Brain Phosphoproteome Obtained by a FASP-Based Method

Centre, St James's University Hospital, Beckett Street, Leeds LS9 7TF, United Kingdom. ...... Marta Strollo , Maria M. Valente , Charlotte Kilstru...
3 downloads 0 Views 4MB Size
Brain Phosphoproteome Obtained by a FASP-Based Method Reveals Plasma Membrane Protein Topology Jacek R. Wis´niewski,*,† Nagarjuna Nagaraj,† Alexandre Zougman,†,‡ Florian Gnad,† and Matthias Mann*,† Department of Proteomics and Signal Transduction, Max-Planck Institute for Biochemistry, Am Klopferspitz 18, D-82152 Martinsried near Munich, Germany Received March 10, 2010

Taking advantage of the recently developed Filter Assisted Sample Preparation (FASP) method for sample preparation, we performed an in-depth analysis of phosphorylation sites in mouse brain. To maximize the number of detected phosphorylation sites, we fractionated proteins by size exclusion chromatography (SEC) or separated tryptic peptides on an anion exchanger (SAX) prior or after the TiO2-based phosphopeptide enrichment, respectively. SEC allowed analysis of minute tissue samples (1 mg total protein), and resulted in identification of more than 4000 sites in a single experiment, comprising eight fractions. SAX in a pipet tip format offered a convenient and rapid way to fractionate phosphopeptides and mapped more than 5000 sites in a single six fraction experiment. To enrich peptides containing phosphotyrosine residues, we describe a filter aided antibody capturing and elution (FACE) method that requires only the uncoupled instead of resin-immobilized capture reagent. In total, we identified 12 035 phosphorylation sites on 4579 brain proteins of which 8446 are novel. Gene Ontology annotation reveals that 23% of indentified sites are located on plasma membrane proteins, including a large number of ion channels and transporters. Together with the glycosylation sites from a recent large-scale study, they can confirm or correct predicted membrane topologies of these proteins, as we show for the examples calcium channels and glutamate receptors. Keywords: Brain • FASP;Proteomics • Ion channel • Phosphoproteomics

Introduction The organization, maintenance, and regulation of the composition of the intracellular milieu are essential for cellular homeostasis and for diverse biological functions. Ion channel and transporter proteins located in the plasma membrane play a pivotal role in establishing this homeostasis. In particular, cell types of the brain have highly specialized functions in neural signal transmission and they are therefore equipped with a large number of specialized plasma proteins regulating ion transport, neurotransmitter uptake and release, and related functions. These activities are frequently regulated by phosphorylation on specific cytoplasmic sites (for recent examples see refs 1-3). So far, these phosphorylation sites have generally been mapped by classical biochemical methods, which have many technological limitations, are very laborious, and often require overexpression of plasma membrane proteins in heterologous expression systems. As a result of these limitations, our knowledge of in vivo phosphorylation sites of brain membrane proteins is still very incomplete. * To whom correspondence should be addressed. E-mail: jwisniew@ biochem.mpg.de (J.R.W.) or [email protected] (M.M.). Fax: +49 89 8578 2219 . † Max-Planck Institute for Biochemistry. ‡ Present affiliation: Cancer Research UK Clinical Centre, St James’s University Hospital, Beckett Street, Leeds LS9 7TF, United Kingdom.

3280 Journal of Proteome Research 2010, 9, 3280–3289 Published on Web 04/26/2010

Mass spectrometry based proteomics should be ideally suited to map in vivo phosphorylation sites in the brain. Accordingly, during the past years much work has been invested in the proteomic mapping of brain proteins and their modifications, including large-scale phosphoproteome analysis. An early study provided information on 460 phosphorylation sites in developing brain.4 Subsequent large-scale work identified 1564 phosphorylation sites in postsynaptic density preparations from murine cortex, midbrain, cerebellum, and hippocampus,5 974 phosphorylation sites on 499 proteins in synaptosomal preparations,6 and 512 phosphorylation sites in 162 cytosolic brain proteins.7 In another study, using an immunoprecipitation based approach, 414 phosphotyrosine sites in 302 proteins were found in Swiss Webster mouse brain.8 Recently, using subcellular fractionation, 2082 unique phosphopeptides were identified in 1062 phosphoproteins.9 A particular challenge for the proteomic analysis of brain proteins is the fact that a large percentage of crucial proteins are embedded in plasma membranes. Standard preparation methods for mass spectrometry are usually not optimal for analyzing this protein class. Recently, we have developed the Filter Aided Sample Preparation (FASP) method, which enables characterization of membrane proteins to the same extent as soluble proteins.10 FASP has already found applications not only for unbiased identification of membrane proteins,10,11 but also in mapping of phosphorylation12,13 and glycosylation 10.1021/pr1002214

 2010 American Chemical Society

Brain Phosphoproteome Reveals Plasma Membrane Protein Topology 14

research articles

sites. Furthermore, the method is easily scalable to large (several mg) and submicrogram sample quantities.11,15

0.2% (w/v) SDS. Fractions were collected between 10.4 and 21.6 mL of elution volume.

Occupancy of the majority of phosphorylation sites varies during the cell cycle,16 during development, and between organelles in a cell. For example, mainly postmitotic tissues such as the brain are expected to show a lower degree of phosphorylation than dividing cells in culture. Exceptions to the rule of relatively low stoichiometry of phosphorylation sites are constitutively phosphorylated sites that confer conformation and stability to proteins and that therefore have 1:1 phosphate/site stoichiometry.17 The majority of phosphorylation sites are nevertheless of low abundance and their detection requires extensive enrichment and consequently relatively high amounts of starting material. This presents challenges in the application of phosphoproteomics to the elucidation of biological questions, because, unlike cultured cells, organ and tissues specimens are often available only in minute amounts. The most frequently used protocols for phosphoproteome analysis involve two basic steps:18 the SCX-chromatography based fractionation of peptides and the enrichment of phosphopeptides either using IMAC19,20 or TiO2-beads21,22 protocols.

Detergent Removal and Protein Digestion: Size Exclusion Fractions. Proteins in SEC fractions were concentrated in 30k Microcon filtration devices (Millipore) to a final volume of 30 µL and were then processed by the FASP procedure II.10 Briefly, the concentrated samples were mixed with 0.2 mL of 8 M urea in 0.1 M Tris/HCl pH 8.5 (UA), loaded into the filtration devices, and centrifuged at 14 000g for 15 min. The concentrates were diluted in the devices with 0.2 mL of UA solution and centrifuged again. After centrifugation, the concentrates were mixed with 0.1 mL of 50 mM iodoacetamide in UA solution and incubated in darkness at room temperature for 30 min. Following a centrifugation for 15 min, the concentrate was diluted with 0.2 mL of UA solution and concentrated again. This step was repeated twice. Next, the concentrate was diluted with 0.1 mL of 40 mM NaHCO3 and concentrated again. This step was repeated once. Subsequently, 2 µg of trypsin in 30 µL of 40 mM NaHCO3 was added to the filter and the samples were incubated at 37 °C overnight. The peptides were collected by centrifugation of the filter units followed by two additional 30 µL washes with 40 mM NaHCO3. The concentration of peptides was determined by UV-spectrometry using an extinction coefficient of 1.1 for 0.1% (g/L) solution at 280 nm.

