a proteome-wide screen for mammalian sxip motif-containing microtubule plus-end tracking proteins

8
Current Biology 22, 1800–1807, October 9, 2012 ª2012 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.cub.2012.07.047 Report A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins Kai Jiang, 1 Grischa Toedt, 2 Susana Montenegro Gouveia, 3 Norman E. Davey, 2 Shasha Hua, 1 Babet van der Vaart, 3 Ilya Grigoriev, 1 Jesper Larsen, 5 Lotte B. Pedersen, 5 Karel Bezstarosti, 4 Mariana Lince-Faria, 6 Jeroen Demmers, 4 Michel O. Steinmetz, 7 Toby J. Gibson, 2 and Anna Akhmanova 1, * 1 Cell Biology, Department of Biology, Faculty of Science, Utrecht University, Padualaan 8, 3584 CH Utrecht, The Netherlands 2 European Molecular Biology Laboratory, Meyerhofstraße 1, 69012 Heidelberg, Germany 3 Department of Cell Biology 4 Proteomics Center Erasmus Medical Center, PO Box 2040, 3000 CA Rotterdam, The Netherlands 5 Department of Biology, University of Copenhagen, Universitetsparken 13, DK-2100 Copenhagen, Denmark 6 Cell Cycle Regulation Laboratory, Instituto Gulbenkian de Cie ˆ ncia, Rua da Quinta Grande, 6, P-2780-156 Oeiras, Portugal 7 Biomolecular Research, Paul Scherrer Institut, 5232 Villigen PSI, Switzerland Summary Microtubule plus-end tracking proteins (+TIPs) are struc- turally and functionally diverse factors that accumulate at the growing microtubule plus-ends, connect them to various cellular structures, and control microtubule dynamics [1, 2]. EB1 and its homologs are +TIPs that can autonomously recognize growing microtubule ends and recruit to them a variety of other proteins. Numerous +TIPs bind to end binding (EB) proteins through natively unstructured basic and serine-rich polypeptide regions containing a core SxIP motif (serine-any amino acid- isoleucine-proline) [3]. The SxIP consensus sequence is short, and the surrounding sequences show high vari- ability, raising the possibility that undiscovered SxIP containing +TIPs are encoded in mammalian genomes. Here, we performed a proteome-wide search for mammalian SxIP-containing +TIPs by combining biochemical and bioinformatics approaches. We have identified a set of previ- ously uncharacterized EB partners that have the capacity to accumulate at the growing microtubule ends, including protein kinases, a small GTPase, centriole-, membrane-, and actin-associated proteins. We show that one of the newly identified +TIPs, CEP104, interacts with CP110 and CEP97 at the centriole and is required for ciliogenesis. Our study reveals the complexity of the mammalian +TIP interac- tome and provides a basis for investigating the molecular crosstalk between microtubule ends and other cellular structures. Results Identification of EB1 Partners by a Combination of Mass Spectrometry and Sequence Analysis To identify new end binding 1 (EB1) binding partners, we per- formed GST-EB1 pull-down assays combined with mass spectrometry using extracts of different cultured cells and whole rat brain. Many SxIP (serine-any amino acid-isoleu- cine-proline) containing plus-end tracking proteins (+TIPs) as well as the CAP-Gly domain containing EB partners, CLIP-170 and dynactin together with the associated dynein complex, were identified by this method (Tables S1A and S1C). We also overexpressed EB1, EB2, and EB3 with a GFP and a biotinylation tag (GFPbio) together with biotin ligase BirA in human embryonic kidney 293T (HEK293T) cells and performed a streptavidin pull-down assay (Figures S1A and S1C; Table S1D). There was a strong overlap of known SxIP motif-containing proteins (17 out of 22) in the data sets obtained with the two types of pull-downs. Almost all of the known SxIP containing +TIPs (21 out of 22) could be found in either of the pull-down experiments. As a complementary approach, we performed a systematic computational search for SxIP motif-containing proteins in the human proteome. The [ST]-X-[IL]-P motif is present in one or more copies in greater than 40% of human proteins. To refine the search, we analyzed the composition of amino acids in the region adjacent to the SxIP motif in 14 well-characterized EB1 partners (Figure S1D). We found that within the nine amino acid stretch surrounding the SxIP consensus (X1-X2-[ST]-X3-[IL]- P-X4-X5-X6), there are no acidic amino acids and at least one basic amino acid is present at one of the positions X1-X4. By using this contextual information (SxIP-9AA) and taking into account that the SxIP-containing peptide should be intrinsically disordered and evolutionarily conserved, we narrowed down the list of potential EB1-binding SxIP proteins to 833 entries (Figure 1A; see Experimental Procedures for details). To rank these proteins, we calculated an ‘‘SxIP score’’ by using a position-specific scoring matrix (PSSM), which was based on the residue preference for each amino acid position within the SxIP motif and the surrounding sequences obtained by quantifying the interaction of purified EB1 with an array of immobilized synthetic peptides [4](Table S1E). Most of the known SxIP EB binding proteins (20 out of 22) were among the top 20% of proteins ranked by this PSSM-based SxIP score (153/833, Tables S1A and S1E). This suggests that the SxIP score provides a good indication of the capacity of a protein to interact with EB1. Next, we plotted the Mascot scores obtained in GST and streptavidin pull-down assays against the SxIP scores of indi- vidual proteins (Figure 1B; Figure S1E). A group of proteins, which includes many known SxIP-containing +TIPs, such as CLASPs, MCAK, STIM1, and MACF had both a high normal- ized Mascot score (exceeding 0.5) and a high SxIP score (exceeding 0.4, Figure 1B; Figure S1E). We hypothesized that other proteins in this group might be EB-binding +TIPs and set out to test this idea. Based on the SxIP score, Mascot score, complemen- tary DNA (cDNA) availability, and potentially interesting *Correspondence: [email protected]

