140 Mouse Brain Proteins Identified by Ca2+-Calmodulin Affinity Chromatography and Tandem Mass Spectrometry Tord Berggård,*,† Giorgio Arrigoni,† Olof Olsson,† Malin Fex,§ Sara Linse,‡ and Peter James† Department of Protein Technology, Lund University, So¨lvegatan 33, Wallenberglaboratoriet, SE-221 00 Lund, Sweden, Department of Biophysical Chemistry, Lund University, SE-221 00 Lund, Sweden, and Department of Cell and Molecular Biology, Section for Molecular Signaling, Lund University, SE-22184 Lund, Sweden Received November 28, 2005
Calmodulin is an essential Ca2+-binding protein that binds to a variety of targets that carry out critical signaling functions. We describe the proteomic characterization of mouse brain Ca2+-calmodulin-binding proteins that were purified using calmodulin affinity chromatography. Proteins in the eluates from four different affinity chromatography experiments were identified by 1-DE and in-gel digestion followed by LC-MS/MS. Parallel experiments were performed using two related control-proteins belonging to the EF-hand family. After comparing the results from the different experiments, we were able to exclude a significant number of proteins suspected to bind in a nonspecific manner. A total of 140 putative Ca2+-calmodulin-binding proteins were identified of which 87 proteins contained calmodulin-binding motifs. Among the 87 proteins that contained calmodulin-binding motifs, 48 proteins have not previously been shown to interact with calmodulin and 39 proteins were known calmodulin-binding proteins. Many proteins with ill-defined functions were identified as well as a number of proteins that at the time of the analysis were described only as ORFs. This study provides a functional framework for studies on these previously uncharacterized proteins. Keywords: calmodulin • calcium • protein-protein interactions • proteomics • brain
Introduction Calmodulin is a ubiquitous protein that is expressed in all eukaryotic cells. It participates in signaling pathways that regulate many crucial processes such as growth, proliferation and movement. Regulation of these events is exerted via direct interactions with a large number of cellular proteins. Calmodulin constitutes at least 0.1% of the total protein in cells and it is expressed at even higher levels in brain and in rapidly growing cells, especially those undergoing division and differentiation. The protein is strongly conserved and the same sequence is found in all vertebrates. Following an extracellular stimulus, Ca2+ moves into the cytosol either from outside the cell (via plasma membrane Ca2+ channels) or from intracellular stores. Binding of Ca2+ releases Mg2+ from calmodulin and causes the protein to undergo a conformational change that increases its binding affinity for a number of target proteins. Because Ca2+ binds to calmodulin in a cooperative fashion, a small change in the level of cytosolic Ca2+ leads to a large change in the level of active protein. Calmodulin contains four EF-hand Ca2+-binding motifs that are coupled pairwise in two globular domains. The Ca2+-bound * To whom correspondence should be addressed. Tel: +46-46-222 38 35. Fax: +46-46-222 14 95. E-mail:
[email protected]. † Department of Protein Technology. ‡ Department of Biophysical Chemistry. § Department of Cell and Molecular Biology, Section for Molecular Signaling. 10.1021/pr050421l CCC: $33.50
2006 American Chemical Society
form of the protein has exposed hydrophobic surfaces and its structure is different from the unbound, or Mg2+-bound form.1 In the resting cell, the two sites in the N-terminal domain are occupied by Mg2+, whereas the two sites in the C-terminal domain are either unoccupied, or bound to Ca2+ with the C-terminal domain pre-bound to target.2 An intriguing feature of calmodulin is the large number of proteins that are bound in a Ca2+-dependent manner with high affinity (KD around nM) without any strong sequence homology of the binding regions. The region that binds to calmodulin is an amphiphatic R-helix with net positive charge, and a number of motifs have been identified3 although these do not conform to all targets. The three-dimensional structures of calmodulintarget complexes are more diverse than anticipated from the initial structures and some proteins bind equally well or only to Ca2+-free calmodulin.4,5 In the most well-known structures observed for many kinases, the calmodulin-binding segment can be produced as a linear peptide that binds as an R-helix in a tunnel formed between the two globular domains of calmodulin.6 In the case of glutamic acid decarboxylase, two copies of the helical binding region are bound in a wider cleft between the domains,7 and for the conductance potassium channel, two helix-turn-helix segments form a 2:2 complex with calmodulin.8 A question that has been puzzling many scientists is how calmodulin is able to interact with such a large and diverse array of targets. Furthermore, many of the reactions regulated Journal of Proteome Research 2006, 5, 669-687
669
Published on Web 02/16/2006
research articles
Berggård et al.
by calmodulin have opposing effects. For example, calmodulin both activates enzymes that synthesize and degrade cyclic AMP. A possible answer to this paradox is that cells may control subcellular pools of calmodulin and in this way regulate the availability of both target protein and calmodulin. One way to achieve this task is the positioning of mRNAs instead of protein, followed by local translation if required.9 Furthermore, calmodulin may itself be subject to regulation by various mechanisms such as phosphorylation.10 It was recently suggested that a key feature in calmodulin-target specificity is negative design through electrostatic repulsion of nonwanted targets.11 The majority of work on calmodulin can be classified into three general areas: (1) elucidating the calmodulin structure and dynamics of Ca2+ interaction with other proteins, (2) determining the expression pattern and regulation of calmodulin mRNA levels in various organisms, and (3) discovering novel calmodulin targets. Each area has given important insights into the function and regulation of calmodulin. In the present report, we have used tandem mass spectrometry to characterize Ca2+-calmodulin-binding proteins that were purified using calmodulin affinity chromatography, a procedure that has been used previously to purify bona fide calmodulinbinding proteins. Using this approach, we have been able to identify many new potential calmodulin-binding proteins and to verify a lot of putative binding targets.