Here, we have used two complementary approaches for analyzing brain phosphoproteomes from small amounts of tissue. Both strategies use FASP for sample digestion and TiO2enrichment but differ in the fractionation strategy used to reduce sample complexity. In the first approach, we use SEC to separate proteins based on their size, and in the second, we use SAX chromatography for peptide fractionation. This second approach also includes the filter based capturing of phosphotyrosine peptides using anti-phosphotyrosine antibodies. Our study provides the most in-depth mouse brain phosphoproteome data set identified so far, providing precise localization information on 12 035 phosphorylation sites on 4579 brain proteins. We also show that our large-scale and high accuracy data can be useful in exploring the function and topology of brain proteins.

Materials and Methods Tissue Lysates. Brain tissue from C57BL/6 mouse was lysed in 2% SDS, 0.1 M DTT in 0.1 M Tris/HCl, pH 7.6. The ratio of buffer to tissue was 10:1. Lysis was facilitated by homogenization using an IKA Ultra Turbax blender at maximum speed for 10 s. After 3 min incubation in boiling water, the suspensions were sonicated using Branson SONIFIER 250 (Heinemann, Germany) for 20 s (output control 3.5; duty cycle 20%). Then, the mixture was incubated at 100 °C for 3 min. The crude extract was clarified by centrifugation at 16 000g at 30 °C for 10 min. The protein content was determined by measurements of tryptophan fluorescence as described previously.23 Briefly, 1 µL of sample or tryptophan standard (100 ng/µL) was mixed with 3 mL of 8 M urea in 20 mM Tris/HCl pH 7.6. Fluorescence was measured at 295 nm for excitation and 350 nm for emission. The slits were set to 10 nm. Static quenching was not observed under these measurement conditions. Protein Fractionation by Size Exclusion Chromatography (SEC). Solid iodoacetamide was added to protein lysates containing 1 mg of protein to a final concentration of 0.2 M. After 30 min incubation, the samples were centrifuged at 130 000g at 30 °C for 15 min. Then, 0.1 mL of the supernatant was loaded onto a Superdex 200 10/300 GL (GE Bioscience) column equilibrated with 25 mM Tris-HCl, pH 8.0, 0.1 M NaCl,

Detergent Removal and Protein Digestion: Large-Scale Preparation. The brain lysate (see above) from 100 mg of tissue (∼8 mg total protein) was mixed with 8 mL of UA buffer and loaded into an Amicon Ultra 15 Ultracel 30k (Millipore) device. After about 10-fold concentration at 4000g, 10 mL of UA was added to the device and the sample was concentrated again. Then, the concentrate was mixed with 2 mL of 50 mM iodoacetamide in UA solution and incubated in darkness at room temperature (RT) for 30 min followed by centrifugation for 15 min. Then, the concentrate was diluted with 10 mL of UA solution and concentrated again. This step was repeated twice. Following two washes with 10 mL of 40 mM NaHCO3, 100 µg of trypsin in 1 mL of 40 mM NaHCO3 was added to the filter. The sample was digested overnight at 37 °C and peptides were collected by centrifugation. To increase the yield of peptides, the filter was washed twice with 0.5 mL of 40 mM NaHCO3. Peptides were desalted on Strata C18E cartridges (0.55 µm; 70 Å, Phenomenex) using water/70% (v/v) CH3CN in 0.1% (v/v) CF3COOH system. The binding capacity of a 200 mg cartridge was estimated to be 1.0-1.5 mg of peptide mixture. Concentration of the peptides was determined as described above. Filter Based Affinity Capturing and Elution (FACE) for Enrichment of Phosphotyrosine Containing Peptides. Four milligrams of peptide was dissolved in 1.2 mL of 50 mM Tris/ HCl pH 8.0 containing 100 mM NaCl (PY-Buffer) The insoluble material was removed by centrifugation at 16 000g for 10 min and the supernatant was incubated with 50 µg of 4G10 antiphosphotyrosine antibodies (Millipore) at 4 °C overnight. Antibody-peptide complexes were concentrated in 50k Microcon filtration devices to a final volume of ∼20 µL. The flow through fraction was subjected to TiO2-based enrichment of phosphopeptides and SAX fractionation as described below. The concentrate was diluted with 200 µL of PY-Buffer and the sample was concentrated again to a final volume of ∼20 µL. This step was repeated 4 times. Following two dilution/ concentration washes with 50 µL of 40 mM NH4HCO3, the antibody bound peptides were eluted with 100 µL of 0.1% (v/ Journal of Proteome Research • Vol. 9, No. 6, 2010 3281

research articles v) CF3COOH. Peptides were concentrated to a final volume of about 4 µL and analyzed by LC-MS/MS. TiO2-Based Enrichment of Phosphopeptides: After Size Exclusion Fractionation. The protein digest were acidified with 0.1% CF3COOH and phosphopeptides were enriched on TiO2beads according to refs 22, 24 with minor modifications.12 Briefly 25 mg of ‘Titansphere TiO2 10 µm’ (GL Sciences, Inc., Japan) was suspended in 50 µL of 3% (m/v) dihydroxybenzoic acid in 80% (v/v) CH3CN, 0.1% CF3COOH and diluted 1:4 with water before use. Ten microliters of this slurry (1 mg beads) was added to each sample and incubated under continuous agitation for 20 min. Then, the titanium beads were sedimented by centrifugation at 5000g for 1 min and the supernatants were collected and mixed with another portion of the beads and incubated as above. The bead-pellets were resuspended in 150 µL of 30% (v/v) CH3CN containing 3% (v/v) CF3COOH and were transferred to a 200 µL pipet tip plugged with one layer of glass microfiber filter GFA (Whatman). The beads were washed 3 times with 30% (v/v) CH3CN, 3% CF3COOH (v/v) solution and 3 times with 80% CH3CN (v/v), 0.3% CF3COOH (v/v) solution. Finally, the peptides were eluted from the beads with 100 µL of 40% CH3CN (v/v) containing 15% NH4OH (m/v) and were vacuum-concentrated to ∼4 µL. TiO2-Based Enrichment of Phosphopeptides: Before SAX Fractionation. The anti-phosphotyrosine-antibody unbound peptides (about 4 mg) (FACE flow-through fraction, see above) were acidified with CF3COOH collected and were mixed and incubated with 75 µL (7.5 mg beads) of the Titansphere-bead slurry (see above) for 20 min. After incubation, the beads were washed and the phosphopeptides were eluted as described above. The enrichment was repeated once using the unbound fraction. The phosphopeptide-enriched fractions were vacuumdried. Anion Exchange-Based Fractionation of Phophopeptides. Peptides were dissolved in 200 µL in Britton & Robinson buffer composed of 20 mM CH3COOH, 20 mM H3PO4, 20 mM H3BO3, and NaOH, pH 11. The peptides were separated by a pipetbased anion exchanger method.11 Briefly, the pipet based column was assemble by stacking 6 layers of a 3 M Empore Anion Exchange disk (Varian, 1214-5012) into a 200 µL micropipet tip. For column equilibration and elution of fractions, Britton & Robinson buffer composed of 20 mM CH3COOH, 20 mM H3PO4, and 20 mM H3BO3 titrated with NaOH to the desired pH was used. Peptides were loaded at pH 11 and fractions were subsequently eluted with buffer solutions of pH 8, 6, 5, 4, and 2, respectively. In the four-step fractionation experiment, peptides were loaded in a buffer of pH 6 and were eluted at pH 5, 4, and 2, respectively. LC-MS/MS Analysis of Peptides. Samples were separated over 15 cm i.d., 75 µm reversed-phase capillary emitter columns using 240 min gradients as described before.10 The MS/MS analysis was carried out on an LTQ-Orbitrap instrument (Thermo Fisher Scientific). Survey MS scans were acquired in the orbitrap analyzer with resolution of 60 000. For accurate mass measurements, the lock-mass option was employed.25 Up to the 10 most intense ions in each full MS scan were fragmented and analyzed in the linear ion trap part of the instrument. Raw MS files were processed with MaxQuant (version. 1.0.13.13), a freely available software suite.26 Peak list files were searched by the MASCOT search engine27 against the IPI-mouse (version 3.46) containing both forward and reversed protein sequences. Initial maximum precursor and fragment mass deviations were set to 7 ppm and 0.5 Da, 3282