Upload: kai-jiang

Post on 25-Nov-2016

213 views

Category:

Documents


0 download

TRANSCRIPT

Page 1: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

A Proteome-wide Screen

Current Biology 22, 1800–1807, October 9, 2012 ª2012 Elsevier Ltd All rights reserved http://dx.doi.org/10.1016/j.cub.2012.07.047

Report

for Mammalian SxIP Motif-ContainingMicrotubule Plus-End Tracking Proteins

Kai Jiang,1 Grischa Toedt,2 Susana Montenegro Gouveia,3

Norman E. Davey,2 Shasha Hua,1 Babet van der Vaart,3

Ilya Grigoriev,1 Jesper Larsen,5 Lotte B. Pedersen,5

Karel Bezstarosti,4 Mariana Lince-Faria,6 Jeroen Demmers,4

Michel O. Steinmetz,7 Toby J. Gibson,2

and Anna Akhmanova1,*1Cell Biology, Department of Biology, Faculty of Science,Utrecht University, Padualaan 8, 3584 CH Utrecht, TheNetherlands2European Molecular Biology Laboratory, Meyerhofstraße 1,69012 Heidelberg, Germany3Department of Cell Biology4Proteomics CenterErasmus Medical Center, PO Box 2040, 3000 CA Rotterdam,The Netherlands5Department of Biology, University of Copenhagen,Universitetsparken 13, DK-2100 Copenhagen, Denmark6Cell Cycle Regulation Laboratory, Instituto Gulbenkiande Ciencia, Rua da Quinta Grande, 6, P-2780-156 Oeiras,Portugal7Biomolecular Research, Paul Scherrer Institut, 5232 VilligenPSI, Switzerland

Summary

Microtubule plus-end tracking proteins (+TIPs) are struc-

turally and functionally diverse factors that accumulate

at the growing microtubule plus-ends, connect them tovarious cellular structures, and control microtubule

dynamics [1, 2]. EB1 and its homologs are +TIPs that canautonomously recognize growing microtubule ends and

recruit to them a variety of other proteins. Numerous +TIPsbind to end binding (EB) proteins through natively

unstructured basic and serine-rich polypeptide regionscontaining a core SxIP motif (serine-any amino acid-

isoleucine-proline) [3]. The SxIP consensus sequenceis short, and the surrounding sequences show high vari-

ability, raising the possibility that undiscovered SxIPcontaining +TIPs are encoded in mammalian genomes.