UV absorption at 280 nm) before elution with EGTA. Bound proteins were eluted with 50 mM Tris, 0.15 NaCl, 1mM DTT, 2mM EGTA, pH 7.5. In a fourth experiment, with the aim of including proteins needing detergent for solubilization, we homogenized mouse brains as above, ultracentrifuged at 100 000 × g for 90 min, collected the pellet and dissolved it by homogenization by a Potter-Elvehjem tissue grinder in 20 mL of a buffer containing detergent (1% β-dodecyl maltoside, 50 mM Tris, 150 mM NaCl, 1 mM DTT + 200 uL protease inhibitor cocktail (Sigma P8340)). The pellet fraction was then incubated for 1 h, ultra-centrifuged at 100 000 × g for 90 min, and the supernatant was collected. CaCl2 (2 mM) was added to the supernatant, which was ultracentrifuged again at 100 000 × g for 90 min, and the supernatant was collected and filtered through a 45 µM filter. Seven milliliters of the supernatant was then incubated with calmodulin-, secretagogin- or calbindin D28k- sepharose, respectively, for 2 h. The affinity columns were then rinsed with 10 column volumes of 50 mM Tris, 0.15 M NaCl, 1mM DTT, 2mM CaCl2, pH 7.5 containing 0,25% β-dodecyl maltoside, followed by elution with 50 mM Tris, 0.15 NaCl, 1mM DTT, 2mM EGTA, 0,25% β-dodecyl maltoside, pH 7.5. The eluates from all four experiments were concentrated by ultra-filtration to 50 µL. Twenty microliters of the eluates from the calmodulin, secretagogin, and calbindin D28k columns was separated on SDS-PAGE (12%) (Figure 1).
Experimental Section
Sample Preparation for Mass Spectrometry. After the separation of proteins by SDS-PAGE (10%), the gel lanes were divided horizontally into equally sized slices that were individually processed to yield in-gel trypsin digestion products. Coomassie-stained protein fragments were excised from the gel and washed with water, followed by 40% acetonitrile in 25 mM NH4HCO3, pH 7.8, until the gel piece was transparent. The gel piece was dried in a SpeedVac vacuum centrifuge. Reduction using 10 mM DTT at 48 °C for 30 min was followed by alkylation in 55 mM iodoacetamide for 30 min in darkness at room temperature. The gel piece was washed and dried again before digestion with sequencing-grade trypsin (Promega) in 25 mM NH4HCO3 overnight at 37 °C. The digestion was terminated and the peptides eluted by adding 10 µL 2% trifluoroacetic acid.
Affinity Chromatography and SDS-PAGE. Human recombinant calmodulin, secretagogin or calbindin D28k (3 mg each) were expressed in Escherichia coli and purified to homogeneity as described12 and immobilized on 1 mL of CNBr activated Sepharose. Mouse brain tissue (3 g) from 10 different mice was homogenized using a mixer followed by a Potter-Elvehjem tissue grinder in 20 mL of 50 mM Tris, 150 mM NaCl, 1 mM DTT + 200 uL protease inhibitor cocktail (Sigma P8340). The cytosol provides a reducing environment that prevents the formation of disulfide linkages by maintaining the Cys residues of cytosolic proteins in the reduced form. To maintain a reducing environment and to prevent the formation of nonspecific intermolecular disulfide cross-links, we have performed the experiments in 1 mM DTT. The homogenate was then centrifuged 3 times for 6 min at 3000 × g, and the supernatant was collected and ultracentrifuged at 100 000 × g for 90 min. 2 mM CaCl2 and 1 mM MgCl2 were added, and the supernatant was filtered through a 0.45 µM nitrocellulose filter. This procedure was repeated three times using brains from different mice each time. Different affinity chromatography experiments were performed as follows to obtain a variation in the binding and washing times. In the first experiment, calmodulinsepharose, secretagogin-sepharose or calbindin D28K sepharose, respectively, were incubated for 2 h with mouse brain homogenate and washed with 10 column volumes of 50 mM Tris, 0.15 M NaCl, 1mM DTT, 2mM CaCl2, pH 7.5 as recommended by the manufacturer of the affinity chromatography resin (Amersham Pharmacia Biotech). In the second experiment, we incubated the calmodulin resin for 2 h with mouse brain extract as in the previous experiment, but in this case we washed extensively with 50 column volumes of 50 mM Tris, 0.15 M NaCl, 1mM DTT, 2mM CaCl2, pH 7.5. In the third experiment, we applied the sample (without incubating), using a flow of 20 cm/h and washed with 10 column volumes of 50 mM Tris, 0.15 M NaCl, 1mM DTT, 2mM CaCl2, pH 7.5. In all cases, we checked that no material appeared in the wash (monitored by 670
Journal of Proteome Research • Vol. 5, No. 3, 2006