Journal of Proteome Research • Vol. 9, No. 6, 2010

Wis´niewski et al.

Figure 1. Workflow for streamlined and sensitive phosphoproteome analysis. (A) Analysis of a 10 µL piece of mouse brain with size exclusion chromatography fractionation of proteins. (B) Analysis of a 100 µL piece of mouse brain with pipet-based SAX fractionation of peptides and with filter based capture and elution (FACE) of peptides containing phosphotyrosine.

respectively, but MaxQuant achieved sub-parts per million (subppm) mass accuracy for the majority of peptide precursors. The search included variable modifications for oxidation of methionine, protein N-terminal acetylation, and phosphorylation of serine, threonine, and tyrosine. Peptides with at least six amino acids were considered for identification. The false discovery rates for both peptides and proteins were set at 0.01. All phosphopeptides identified in this study are listed in Supplementary Table 1.

Results Size Exclusion Chromatography-Based Phosphoproteomics. Size exclusion chromatography is a powerful technique for protein separation allowing fractionation by size. In contrast to gel electrophoresis, which is widespread in proteomics, gel filtration can easily be scaled up to large starting amounts. Similarly to electrophoresis, it tolerates high concentrations of denaturants making it suitable to the fractionation of membrane proteins. Despite these features, there are relatively few reports of size exclusion chromatography in proteomic investigations.28-31 Here, we used the Superdex 200 column to separate SDS solubilized brain proteins. A total of 100 µL of brain lysates containing about 1 mg of total protein was fractionated into 7 or 8 fractions (Figure 1) and each fraction

Brain Phosphoproteome Reveals Plasma Membrane Protein Topology

Figure 2. Identification of phosphopeptides in chromatographic fractions. (A) Size exclusion chromatography approach; (B and C) pipette-based SAX approach with 6 and 4 fractions, respectively. The bars indicate the number of peptides identified in a single fraction by a single LC-MS/MS analysis. Data from two independent experiments. (D) Overlap between proteins identified with the SEC and SAX methods.

was processed by the FASP procedure yielding pure mixtures of tryptic peptides. Phosphopeptides were enriched with TiO2beads and analyzed by LC-MS/MS on an LTQ-Orbitrap. Because the optimal ratio of peptide to bead in the TiO2 phospho-enrichment experiments is difficult to determine beforehand,32 each of the fractions was extracted with the beads for a second time. Analysis of the latter fractions yielded 3376 phosphorylation sites of which 32% were not detected after the first extraction. A single chromatographic fraction analyzed in a single LC-MS/MS run allowed identification of 1000-2700 nonredundant phosphosites (Figure 2A). The same phosphosite was identified on average in 2.1 fractions. In total, in a single analysis using 1 mg of total protein (10-20 mg of wet tissue), about 4000 phosphosites were identified in 7-8 fractions (Supplemental Table 1). Combining the results from three independent experiments (analyzed together in MaxQuant), we identified 8014 phosphorylation sites in 2748 proteins (Table 1). Anion Exchange Chromatography-Based Phosphoproteomics. Ion exchange chromatography is frequently chosen as a separation method for complexity reduction in proteomic

research articles

studies, usually in a strong cation exchange (SCX) format. Less frequently strong anion exchange chromatography has been used.20,33 Recently, we have described a pipet-based strong anion exchange technique for separation of peptides and have shown that it is a simple and robust way to fractionate minute amounts of peptides.11 Here, we used this pipet tip-based SAX column to fractionate phosphopeptide-enriched fractions either into 4 or 6 fractions. We identified 1601 ( 273 unique phosphosites per single SAX fraction (Figure 2B,C). In duplicate analyses, we mapped 4638 unique phosphorylation sites (4308 phosphopeptides) in a 4-fraction experiment (Table 1). Seventyfive percent of these sites were identified in the duplicate 6-fraction experiment (Supplemental Figure 1). To achieve even greater depths of analysis, we employed sixfraction-protocol and also performed the TiO2 enrichment once more with the peptides that were not bound in the first enrichment step. These peptides were fractionated in the same way as the peptides captured by the first incubation (Figure 2C). This re-extraction experiment added about 30% new phosphorylation sites. In total, in the duplicate experiment comprising 24 LC-MS/MS runs, this procedure yielded 8565 unique brain phosphorylation sites. Dihydroxybenzoic acid (DHB) is frequently used to reduce binding of nonphosphopeptides to TiO2 beads.22 Although it is highly efficient, it can cause practical difficulties because it is also a contaminant adding chemical noise to mass spectrometric analysis. Remarkably, we found that under our fractionation conditions DHB bound irreversibly to the SAX membranes eliminating this problem and facilitating identification of phosphopeptides. Filter-Based Capture and Elution of Phosphotyrosine Containing Peptides. The number of tyrosine phosphorylation sites is about 2 orders of magnitude lower than that of serine and threonine phosphorylation sites. For this reason, it is useful to employ higher amounts of starting material and affinity enrichment for the identification of phosphotyrosine sites. Accordingly, we prepared lysate from 100 mg of brain (about 8 mg of total protein) that were processed using the FASP procedure10 in 15 mL concentrators. From this amount of sample, we obtained 4-5 mg of peptides. For the affinity enrichment, we developed the Filter Aided Capture and Elution (FACE) protocol, in which the affinity reagent is loaded on top of the FASP filter and binds to cognate peptides, whereas a majority of peptides are washed through the filter. Importantly, the antibody recognizing the desired peptide class (monoclonal phosphotyrosine antibody 4G10 in this case), only needs to be added in solution on top of the filter and does not need to be coupled to beads. The antibody-peptide complexes were retained on 50k Microcon filtration devices and the bound peptides were eluted with 0.1% CF3COOH directly for LC-MS/ MS analysis. Two single LC-MS/MS runs allowed identification of 174 unique phosphotyrosine sites. Comparison of SEC and SAX Based Phosphoproteomics. SEC is frequently considered a low resolution fractionation technique, but in fact, separation by this technique has a resolution comparable with that of ion exchange chromatography. As an alternative to SEC, preparative electrophoresis could in principle be used. However, whereas that technique is superior for the separation of small proteins, its performance for larger proteins is rather weak in our experience. Here, we show that SEC is a useful tool for prefractionation of proteins prior to protein digestion and phospho-enrichment. It is not only suitable for small amounts of sample, but can easily be Journal of Proteome Research • Vol. 9, No. 6, 2010 3283