Here, we performed a proteome-wide search for mammalianSxIP-containing +TIPs by combining biochemical and

bioinformatics approaches.We have identified a set of previ-ously uncharacterized EB partners that have the capacity to

accumulate at the growing microtubule ends, includingprotein kinases, a small GTPase, centriole-, membrane-,

and actin-associated proteins. We show that one of thenewly identified +TIPs, CEP104, interacts with CP110 and

CEP97 at the centriole and is required for ciliogenesis. Ourstudy reveals the complexity of the mammalian +TIP interac-

tome and provides a basis for investigating the molecularcrosstalk between microtubule ends and other cellular

structures.

*Correspondence: [email protected]

Results

Identification of EB1 Partners by a Combination of Mass

Spectrometry and Sequence AnalysisTo identify new end binding 1 (EB1) binding partners, we per-formed GST-EB1 pull-down assays combined with massspectrometry using extracts of different cultured cells andwhole rat brain. Many SxIP (serine-any amino acid-isoleu-cine-proline) containing plus-end tracking proteins (+TIPs)as well as the CAP-Gly domain containing EB partners,CLIP-170 and dynactin together with the associated dyneincomplex, were identified by this method (Tables S1A andS1C). We also overexpressed EB1, EB2, and EB3 with a GFPand a biotinylation tag (GFPbio) together with biotin ligaseBirA in human embryonic kidney 293T (HEK293T) cells andperformed a streptavidin pull-down assay (Figures S1A andS1C; Table S1D). There was a strong overlap of known SxIPmotif-containing proteins (17 out of 22) in the data setsobtained with the two types of pull-downs. Almost all of theknown SxIP containing +TIPs (21 out of 22) could be found ineither of the pull-down experiments.As a complementary approach, we performed a systematic

computational search for SxIP motif-containing proteins in thehuman proteome. The [ST]-X-[IL]-P motif is present in one ormore copies in greater than 40% of human proteins. To refinethe search, we analyzed the composition of amino acids in theregion adjacent to the SxIP motif in 14 well-characterized EB1partners (Figure S1D).We found that within the nine amino acidstretch surrounding the SxIP consensus (X1-X2-[ST]-X3-[IL]-P-X4-X5-X6), there are no acidic amino acids and at leastone basic amino acid is present at one of the positionsX1-X4. By using this contextual information (SxIP-9AA) andtaking into account that the SxIP-containing peptide shouldbe intrinsically disordered and evolutionarily conserved, wenarrowed down the list of potential EB1-binding SxIP proteinsto 833 entries (Figure 1A; see Experimental Proceduresfor details). To rank these proteins, we calculated an ‘‘SxIPscore’’ by using a position-specific scoring matrix (PSSM),which was based on the residue preference for each aminoacid position within the SxIP motif and the surroundingsequences obtained by quantifying the interaction of purifiedEB1 with an array of immobilized synthetic peptides [4](TableS1E). Most of the known SxIP EB binding proteins (20 out of22) were among the top 20% of proteins ranked by thisPSSM-based SxIP score (153/833, Tables S1A and S1E).This suggests that the SxIP score provides a good indicationof the capacity of a protein to interact with EB1.Next, we plotted the Mascot scores obtained in GST and

streptavidin pull-down assays against the SxIP scores of indi-vidual proteins (Figure 1B; Figure S1E). A group of proteins,which includes many known SxIP-containing +TIPs, such asCLASPs, MCAK, STIM1, and MACF had both a high normal-ized Mascot score (exceeding 0.5) and a high SxIP score(exceeding 0.4, Figure 1B; Figure S1E). We hypothesized thatother proteins in this group might be EB-binding +TIPs andset out to test this idea.Based on the SxIP score, Mascot score, complemen-

tary DNA (cDNA) availability, and potentially interesting

Page 2: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Figure 1. Identification of EB1 Partners by Combining Biochemical and Bioinformatics Approaches

(A) Definition of the SxIP-9AAmotif and an overview of the bioinformatics screen for EB binding proteins. We performed a human proteome-wide short linear

motif (SLiM) search using the regular expression [ST].[IL]P. The motif instances were filtered based on the SxIP-9AA consensus, accessibility, and conser-

vation and then ranked.