Protein Identification by Mass Spectrometry. The resulting peptide mixtures were separated and analyzed in an automated system by nanoscale liquid chromatography quadrupole timeof-flight MS/MS using an Ultima Q-TOF (Micromass) connected with CapLC pumps (Waters) and a C18 PepMap100 capillary column (3 µm, 100 Å, 15 cm, 75 µm I.D., LC Packings, CA). The peptides were separated with a linear gradient of acetonitrile/formic acid 0.1% from 7% to 85% for 60 min. After fragmentation in the MS/MS mode, spectra from all bands from each gel-lane were merged to a single.pkl-file and analyzed by MASCOT (Mascot Demon) software to identify tryptic peptide sequences matched to the National Center for Biotechnology Information (NCBI) database. The following parameters were used in the Mascot search (http://www.proteomics.com.cn/ mascot/ search_form_MSMS.html); fixed modifications: carbamidomethyl (Cys), variable modifications: oxidation (Met); mass values: monoisotopic; protein mass: unrestricted; peptide mass tolerance 0.1 Da; fragment mass tolerance: 0.1 Da; maximum number of missed cleavages: 1; enzyme: trypsin. Identification of Calmodulin Binding Motifs. A web-based database (http://calcium.uhnres.utoronto.ca/ctdb) was used for identification of putative calmodulin-binding motifs.
Characterization of the Brain Calmodulin Interactome
Figure 1. (A). Separation and identification of calmodulin-binding proteins. Proteins binding to Ca2+-calmodulin, Ca2+-secretagogin, and Ca2+-calbindin D28k from mouse brain were purified by affinity chromatography. In the first experiment, the calmodulin-, secretagogin- and calbindin D28k-columns were incubated for 2 hours with mouse brain homogenate and washed with 10 column volumes of wash buffer. In the second experiment, the calmodulin resin was incubated for 2 hours with mouse brain extract and washed extensively with 50 column volumes of wash buffer. In the third exeriment, brain-homogenate was applied onto a calmodulin affinity column without incubating, using a flow of 20 cm/h and washed with 10 column volumes of wash buffer. In the fourth experiment a detergent (β-dodecyl maltoside) was included when solubilizing the brain extract which was then incubated for 2 hours with calmodulin-sepharose, secretagoginsepharose, or calbindin D28k-sepharose and rinsed with 10 column volumes of wash buffer containing detergent. Each column was eluted with 5 mL of a buffer containing 2 mM EGTA. All EGTA-eluates were concentrated to 50 µL by ultrafiltration. Twenty µL of the eluate was separated by SDS/PAGE (12%). Gel slices were cut from each gel, digested with trypsin and identified by LC-MS/MS followed by database searching.
Results Affinity based interaction data are biased toward proteins of high abundance and selects for proteins interacting with high affinity and slow kinetics of dissociation. As calmodulin binds most of its targets with high affinity, affinity-based approaches such as affinity chromatography can be used to isolate stable multi-protein complexes that contain calmodulin. Upon binding of Ca2+, calmodulin undergoes a conformational change, which allows it to bind to target-proteins via an exposed hydrophobic patch. Ca2+ chelators such as EGTA remove Ca2+ from calmodulin thereby reversing the conformational change that exposed the hydrophobic patch. In this report, we have purified mouse cytosolic brain proteins that bind to Ca2+calmodulin using immobilized recombinant protein and elution with EGTA. As there is no generally applicable affinity chromatography scheme for calmodulin-binding proteins, we repeated the experiments four times at different occasions using brain homogenates from different mice. In each experiment, the conditions were changed slightly (see Materials and Meth-
research articles ods Section). Negative controls were conducted in parallel using sepharose 4B cross-linked to calbindin D28k or secretagogin, which are EF-hand containing Ca2+ binding proteins belonging to the calmodulin protein-super family.13,14 Experiments with calbindin D28k and secretagogin were performed under two different conditions (see materials and methods section). SDSPAGE showed a number of bands present in the eluate from the calmodulin column, while the eluates from the calbindin D28k- and secretagogin-columns contained significantly less proteins (Figure 1). Each lane with eluted proteins from the 8 columns (four calmodulin-columns, two secretgogin- and two calbindin D28K columns) was cut in equally sized gel slices followed by tryptic digestion and peptide identification by LCMS/MS and database searches. The experience in our lab is that 1D gel electrophoresis is a superior approach if the purpose is identification of all proteins in a complex mixture without quantification. When using additional separation (liquid chromatography) of peptide mixtures from 1-D gel bands it is not necessary to resolve every individual protein due to the high resolving power of the capillary liquid phase separation. Therefore, we have chosen to cut bands directly from 1D gels instead of first separating eluted proteins by 2D electrophoresis. Proteins identified with Mascot scores of > 28 (corresponding to a p value