research articles

Wis´niewski et al.

Table 1. Summary of the Phosphoproteome Profiling of the Mouse Brain size exclusion ‘GO’-annotation

phosphosites

Total Identified

8014

‘GO’-Annotated Plasma membrane Integral to membrane Nucleus Mitochondrion Golgi Endoplasmic reticulum

6630 1513 1237 1282 220 369 255

anion exchanger 6 fractions %

phosphosites

23 19 19 3.3 5.6 3.8

7063 1666 1389 1343 228 352 287

scaled up for bulk separation. Moreover, it provides valuable protein size information, which, even if it is only approximate, can be useful in data interpretation. Pipette based SAX fractionation of phosphopeptides, on the other hand, has several important practical advantages. It does not require any special instrumentation for chromatography of peptides and the entire procedure can be accomplished within 30 min. In addition, the phosphopeptide-enrichment is reduced to one or two incubations (if a reincubation is desired). TiO2 enrichment is performed at a stage of sample preparation where peptide content can be more reliably determined. This, in turn, can facilitate optimization of the phospho-enrichment step. Taking these observations together, the “pipette SAX” approach appears to be more convenient and faster than the SEC based one. Importantly, it can easily be coupled to the phosphotyrosine affinity enrichment step described above (FACE method). The entire analysis of brain phosphorylation sites consisted of data from 61 LC-MS/MS runs of fractions obtained in three SEC separations of proteins and six SAX chromatographic fractionations of peptides. Combined analysis of the data using the MaxQuant software revealed 16 296 events of phosphorylation in mouse brain proteins (Supplementary Table 1). Out of those, 12 035 were identified with a site localization probability of g0.75 (median value 0.999 and average 0.965) (Supplementary Figure 2A) and only these sites are presented and discussed in this work (Supplementary Table 1). A distribution of the 5490 of these sites were identified in both SAX and SEC experiments, whereas 4343 were identified only in SAX and 1963 only in SEC (Figure 2D). Features of the Brain Phosphoproteome. Our study detected 12 035 phosphorylation sites in 4579 proteins and 70% of the sites were identified in at least two experiments with very high average localization probability (Supplementary Table 1). This outnumbers the data sets of other studies on the mouse brain by a factor of 10 and should therefore be an excellent resource to derive features that are characteristic for the whole mouse brain. Overall, 70% of our phosphosites have not been reported in mouse before according to Uniprot annotation (version 55.2) The distribution of phosphorylated serines, threonines, and tyrosines was calculated to be 83%, 14.7%, and 2.3%, respectively (Supplementary Table 1), which is very close to the distributions in other eukaryotic organisms. However, even in the absence of any external stimulus, the tyrosine phosphoproteome is highly represented, due to the FACE enrichment method that was applied in our study. For the 4579 phosphoproteins detected, 11 437 nonredundant phosphopeptides were identified. Overall, 96% of the phosphorylated peptides had Mascot identification scores higher than 15. Although peptides were initially identified by 3284

%

8565

Journal of Proteome Research • Vol. 9, No. 6, 2010

anion exchanger 4 fractions phosphosites

%

4638 24 20 19 3.2 5.0 4.1

3827 936 727 708 126 163 123

all approaches together phosphosites

%

12035 24 19 19 3.3 4.3 3.2

9897 2270 1890 1938 309 516 385

23 19 20 3.1 5.2 3.9

the Mascot algorithm, final scoring is done in MaxQuant as described.26 The distribution of Mascot and PTM scores of identifiedphosphopeptidesisshowninSupplementaryFigure2B,C. A majority of the identified phosphopeptides were singly phosphorylated (88%) and 10% were doubly phosphorylated (Supplementary Table 1). Bioinformatic Analysis. Using CytoScape and BinGO34,35 for gene ontology analysis, we found that the identified phosphoproteins were distributed broadly in organelles of brain cells (Supplementary Table 2). Cell compartments that are specific to nerve cells including axon, dendrite, and synapse were significantly overrepresented compared to the entire IPI mouse database (Supplementary Figure 3). Interestingly, membrane proteins were also significantly overrepresented in our phospho-data set (Supplementary Figure 3, Supplementary Table 2). In contrast, in our previous large-scale phosphoproteomics study of HeLa cells, which did not apply the FASP method, we had found underrepresentation of this class of protein.24 Functional analysis revealed that 190 protein kinases and 65 phosphatases were found to be phosphorylated in our set, which presents a significant enrichment (p ) 6.7 × 10-14 and p ) 2.3 × 10-5, respectively). Matching mouse phosphosites with annotated kinase motifs showed that the identified phosphoproteins were putative substrates of a variety of kinases including PKA, CDK1, and Aurora-A (Supplementary Table 3). The number of mouse phosphosites that matched with the CDK2 motif, for example, was more than 12 times higher than would be expected by chance. In agreement with previous studies,7,36,37 phosphosites were predominantly located in loops and turns on the protein surface (Supplementary Figure 4) according to the prediction method SABLE 2.0.38 Overall, 92% of phosphorylated serines are located in loops in comparison to 75% of their nonphosphorylated counterparts. Using the EnsEMBL Compara database,39 we found that phosphorylated proteins in our data set have more orthologs than nonphosphorylated proteins (Supplementary Table 4), which is in agreement with our previous phosphoproteomic studies.36 For example, 87% of the phosphorylated mouse brain proteins have one-to-one orthologs in chimp in comparison to 61% of all other mouse proteins. Moreover, the proportion of nonsynonymous to synonymous changes in genes that encode phosphorylated proteins was lower than the one of triplets which encode nonphosphorylated proteins. Determining Plasma Membrane Protein Topology by PTM Analysis. According to GO annotation, 23% of the identified phosphorylation sites are located on plasma membrane proteins including many brain specific proteins such as channels and transporters (Table 2). Interestingly, more than 90% of the identified sites on these proteins are not recorded by Swiss-Prot40 (version 55.2), reflecting previous analytical