(B) A plot of the normalizedMascot scores versus the SxIP scores of the proteins identified in GST-EB pull-down. The color codes for the dots are indicated

in the figure.

Proteome-wide Screen for +TIPs with an SxIP Motif1801

microtubule (MT)-related functions, we chose 30 proteins forfurther analysis (Table S1B). These 30 proteins included twogroups: 17 proteins or homologs of the proteins that hadboth a reasonably high Mascot score in at least one of theexperiments (normalized Mascot score >0.1, Mascot score>500) and a reasonably high SxIP score, and 13 proteins thathad a high or intermediate SxIP score but very low or zeroMascot score (Table S1B). We generated GFP fusions of theseproteins and tested them for EB binding by GST pull-downassay. As a control, we included DDA3, which was shown tobind to EB3 but was not yet characterized as a +TIP [5]. Wealso tested MT plus-end tracking of all GFP fusions by livecell imaging and colocalization with endogenous EB1 in fixedcells (see below). Most of the proteins that were identified byboth a high SxIP score and a high Mascot score could bindto EB1 and also track growing MT ends in cells (Figure S1F;Table S1B). Only two proteins, MARK1 and FAM181A, boundto EB1 but showed no plus-end tracking when fused to eitherN- or C-terminal GFP tag (Figure S1F; Table S1B). Among theproteins identified solely on the basis of their SxIP score, wefound five EB1-binding and plus-end tracking proteins,whereas the others did not bind to EB1 and showed no plus-end tracking (Figure S1F). Taken together, we were able toidentify 22 novel EB binding proteins from 30 candidates,and 20 of them could track growing MT plus-ends in cells

(see below). We conclude that +TIPs comprise a much largergroup of proteins than previously recognized.

+TIPs Implicated in MT Binding and OrganizationAmong the previously described MT-associated factors iden-tified in our screen, GTSE1 [6], tastin [7, 8], and DDA3, whichwas previously shown to bind to EB3 [5], strongly accumulatedat growing MT ends (Figure 2A; Figure S2A). Interestingly, tas-tin and DDA3 were the only two tested +TIPs that tracked notonly growing, but also depolymerizing MT ends (Figures 2B–2E; Movie S1). Tastin and DDA3 have been shown to playa role in controlling organization of mitotic spindles and tobind to dynein-dynactin complex and the MT depolymeraseKIF2A, respectively [7, 9]. It is possible that ‘‘back-tracking’’of tastin and DDA3 contributes to the binding of these motorsto depolymerizing MT ends within the spindle.Another notable EB partner identified in our screen was the

mitotic kinesin-5 Eg5, a motor that plays an important role inspindle bipolarity by sliding antiparallel MTs (see [10] forreview). An interaction of Eg5 with EB1 was observed inextracts of both asynchronous as well as mitotically synchro-nized cells and was somewhat enhanced when the Eg5 motorwas inhibited by S-trityl-L-cysteine (STLC, Figure S2B).Although weak, this interaction was specific because it wasstrongly reduced by mutating the Ile and Pro residues of the

Page 3: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Figure 2. +TIPs Implicated in MT Binding and Organization

(A) Schemes of the domain structures of EB binding proteins and the interactions between CEP104, CEP97, and CP110. Abbreviations: CC, coiled coil; LRR,

leucine rich repeat. The data on CP110-CEP97 interaction are derived from reference [14].

(B and C) Live cell images of GFP-tastin (B) and GFP-DDA3 (C) in COS-7 cells. The images are acquired with 500 ms exposure time at the indicated time

points.

(B) GFP-tastin showed a punctate localization along MTs, and when MTs were depolymerized, GFP-tastin puncta remained associated with shrinking MT

ends (arrowhead), where they gradually accumulated and could slow MT depolymerization.

(C) GFP-DDA3 was present in homogeneous comet-like accumulations at growing MT plus ends and shrinking plus and minus ends (see also Movie S1);

DDA3 had no significant effect onMT depolymerization. A depolymerizingMT plus-end is markedwith an arrowhead. A polymerizing MT end is marked with

an arrow.

