked in lysis buffer (150 mM NaCl, 0.5% Triton X-100, 0.5% DOC, 50 mM Tris pH 7.5) on a nitrocellulose membrane. After drying for 30 min at room temperature, the membrane was incubated in lysis buffer containing 5 μg/ml proteinase K (PK) for 90 min at 37°C, rinsed twice with water, and incubated in 2 mM PMSF for 10 min. The membrane was shaken in 3 M guanidinium thiocyanate, 10 mM Tris-HCl (pH 8.0) for 10 min, followed by rinsing five times with water. 5% nonfat dry milk in TBS-T was used for blocking for 1 h at room temperature. PrP was detected using an anti-PrP

SILAC phosphoproteomics analysis
SILAC ready-to-use cell culture media and dialyzed FBS were obtained from Dundee Cell Products Ltd, UK. While N2a58# cells were cultured in control SILAC DMEM media containing unlabelled arginine and lysine amino acids (R0K0), N2a58/22L cells were cultured in ready-to-use SILAC DMEM medium containing 13 C labeled arginine and lysine amino acids (R6K6) for seven cell division cycles. After preparation of cell lysates and measurement of protein concentration, lysates of N2a58# and N2a58/22L cells were mixed in a ratio 1:1.
Each sample was reduced in SDS PAGE loading buffer containing 10 mM DTT and alkylated in 50 mM iodoacetamide prior to separation by one-dimensional SDS-PAGE (4-12% Bis-Tris Novex mini-gel, Invitrogen) and visualization by colloidal Coomassie staining (Novex, Invitrogen). The entire protein gel lane was excised and cut into 10 gel slices each. Every gel slice was subjected to in-gel digestion with trypsin [27]. The resulting tryptic peptides were extracted by 1% formic acid, acetonitrile, lyophilized in a speedvac (Helena Biosciences).

Phosphopeptide enrichment
The lyophilized peptides above were resuspended in 5% acetic acid (binding buffer) and phosphopeptide enrichment was carried out using immobilized metal ion affinity chromatography (IMAC). Immobilized gallium in the Pierce Ga-IDA Phosphopeptide Enrichment Kit was used to enrich for phosphopeptides prior to MS/MS analysis according to the manufacturer's instructions (Thermo Scientific).

LC-MS/MS
Trypsin digested peptides were separated using an Ultimate U3000 (Dionex Corporation) nanoflow LC-system consisting of a solvent degasser, micro and nanoflow pumps, flow control module, UV detector and a thermostated autosampler. 10 μl of sample (a total of 2 μg) was loaded with a constant flow of 20 μl/min onto a PepMap C18 trap column (0.3 mm id × 5 mm, Dionex Corporation). After trap enrichment peptides were eluted off onto a PepMap C18 nano column (75 μm × 15 cm, Dionex Corporation) with a linear gradient of 5-35% solvent B (90% acetonitrile with 0.1% formic acid) over 65 minutes with a constant flow of 300 nl/min. Figure 2 Specific regulation of cofilin, Cdc2, and stathmin phosphorylation in scrapie-infected neuronal cells. Cell lysates of N2a58# and 22L-infected N2a58/22L cells were analyzed by Western blot using phospho-specific antibodies to detect p-cofilin S3 , p-cdc2 Y15 , and p-stathmin S38 (left panels). As loading controls, equal amounts of cofilin, Cdc2 and stathmin were shown. Quantification of intensities of phosphorylation signal was performed by normalizing the corresponding loading control (* p < 0.05) (right panels).
The HPLC system was coupled to a LTQ Orbitrap XL (Thermo Fisher Scientific Inc) via a nano ES ion source (Proxeon Biosystems). The spray voltage was set to 1.2 kV and the temperature of the heated capillary was set to 200°C. Full scan MS survey spectra (m/z 335-1800) in profile mode were acquired in the Orbitrap with a resolution of 60,000 after accumulation of 500,000 ions. The five most intense peptide ions from the preview scan in the Orbitrap were fragmented by collision induced dissociation (normalised collision energy 35%, activation Q 0.250 and activation time 30 ms) in the LTQ after the accumulation of 10,000 ions. Maximal filling times were 1,000 ms for the full scans and 150 ms for the MS/MS scans. Precursor ion charge state screening was enabled and all unassigned charge states as well as singly charged species were rejected. The dynamic exclusion list was restricted to a maximum of 500 entries with a maximum retention period of 90 seconds and a relative mass window of 10 ppm. The lock mass option was enabled for survey scans to improve mass accuracy [28]. Data were acquired using the Xcalibur software.

Quantification and Bioinformatic Analysis
Quantification was performed with MaxQuant version 1.0.7.4 [29], and was based on two-dimensional centroid of the isotope clusters within each SILAC pair.
To minimize the effect of outliers, protein ratios were calculated as the median of all SILAC pair ratios that belonged to peptides contained in the protein. The percentage variability of the quantitation was defined as the standard deviation of the natural logarithm of all ratios used for obtaining the protein ratio multiplied by a constant factor 100. The generation of peak list, SILAC-and extracted ion current-based quantitation, calculated posterior error probability, and false discovery rate based on search engine results, peptide to protein group assembly, and data filtration and presentation was carried out using MaxQuant. The derived peak list was searched with the Mascot search engine (version 2.1.04; Matrix Science, London, UK) against a concatenated database combining 80,412 proteins from International Protein Index (IPI) human protein database version 3.6 (forward database), and the reversed sequences of all proteins (reverse database). Alternatively, database searches were done using Mascot (Matrix Science) as the database search engine and the results saved as a peptide summary before quantification using MSQuant http://msquant.sourceforge. net/. Parameters allowed included up to three missed cleavages and three labeled amino acids (arginine and lysine). Initial mass deviation of precursor ion and fragment ions were up to 7 ppm and 0.5 Da, respectively.
The minimum required peptide length was set to 6 amino acids. To pass statistical evaluation, posterior error probability (PEP) for peptide identification (MS/MS spectra) should be below or equal to 0.1. The required false positive rate (FPR) was set to 5% at the peptide level. False positive rates or PEP for peptides were calculated by recording the Mascot score and peptide sequence length-dependent histograms of forward and reverse hits separately and then using Bayes' theorem in deriving the probability of a false identification for a given top scoring peptide. At the protein level, the false discovery rate (FDR) was calculated as the product of the PEP of a protein's peptides where only peptides with distinct sequences were taken into account. If a group of identified peptide sequences belong to multiple proteins and these proteins cannot be distinguished, with no unique peptide reported, these proteins are reported as a protein group in MaxQuant. Proteins were quantified if at least one MaxQuant-quantifiable SILAC pair was present. Identification was set to a false discovery rate of 1% with a minimum of two quantifiable peptides. The set value for FPR/PEP at the peptide level ensures that the worst identified peptide has a probability of 0.05 of being false; and proteins are sorted by the product of the false positive rates of their peptides where only peptides with distinct sequences are recognized. During the search, proteins are successively included starting with the bestidentified ones until a false discovery rate of 1% is reached; an estimation based on the fraction of reverse protein hits.
Enzyme specificity was set to trypsin allowing for cleavage of N-terminal to proline and between aspartic acid and proline. Carbamidomethylation of cysteine was searched as a fixed modification, whereas N-acetyl protein, oxidation of methionine and phosphorylation of serine, threonine and tyrosine were searched as variable modifications.