Brain Phosphoproteome Reveals Plasma Membrane Protein Topology Table 2. Phosphorylation Sites on Major Groups of Channel and Transporter Proteins number of phosphorylation sites gene/protein family on all subunits

Glutamate receptors Ionotropic NMDA Metabotropic delta GABA receptors Type A Type B ABC transporters

6 32 25 13

Gria1,2 and 4 Grin2a and b Grm1,2,3,5 and 7 Grid 1 and 2

8 9 16

Gabbr1 and 2, Gabra1,3 and 6 Abca1; Abca2; Abcb1a; Abcc1; Abcc5; Abcd2; Abcd3; Abce1 Slc1a2; Slc4a4; Slc4a10; Slc6a17; Slc7a14; Slc9a1; Cacna1a-e; Cacnag; Cacnai; Cacnb1-4; Cacng Scn1a; Scn2a1; Scn2b; Scn3b; Scn8a; Scn9a Kcna1,2,4 and 6; Kcnb1and 2; Kcnc1-4, Kcnj3,6,10,and 11;Kcma1; Kcnq2,3 and 5

SLC transportersa

290

Calcium channels

94

Sodium channels

38

Potassium channels

a

examples (gene names)

135

Not all slc transporters are located at the plasma membrane.

difficulties in analyzing phosphorylation sites on this important class of proteins. Membrane proteins typically span the membrane one or more times and the hydrophobic, membrane-spanning regions are relatively easily predicted by bioinformatics. Membrane topology, which parts of the protein face the cytosolic and extracellular region of the cell, is more difficult to determine either experimentally or by bioinformatics. However, this knowledge is crucial for studying the function of membrane proteins. The Swiss-Prot database contains topology information for membrane proteins, in almost all cases derived from bioinformatic prediction. We mapped our phosphorylation sites onto Swiss-Prot protein topology information to determine the location of phosphorylation sites in plasma membrane proteins. For 1076 sites, we found Swiss-Prot protein topology information and the analysis revealed that 95% of phosphorylation events localized to predicted cytosolic domains and 5% to predicted extracellular locations (Supplementary Table 5). Almost all of the latter represent sites on proteins for which little experimental evidence exists, and therefore, we suspected that the predicted extracellular locations are in fact cytosolic. To shed some light on the correctness of the topology prediction, we used the TMHMM program41 to predict the topology of GABAA3 subunit, sodium channel Nav1.2, and solute carrier family 12 member 5. For the GABA receptor subunit, the phosphorylation sites are located in the cytoplasmic region of the protein as expected (Figure 3A). For independent verification of membrane topology, we also mapped N-glycosylation sites on this protein (obtained from a recent large-scale study).14 N-glycosylation is restricted to the extracellular part of plasma membrane proteins, and therefore, these sites also contain information on the actual protein topology. Again, as expected, N-glycosylation sites mapped to the predicted extracellular part, verifying our approach. For sodium channel Nav1.2, the situation was exactly opposite: All phosphorylation sites of this protein with many

research articles

predicted extracellular and intracellular loops occurred on predicted extracellular parts, whereas all N-glycosylations occurred on predicted intracellular parts (Figure 3B). This demonstrates that the predicted structure is incorrect and suggests that it is simply flipped between intra and extracellular parts, a typical error of prediction. The situation was more complex for solute carrier family 12 member 5 (Swiss-Prot entry number Q91V14-2). In neither the Swiss-Prot nor the TMHMM predictions did the phosphorylation or glycosylation sites locate to the correct compartment, showing that both models were incorrect (Figure 3C). These examples illustrate some of the difficulties in predicting membrane topology and demonstrate that large-scale PTM analysis can verify or disprove suggested structures. Importantly, it is not necessary to obtain exhaustive information because errors in one segment of the protein are transmitted to the next. Locating a few PTMs with high accuracy can provide ‘fixed points’ that the membrane structure model needs to be consistent with, greatly improving the chances that the overall structure is correct. Several previous proteomic studies already shown that phosphoproteomics data can be used as empirical evidence for validating the topology of transmembrane proteins.42,43 Interestingly, for 25% of plasma membrane proteins from Arabidopsis, the cytosolic protein portions had been incorrectly assigned by computational predictions.42 Phosphorylation of Brain Specific Complexes. Integral membrane proteins frequently associate with each other or with cytoplasmic proteins in specific and functional complexes. For example, high voltage activated calcium channels are composed of the pore forming 1R subunit,44 membrane anchored extracellular R2δ subunit,45,46 and the cytoplasmic β-subunit.2 Our results revealed extensive phosphorylation of the cytoplasmic parts of the 1R and the β subunit. Modifications on the 1R subunit were found in the N- and C-terminal polypeptides and the inter-repeat loops I/II and III/IV (Figure 4). From our phosphorylation data, we were able to map up to 13 phosphorylation sites (Cav2.1) to the C-terminal polypeptides and up to 11 to the II/III linker of Cav2.2. Phosphorylation sites on β-subunit are located in the sequences flanking the SH3 and GK domains and could therefore modulate binding mediated by them (Figure 4). Interestingly, these sites conform to Calmodulin kinase II and PKA motifs, congruent with the known involvement of these kinases in regulation of the calcium channel.47,48 The majority of the phosphorylation sites on 1R and β-subunits match consensus motifs of a number of different kinases (Supplemental Table 1). It is possible that these kinases are also involved in regulation of this family of calcium channel. Other examples of complexes formed by plasma membrane proteins are the metabotropic glutamate receptors (mGlurs). These proteins influence a variety of second messenger systems that modulate neuronal excitability, synaptic plasticity, and are involved in neurodegeneration. They are multipass transmembrane proteins with an about 600 residue long extracellular N-terminal part and a cytoplasmic C-terminal region of variable length (Figure 5A). The C-terminal tails have known binding sites for a number of proteins including receptors, kinases, and scaffold proteins.49 Past work already provided evidence that mGlur is phosphorylated,50 but our work now identifies more than 20 specific phosphorylation sites located in the tails of different mGlurs. Of potential functional importance, we mapped several phosphorylation sites to epitopes involved in Journal of Proteome Research • Vol. 9, No. 6, 2010 3285

research articles

Wis´niewski et al.

Figure 3. PTM data can improve topology prediction of plasma membrane proteins. (A) Gabra3, Gamma-aminobutyric acid (GABA-A) receptor, subunit a3; (B) Scn2a1, Sodium channel Nav1.2; (C and D) Slc12a5, Solute carrier family 12 member 5. Phosphorylation (this work) and N-glycosylation sites14 are indicated by circles and triangles, respectively. (A and D) Protein topology according to SwissProt; (B and C) TMHMM predictions.

binding of Cav2a, Calmodulin, Filamin A, β tubulin, Homer, and Tamalin (Figure 5B,C). Phosphorylation of these epitopes may modulate binding to cognate interaction partners. Since up to 9 different sites were found on a single receptor (mGlur1; Figure 5A), the phosphorylation of mGlurs may play an important role in their regulation.