(D and E) Kymographs illustrating the behavior of GFP-tastin and GFP-DDA3 at a depolymerizing (top) and growing (bottom) MT plus ends.

(F and G) Live cell images of the indicated GFP fusion proteins in MRC-5 (F) or COS-7 (G) cells. Panels on the left show the averaging of five consecutive

images acquired with 500 ms exposure time. Middle panels show the maximum intensity projections of four consecutive averaged images, shown in red,

green, blue, and then red again, respectively. Panels on the right show enlargements of the areas boxed in themiddle panel showing in alternating colors the

tracks of individual comets.

(H) Verification of the binding between CEP104 and CEP97/CP110 complex by coIP from extracts of HEK293T cells.

(I) GFP-CEP104 overlaps with the endogenous CP110 but not with centrin at the distal end of both centrioles in interphase U2OS cells.

(J) U2OS cells were transfected with either GFP-CEP97 alone or a combination of GFP-CEP97 and mCherry-CEP104. GFP-CEP97 is recruited to MT plus

ends by mCherry-CEP104.

(K) Western blots with the indicated antibodies showing that endogenous CEP104 can be effectively knocked down by two different siRNAs.

(L) Quantification of primary cilia numbers in control andCEP104 depleted serum-starvedRPE cells. Values significantly different from controls are indicated

by asterisks. **p < 0.01.

Current Biology Vol 22 No 191802

Page 4: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Proteome-wide Screen for +TIPs with an SxIP Motif1803

single SxIP motif of Eg5 to asparagines (Figure S2B). Weobserved weak but clearly detectable accumulation of Eg5 atthe growing MT tips in interphase cells (Figure 2F); however,because of the much denser MT system and association ofEg5 with the MT lattice, it was not possible to distinguisha MT tip-bound Eg5 pool in mitosis. The affinity of Eg5 forgrowing MT ends could be functionally relevant, becauseEg5 interacts with the abundant +TIPs dynein and dynactin[11] and its yeast homolog was implicated in control of MTplus-end dynamics [12].

Another newly identified +TIP was an evolutionaryconserved protein CEP104/KIAA0562 (Figures 2A and 2G),which was previously shown to localize to centrioles [13].Because of a potential importance of a MT regulator incentriole assembly, we performed a secondary mass spec-trometry-based screen for CEP104 partners and identifiedCP110 and its binding partner CEP97 [14] as CEP104-bindingproteins (Figure S2C). We confirmed the interactions by coim-munoprecipitation (coIP) and mapped the binding domains ofCEP104 for CP110 and CEP97 to two protein regions that didnot overlap with each other or the SxIP motif (Figures 2A and2H; Figures S2D–S2G). We note that because all immunopre-cipitations were performed using cell extracts, the interactionsmight be indirect. We found that CEP104 colocalized withCP110 at the distal end of the each centriole in interphaseU2OS, HeLa, and RPE cells (Figure 2I; Figure S2H; data notshown). In ciliated RPE cells, CEP104 was never found at thebasal body but could be found at the adjacent daughtercentriole, similar to CP110 [14] (Figure S2I). In mammaliancells, CP110 and CEP97 act as a ‘‘cap’’ at the distal end ofthe centriole, which regulates elongation of the plus ends ofcentriolar MTs [14, 15], but it is still poorly understood howthese proteins are connected to MTs. We found that whenoverexpressed, CEP104 could efficiently recruit the normallydiffuse CEP97 to the growing MT tips (Figure 2J), suggestingthat CEP104 can mediate the connection between CP110-CEP97 complex and dynamic MT ends. Depletion of CEP104with two different siRNAs (Figure 2K) revealed no obviousphenotypes in distribution or dynamics of EB1 and EB3, cell-cycle progression and centriole number or length in U2OSand RPE cells (data not shown). However, the ability of cellsto form cilia in RPE cells was reduced when CEP104 wasdepleted (Figure 2L; Figure S2J). CP110 and CEP97 areremoved from the basal body during cilia formation [14], andCEP104 possibly regulates the function of CP110-CEP97complex when cilia outgrowth is initiated. This process mightinvolve EB proteins because they are required for cilia forma-tion [16].

Membrane-Associated +TIPs

Among the new EB partners, we identified AMER2/FAM123A(Figure 3A; Figure S2K), which was described as an APCbinding protein targeted to the plasma membrane by a lipid-binding domain [17]. We found that AMER2 accumulated atthe plasma membrane through its N-terminal domain and atMT tips through its C terminus (Figures 3A and 3B; Figure S2L).Dual color imaging of GFP-AMER2 together with EB3-mRFPin the vicinity of the cell cortex by total internal reflectionfluorescence (TIRF) microscopy showed that only a part ofEB3-mRFP-labeled growing MT ends were decorated withGFP-AMER2, and such decoration was associated withreduced MT growth rate (Figures 3B and 3C; Movie S2).

Because the plasma membrane is separated from the cyto-plasm by the actin-rich cortex, we hypothesized that actin

structures form a barrier preventing MT tip-plasma membraneinteractions. Indeed, triple color live cell TIRF imaging showedthat actin structures could form obstacles blocking progres-sion of AMER2-positive EB3 comets and that these cometswere present in regions with lower actin abundance than theAMER2-negative ones (Figures 3D–3F). Actin depolymeriza-tion with latrunculin B dramatically increased the number ofEB3-mRFP comets that were AMER2-positive (Figures 3Gand 3H; Movie S3), providing direct support for the view thatactin-rich cortex can prevent MT interaction with the plasmamembrane. AMER2 can potentially function as a MT-plasmamembrane linker and can also be a useful reporter of MT-plasma membrane contacts.Another +TIP implicated in membrane functions identified in

our screen was syntabulin, a protein involved in transport ofmitochondria and synaptic cargo in neurons [18] (Figure 3A;Figure S2M). In cultured cells, GFP-syntabulin was enrichedat the mitochondria (Figure S2N) and concentrated at the siteswhere MT tips colocalized with mitochondria (Figure 3I). Inter-action of syntabulin with growing MT ends could potentiallypromote cargo loading and transport by MT-based motors.

+TIPs Involved in Signaling, Actin-Binding,

and Other FunctionsWe identified three EB-binding kinases, MARK1 [19], TTBK1[20], and TTBK2 [21] (Figure 4A; Figure S1F). WhereasMARK1 showed no plus-end localization (data not shown),TTBK1, a kinase previously shown to phosphorylate MT-stabi-lizing protein tau [20], and its homolog TTBK2 robustly trackedgrowing MT ends in a manner independent of their kinaseactivity (Figure S3A; Figure 4B). At high expression levels,TTBK1 decorated and bundled MTs, while TTBK2 wasdiffusely located in the cytosol, a difference likely due to auto-phosphorylation of TTBK2, because both kinase deadmutantsdecorated MTs (Figures S3B–S3D). Overexpressed TTBK2displaced EB1 from theMTs (Figure S3B), suggesting a poten-tial role of TTBKs in regulating the association of +TIPs andother MT-associated proteins, such as tau, with freshly poly-merized MT ends.We also identified a tip-tracking small GTPase, RasL11B,

which was previously implicated in embryonic developmentin zebrafish [22] (Figures 4A and 4C). Interestingly, its closehomolog RasL11A contains a highly similar SxIP motif butcould neither track growing MT ends (Figure 4C) nor bind toEB1 (Figure S1F), possibly due to an unfavorable sequencecontext of the SxIP motif (Figure S3E).An important +TIP function is formation of molecular

links between MT and actin networks. Among the newlyidentified +TIPs, GAS2L1 and GAS2L2 proteins (Figure 4A)were already implicated in MT and actin binding [23]. Wefound that both proteins could track growing MT ends at lowexpression levels (Figure 4D), colocalized with actin fibers atelevated expression levels, and could promote MT-actin coal-ignment, especially in the presence of elevated levels of EBs(Figure 4E; Figure S3F). These results suggest that, similar tospectraplakins [24, 25], GAS2L1 and GAS2L2 proteins canpotentially promote interactions of dynamic MTs with theactin cytoskeleton.Other newly identified +TIPs, such as filamin-destabilizing

protein FILIP1 [26], cortactin-binding protein 2 [27], and Sickletail [28], the paralogue of a neuronal actin regulator p140Cap[29, 30], are also likely involved in MT-actin crosstalk (FiguresS4A and S4B). Among the remaining +TIPs identified here, themajority represented poorly characterized proteins containing