Conclusions and Outlook Brain cells contain high numbers of membrane and plasma membrane proteins that have unique functional roles in this organ. Analysis of the brain membrane proteome in general and the brain phosphoproteome in particular have presented great analytical challenges leaving us with a fragmented picture of this important modification. The FASP-based methods employed here ensured complete solubilization of the brain membrane proteome, which is essential for unbiased analysis.51,52 Lysis of samples in buffer containing high concentration of SDS 3286

Journal of Proteome Research • Vol. 9, No. 6, 2010

is not only advantageous for analysis of membrane proteins, but also efficiently inactivates degradation of proteins and other enzymatic activities such as dephosphorylation, serving in effect as powerful protease and phosphatase inhibitor. The FASP sample preparation method was combined with either SECbased protein fractionation or SAX-based peptide fractionation. These methods both have unique features and appear to provide somewhat complementary results. However, the SAXbased peptide fractionation method is more practical. It is also readily combined with a peptide affinity step on top of the FASP filter (that we term FACE), which can be used to capture glycopeptides,14 or as shown here, tyrosine phosphorylated peptides. Other solution based affinity enrichment steps, including nucleic acid based ones, could readily be coupled with FASP in this way. Applying the above methods, we identified more than 16 000 phosphorylation events and mapped more than 12 000 phos-

Brain Phosphoproteome Reveals Plasma Membrane Protein Topology

research articles

Figure 4. Subunits 1R and β of the high voltage gated calcium channel are extensively phosphorylated. (A) Scheme of the organization of the constitutive subunits of the channel with distribution of the measured phosphorylation sites. (B and C) Identified phosphorylation sites mapped to β and 1R subunits are listed in B and C, respectively. Red and blue fonts refer to sites with consensus sequence of CamK2 and PKA.

Figure 5. Phosphorylation sites mapped to C-terminal tails of the metabotropic glutamate receptors. (A) Scheme of the topology of mGlurs. Nine phosphorylation sites identified in mGlur1 are indicated. Phosphorylation sites identified in mGlur1 and mGlur5 (B), and mGlur2 and mGlur3 (C). The phosphorylated residues are highlighted. Binding sites of Cav2.1, Calmodulin, Filamin A, β Tubulin, Homer and Tamalin are marked by colored lines.

Journal of Proteome Research • Vol. 9, No. 6, 2010 3287

research articles phorylation sites with high localization probabilities, providing an in-depth map of phosphorylation sites in the mouse brain proteome. Importantly, 23% of the phosphorylation sites are located in plasma membrane proteins, many on brain specific proteins that are integral to the plasma membrane. We already singled out one promising application of the brain phosphoproteome, namely, the determination of membrane protein topology. Although only described for selected examples, our data already contain a large number of sites that point to incorrect structure models in the Swiss-Prot database. Dedicated efforts can likely greatly expand the number of topological ‘fixed points’ for membrane proteins. Our largescale phosphorylation data, in combination with glycosylation data, can therefore be used for improved prediction of membrane protein topology. Detailed mapping of phosphorylation in individual integral plasma membrane proteins is a subject of many recent studies (see, for example,1, 53, 54). This indicates the potential impact and usefulness of large-scale studies providing information on in vivo modification status of thousands of proteins. Such a resource of phosphorylation sites, combined with bioinformatic analysis and literature knowledge, should inspire and facilitate functional neurobiological studies. Abbreviations: FASP, filter aided sample preparation; FACE, filter aided capture and elution.

Acknowledgment. The work was supported by MaxPlanck Society for the Advancement of Science, Munich Center for Integrated Protein Science (CIPSM), and PROSPECT, a 7th framework program of the European Union. Supporting Information Available: Supplementary Figure 1, overlap between phosphopeptides identified with the pipette-based SAX approach with 6 and 4 fractions; Supplementary Figure 2, distribution of the site localization probabilities, Mascot scores, and PTM scores of the mapped phosphoproteome; Supplementary Figure 3, gene ontology analysis; Supplementary Figure 4, structure analysis; Supplementary Table 1, the mouse brain phosphoproteome; Supplementary Table 2, gene ontology analysis; Supplementary Table 3, motif analysis; Supplementary Table 4, conservation analysis. This material is available free of charge via the Internet at http://pubs.acs.org. References (1) Rinehart, J.; Maksimova, Y. D.; Tanis, J. E.; Stone, K. L.; Hodson, C. A.; Zhang, J.; Risinger, M.; Pan, W.; Wu, D.; Colangelo, C. M.; Forbush, B.; Joiner, C. H.; Gulcicek, E. E.; Gallagher, P. G.; Lifton, R. P. Sites of regulated phosphorylation that control K-Cl cotransporter activity. Cell 2009, 138 (3), 525–36. (2) Dolphin, A. C. Beta subunits of voltage-gated calcium channels. J. Bioenerg. Biomembr. 2003, 35 (6), 599–620. (3) Yukutake, Y.; Hirano, Y.; Suematsu, M.; Yasui, M. Rapid and reversible inhibition of aquaporin-4 by zinc. Biochemistry 2009, 48 (51), 12059–61. (4) Ballif, B. A.; Villen, J.; Beausoleil, S. A.; Schwartz, D.; Gygi, S. P. Phosphoproteomic analysis of the developing mouse brain. Mol. Cell. Proteomics 2004, 3 (11), 1093–101. (5) Trinidad, J. C.; Thalhammer, A.; Specht, C. G.; Lynn, A. J.; Baker, P. R.; Schoepfer, R.; Burlingame, A. L. Quantitative analysis of synaptic phosphorylation and protein expression. Mol. Cell. Proteomics 2008, 7 (4), 684–96. (6) Munton, R. P.; Tweedie-Cullen, R.; Livingstone-Zatchej, M.; Weinandy, F.; Waidelich, M.; Longo, D.; Gehrig, P.; Potthast, F.; Rutishauser, D.; Gerrits, B.; Panse, C.; Schlapbach, R.; Mansuy, I. M. Qualitative and quantitative analyses of protein phosphorylation in naive and stimulated mouse synaptosomal preparations. Mol. Cell. Proteomics 2007, 6 (2), 283–93.