Page 5: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Figure 3. Membrane-Associated +TIPs

(A) Schemes of the domain structure of EB binding proteins. Abbreviations: CC, coiled coil.

(B) Live images of GFP-AMER2 and EB3-mRFP obtained by TIRF microscopy in MCF7 cells. The images are acquired with 500 ms exposure time at the

indicated time points. Red arrowhead represents an EB3 comet that acquires the GFP signal as it comes into contact with the membrane, yellow arrowhead

represents an EB3 comet that leaves the membrane, and green arrowhead represents an EB3 comet that does not interact with the membrane. See also

Movie S2.

(C) Two color kymographs illustrating the accumulation of GFP-AMER2 (green) at the EB3-mRFP-positive MT ends (red). Note that the growth rate of EB3-

mRFP labeled MT tips was significantly reduced during the periods when they were strongly decorated with GFP-AMER2.

(D and E) AMER2 colocalizes with EB3 in cortical regions with low actin accumulation. Images were obtained by TIRF microscopy in U2OS cells expressing

the indicated constructs (D). The arrow indicates AMER2-positive EB3 comets at the cell edge outside of the actin-enriched areas. The inset shows that an

enlargement of an AMER2-positive growing MT plus end encountering an actin fiber, which prevents its progression as shown in the kymograph in (E).

(F) Quantification of actin intensity at the sites corresponding to EB3 and AMER2 double-positive MT tips and EB3-positive but AMER2-negative comets.

The intensities were normalized by themean actin intensity in each cell after background subtraction. Values are significantly different as indicated by aster-

isks (***p < 0.001).

(G and H) Actin depolymerization increases the attachment between membrane and dynamic MTs. Images were obtained in U2OS cells expressing GFP-

AMER2, EB3-TagBFP, andmCherry-actin by TIRFmicroscopy. (G) Before adding actin depolymerizing drug latrunculin B (2 mM), only a few AMER2-positive

comets can be found at the edge of cell (20:05’, green arrowheads). Actin depolymerization coincides with the increase in the number of AMER2-positive

EB3 comets at the plasmamembrane (see alsoMovie S3). The cell edge is indicated by red stippled line. (H) A representative curve showing quantification of

the percentage of AMER2-positive EB3 labeled growing MT ends before and after latrunculin B treatment in U2OS cells.

(I) Live cell images of GFP-syntabulin and EB3-mRFP in HeLa cells. The images are acquired with 500 ms exposure time at the indicated time points. Two

EB3 comets aremarked by an arrow and arrowhead, respectively. GFP-syntabulin colocalizes with the rear half of the EB3 comet in the last frame (white and

red arrowheads).

Current Biology Vol 22 No 191804

Page 6: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Figure 4. +TIPs Involved in Signaling and Actin-Binding

(A) Schemes of the domain structure of EB binding proteins. Abbreviations: CH, calponin homology (actin binding domain in GAS2L1 and GAS2L2 proteins);

GAR, GAS2 related (MT-binding domain in GAS2L1 and GAS2L2 proteins); Note that in MARK1, the SxIP motifs are located next to the kinase domain and in

the middle of the protein, which might sterically hinder the interaction with MTs.

(B) Images of fixed U2OS cells expressing the indicatedGFP fusion proteins. All constructs show clear plus-end localization at low expression levels. Kinase

dead (KD) mutants were generated by introducing K63R or K50R mutations in TTBK1 or TTBK2, respectively.

(C and D) Live cell images of the indicated GFP fusion proteins in COS-7 cells prepared in the same way as in Figures 2F and 2G.