3288

Journal of Proteome Research • Vol. 9, No. 6, 2010

Wis´niewski et al. (7) Collins, M. O.; Yu, L.; Campuzano, I.; Grant, S. G.; Choudhary, J. S. Phosphoproteomic analysis of the mouse brain cytosol reveals a predominance of protein phosphorylation in regions of intrinsic sequence disorder. Mol. Cell. Proteomics 2008, 7 (7), 1331–48. (8) Ballif, B. A.; Carey, G. R.; Sunyaev, S. R.; Gygi, S. P. Large-scale identification and evolution indexing of tyrosine phosphorylation sites from murine brain. J. Proteome Res. 2008, 7 (1), 311–8. (9) Tweedie-Cullen, R. Y.; Reck, J. M.; Mansuy, I. M. Comprehensive mapping of post-translational modifications on synaptic, nuclear, and histone proteins in the adult mouse brain. J. Proteome Res. 2009, 8 (11), 4966–82. (10) Wisniewski, J. R.; Zougman, A.; Nagaraj, N.; Mann, M. Universal sample preparation method for proteome analysis. Nat. Methods 2009, 6 (5), 359–62. (11) Wisniewski, J. R.; Zougman, A.; Mann, M. Combination of FASP and StageTip-based fractionation allows in-depth analysis of the hippocampal membrane proteome. J. Proteome Res. 2009, 8 (12), 5674–8. (12) Zielinska, D. F.; Gnad, F.; Jedrusik-Bode, M.; Wisniewski, J. R.; Mann, M. Caenorhabditis elegans has a phosphoproteome atypical for metazoans that is enriched in developmental and sex determination proteins. J. Proteome Res. 2009, 8 (8), 4039–49. (13) Soufi, B.; Kelstrup, C. D.; Stoehr, G.; Frohlich, F.; Walther, T. C.; Olsen, J. V. Global analysis of the yeast osmotic stress response by quantitative proteomics. Mol. BioSyst. 2009, 5 (11), 1337–46. (14) Zielinska, D. F.; Gnad, F.; Wisniewski, J. R.; Mann, M. Precision mapping of an in vivo N-linked glycoproteome reveals rigid topological and sequence constraints. Cell 2010, 141 (May 28), in press. (15) Wisniewski, J. R.; Mann, M. Spin filter-based sample preparation for shotgun proteomics (Reply). Nat. Methods 2009, 6 (11), 785–6. (16) Olsen, J. V.; Vermeulen, M.; Santamaria, A.; Kumar, C.; Miller, M. L.; Jensen, L. J.; Gnad, F.; Cox, J.; Jensen, T. S.; Nigg, E. A.; Brunak, S.; Mann, M. Quantitative phosphoproteomics reveals widespread full phosphorylation site occupancy during mitosis. Sci. Signal. 2010, 3 (104), ra3. (17) Wisniewski, J. R.; Szewczuk, Z.; Petry, I.; Schwanbeck, R.; Renner, U. Constitutive phosphorylation of the acidic tails of the high mobility group 1 proteins by casein kinase II alters their conformation, stability, and DNA binding specificity. J. Biol. Chem. 1999, 274 (29), 20116–22. (18) Olsen, J. V.; Macek, B. High accuracy mass spectrometry in largescale analysis of protein phosphorylation. Methods Mol. Biol. 2009, 492, 131–42. (19) Villen, J.; Beausoleil, S. A.; Gerber, S. A.; Gygi, S. P. Large-scale phosphorylation analysis of mouse liver. Proc. Natl. Acad. Sci. U.S.A. 2007, 104 (5), 1488–93. (20) Nuhse, T. S.; Stensballe, A.; Jensen, O. N.; Peck, S. C. Large-scale analysis of in vivo phosphorylated membrane proteins by immobilized metal ion affinity chromatography and mass spectrometry. Mol. Cell. Proteomics 2003, 2 (11), 1234–43. (21) Pinkse, M. W.; Uitto, P. M.; Hilhorst, M. J.; Ooms, B.; Heck, A. J. Selective isolation at the femtomole level of phosphopeptides from proteolytic digests using 2D-NanoLC-ESI-MS/MS and titanium oxide precolumns. Anal. Chem. 2004, 76 (14), 3935–43. (22) Larsen, M. R.; Thingholm, T. E.; Jensen, O. N.; Roepstorff, P.; Jorgensen, T. J. Highly selective enrichment of phosphorylated peptides from peptide mixtures using titanium dioxide microcolumns. Mol. Cell. Proteomics 2005, 4 (7), 873–86. (23) Nielsen, P. A.; Olsen, J. V.; Podtelejnikov, A. V.; Andersen, J. R.; Mann, M.; Wisniewski, J. R. Proteomic mapping of brain plasma membrane proteins. Mol. Cell. Proteomics 2005, 4 (4), 402–8. (24) Olsen, J. V.; Blagoev, B.; Gnad, F.; Macek, B.; Kumar, C.; Mortensen, P.; Mann, M. Global, in vivo, and site-specific phosphorylation dynamics in signaling networks. Cell 2006, 127 (3), 635–48. (25) Olsen, J. V.; de Godoy, L. M.; Li, G.; Macek, B.; Mortensen, P.; Pesch, R.; Makarov, A.; Lange, O.; Horning, S.; Mann, M. Parts per million mass accuracy on an Orbitrap mass spectrometer via lock mass injection into a C-trap. Mol. Cell. Proteomics 2005, 4 (12), 2010– 21. (26) Cox, J.; Mann, M. MaxQuant enables high peptide identification rates, individualized p.p.b.-range mass accuracies and proteomewide protein quantification. Nat. Biotechnol. 2008, 26 (12), 1367– 72. (27) Perkins, D. N.; Pappin, D. J.; Creasy, D. M.; Cottrell, J. S. Probabilitybased protein identification by searching sequence databases using mass spectrometry data. Electrophoresis 1999, 20 (18), 3551–67. (28) Old, W. M.; Meyer-Arendt, K.; Aveline-Wolf, L.; Pierce, K. G.; Mendoza, A.; Sevinsky, J. R.; Resing, K. A.; Ahn, N. G. Comparison of label-free methods for quantifying human proteins by shotgun proteomics. Mol. Cell. Proteomics 2005, 4 (10), 1487–502.

research articles

Brain Phosphoproteome Reveals Plasma Membrane Protein Topology (29) Mootha, V. K.; Bunkenborg, J.; Olsen, J. V.; Hjerrild, M.; Wisniewski, J. R.; Stahl, E.; Bolouri, M. S.; Ray, H. N.; Sihag, S.; Kamal, M.; Patterson, N.; Lander, E. S.; Mann, M. Integrated analysis of protein composition, tissue diversity, and gene regulation in mouse mitochondria. Cell 2003, 115 (5), 629–40. (30) Whitelegge, J. P.; Zhang, H.; Aguilera, R.; Taylor, R. M.; Cramer, W. A. Full subunit coverage liquid chromatography electrospray ionization mass spectrometry (LCMS+) of an oligomeric membrane protein: cytochrome b(6)f complex from spinach and the cyanobacterium Mastigocladus laminosus. Mol. Cell. Proteomics 2002, 1 (10), 816–27. (31) Kuk, C.; Kulasingam, V.; Gunawardana, C. G.; Smith, C. R.; Batruch, I.; Diamandis, E. P. Mining the ovarian cancer ascites proteome for potential ovarian cancer biomarkers. Mol. Cell. Proteomics 2009, 8 (4), 661–9. (32) Li, Q. R.; Ning, Z. B.; Tang, J. S.; Nie, S.; Zeng, R. Effect of peptideto-TiO2 beads ratio on phosphopeptide enrichment selectivity. J. Proteome Res. 2009, 8 (11), 5375–81. (33) Wu, Y. B.; Dai, J.; Yang, X. L.; Li, S. J.; Zhao, S. L.; Sheng, Q. H.; Tang, J. S.; Zheng, G. Y.; Li, Y. X.; Wu, J. R.; Zeng, R. Concurrent quantification of proteome and phosphoproteome to reveal system-wide association of protein phosphorylation and gene expression. Mol. Cell. Proteomics 2009, 8 (12), 2809–26. (34) Shannon, P.; Markiel, A.; Ozier, O.; Baliga, N. S.; Wang, J. T.; Ramage, D.; Amin, N.; Schwikowski, B.; Ideker, T. Cytoscape: a software environment for integrated models of biomolecular interaction networks. Genome Res. 2003, 13 (11), 2498–504. (35) Maere, S.; Heymans, K.; Kuiper, M. BiNGO: a Cytoscape plugin to assess overrepresentation of gene ontology categories in biological networks. Bioinformatics 2005, 21 (16), 3448–9. (36) Gnad, F.; Ren, S.; Cox, J.; Olsen, J. V.; Macek, B.; Oroshi, M.; Mann, M. PHOSIDA (phosphorylation site database): management, structural and evolutionary investigation, and prediction of phosphosites. Genome Biol. 2007, 8 (11), R250. (37) Gnad, F.; de Godoy, L. M.; Cox, J.; Neuhauser, N.; Ren, S.; Olsen, J. V.; Mann, M. High-accuracy identification and bioinformatic analysis of in vivo protein phosphorylation sites in yeast. Proteomics 2009, 9 (20), 4642–52. (38) Wagner, M.; Adamczak, R.; Porollo, A.; Meller, J. Linear regression models for solvent accessibility prediction in proteins. J. Comput. Biol. 2005, 12 (3), 355–69. (39) Hubbard, T. J.; Aken, B. L.; Ayling, S.; Ballester, B.; Beal, K.; Bragin, E.; Brent, S.; Chen, Y.; Clapham, P.; Clarke, L.; Coates, G.; Fairley, S.; Fitzgerald, S.; Fernandez-Banet, J.; Gordon, L.; Graf, S.; Haider, S.; Hammond, M.; Holland, R.; Howe, K.; Jenkinson, A.; Johnson, N.; Kahari, A.; Keefe, D.; Keenan, S.; Kinsella, R.; Kokocinski, F.; Kulesha, E.; Lawson, D.; Longden, I.; Megy, K.; Meidl, P.; Overduin, B.; Parker, A.; Pritchard, B.; Rios, D.; Schuster, M.; Slater, G.; Smedley, D.; Spooner, W.; Spudich, G.; Trevanion, S.; Vilella, A.; Vogel, J.; White, S.; Wilder, S.; Zadissa, A.; Birney, E.; Cunningham, F.; Curwen, V.; Durbin, R.; Fernandez-Suarez, X. M.; Herrero, J.; Kasprzyk, A.; Proctor, G.; Smith, J.; Searle, S.; Flicek, P. Ensembl 2009. Nucleic Acids Res. 2009, 37 (Database issue), D690–7.

(40) Bairoch, A.; Apweiler, R. The SWISS-PROT protein sequence data bank and its new supplement TREMBL. Nucleic Acids Res. 1996, 24 (1), 21–5. (41) Krogh, A.; Larsson, B.; von Heijne, G.; Sonnhammer, E. L. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 2001, 305 (3), 567– 80. (42) Nuhse, T. S.; Stensballe, A.; Jensen, O. N.; Peck, S. C. Phosphoproteomics of the Arabidopsis plasma membrane and a new phosphorylation site database. Plant Cell 2004, 16 (9), 2394–405. (43) Wu, C. C.; MacCoss, M. J.; Howell, K. E.; Yates, J. R., III. A method for the comprehensive proteomic analysis of membrane proteins. Nat. Biotechnol. 2003, 21 (5), 532–8. (44) Dolphin, A. C. Calcium channel diversity: multiple roles of calcium channel subunits. Curr. Opin. Neurobiol. 2009, 19 (3), 237–44. (45) Davies, A.; Kadurin, I.; Alvarez-Laviada, A.; Douglas, L.; NietoRostro, M.; Bauer, C. S.; Pratt, W. S.; Dolphin, A. C. The {alpha}2{delta} subunits of voltage-gated calcium channels form GPI-anchored proteins, a posttranslational modification essential for function. Proc. Natl. Acad. Sci. U.S.A. 2010, 107 (4), 1654–9. (46) Klugbauer, N.; Marais, E.; Hofmann, F. Calcium channel alpha2delta subunits: differential expression, function, and drug binding. J. Bioenerg. Biomembr. 2003, 35 (6), 639–47. (47) Gerhardstein, B. L.; Puri, T. S.; Chien, A. J.; Hosey, M. M. Identification of the sites phosphorylated by cyclic AMP-dependent protein kinase on the beta 2 subunit of L-type voltagedependent calcium channels. Biochemistry 1999, 38 (32), 10361– 70. (48) Abiria, S. A.; Colbran, R. J. CaMKII associates with CaV1.2 L-type calcium channels via selected beta subunits to enhance regulatory phosphorylation. J. Neurochem. 2010, 112 (1), 150–61. (49) Enz, R. The trick of the tail: protein-protein interactions of metabotropic glutamate receptors. BioEssays 2007, 29 (1), 60–73. (50) Mao, L. M.; Liu, X. Y.; Zhang, G. C.; Chu, X. P.; Fibuch, E. E.; Wang, L. S.; Liu, Z.; Wang, J. Q. Phosphorylation of group I metabotropic glutamate receptors (mGluR1/5) in vitro and in vivo. Neuropharmacology 2008, 55 (4), 403–8. (51) Lu, A.; Wisniewski, J. R.; Mann, M. Comparative proteomic profiling of membrane proteins in rat cerebellum, spinal cord, and sciatic nerve. J. Proteome Res. 2009, 8 (5), 2418–25. (52) Nagaraj, N.; Lu, A.; Mann, M.; Wisniewski, J. R. Detergent-based but gel-free method allows identification of several hundred membrane proteins in single LC-MS runs. J. Proteome Res. 2008, 7 (11), 5028–32. (53) Zvonok, N.; Xu, W.; Williams, J.; Janero, D.; Krishnan, S.; Makriyannis, A. Mass spectrometry-based GPCR proteomics: comprehensive characterization of the human cannabinoid 1 receptor. J. Proteome Res. 2010, 9 (4), 1746–53. (54) Yan, J.; Olsen, J. V.; Park, K. S.; Li, W.; Bildl, W.; Schulte, U.; Aldrich, R. W.; Fakler, B.; Trimmer, J. S. Profiling the phospho-status of the BKCa channel alpha subunit in rat brain reveals unexpected patterns and complexity. Mol. Cell. Proteomics 2008, 7 (11), 2188–98.

PR1002214

Journal of Proteome Research • Vol. 9, No. 6, 2010 3289