(E) Images of fixed COS-7 cells coexpressing the indicated GFP fusion proteins and EB3-TagBFP. The cells that are cotransfected with GFP-GAS2L1 and

GAS2L2 and EB3-TagBFP show strong overlap between F-actin and EB3-decoratedMTs, indicating that GAS2L1 and GAS2L2 can promote EB-dependent

interactions of growing MT tips with the actin network or MT guidance along actin fibers.

Proteome-wide Screen for +TIPs with an SxIP Motif1805

no conspicuous domains with the exception of coiled coils(Figures S4A and S4B).

EB Proteins Display Different Affinity for SxIP-Containing +TIPs

Previous studies showed that EB2 displayed lower affinity forseveral partners compared to EB1 and EB3 [31–33]. To quan-tify the differences in binding partners between the three EBs,we focused on the results of biotinylation-tag-based pull-downs. In this experiment, the three EBs were expressed ineukaryotic cells at similar levels and isolated together withtheir partners in a single-step procedure, making biases dueto protein purity, stability, or posttranslational modificationsless likely. Although the amounts of isolated EB1, EB2, andEB3 were very similar (Figure S1B), more +TIPs were identifiedin the EB1 and EB3 pull-downs than in the EB2 sample, andmost of the ones identified in all three samples were least

abundant in the EB2 sample (Figures S4C and S4D),whereas +TIP abundance in EB1 and EB3 pull-downs wassimilar. These data confirm the view that EB2 binds to SxIPmotifs with a lower affinity and that its functions might bedistinct from those of EB1 and EB3 [34].

Discussion

By using a combination of mass spectrometry and bio-informatics, we have identified a broad group of previouslyunknown EB partners that can track growing MT ends. Thenumber of SxIP-containing +TIPs will likely increase further,because some of the newly identified +TIPs were part ofa considerable and not yet fully tested protein list based exclu-sively on the bioinformatics search. However, it is unlikely thata large new group of EB-binding +TIPs will be found: themajority of the proteins identified as potential EB partners by

Page 7: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Current Biology Vol 22 No 191806

mass spectrometry either contain the canonical SxIP or CAP-Gly motifs or bind to EBs through SxIP or CAP-Gly proteins.SxIP thus represents the major MT tip localization sequence(MtLS) (Figure S4E). Global quantification of protein expres-sion [35] showed that EB1 is significantly more abundantthan any of the SxIP partners (see legend to Table S1E),suggesting that it can recruit multiple proteins to MT tipssimultaneously.

Do all EB partners that can track MT ends when expressedas GFP fusions represent true +TIPs? Imaging experimentssuggest that when a given EB partner can diffuse withina compartment accessible to growing MT ends, it will concen-trate on these ends. If a protein is tightly bound to anothercellular structure, it will show no plus-end tracking behavior;however, its interaction with EBs can still be relevant toconnect growing MT ends to other cellular structures. There-fore, the really important question is whether the particularSxIP motif-mediated interaction with EB-bound MT tips isfunctionally relevant. For SxIP proteins as a whole, this ques-tion can be addressed by replacing the cellular pool of EBproteins with an EB mutant containing substitutions in theC-terminal domain that disrupt the SxIP-EB interaction [36].For each individual SxIP protein, substitution of the wild-type protein with an SxIP motif mutant (for example, byexchanging the isoleucine and proline residues to aspara-gines; [3]) can provide information on the functional signifi-cance of its binding to EB.

A key outcome of this study is the demonstration that +TIPsare a significantly larger group of proteins than previouslythought. These proteins will need to be taken into accountwhen searching for the molecular basis of complex cytoskel-eton-based phenomena. In a broader context, the approachesused in this study can be applied to other classes of proteinmotifs to discover their proteome-wide occurrence and theircontribution to the global protein interaction networks.

Supplemental Information

Supplemental Information includes four figures, one table, Supplemental

Experimental Procedures, and three movies and can be found with this

article online at http://dx.doi.org/10.1016/j.cub.2012.07.047.

Acknowledgments

We are grateful to C. Schneider, R. Tsien, and Kazusa DNA Research Insti-

tute for sharingmaterials and toM. Singh for generation of GFP-FILIP1. This

study was supported by the Netherlands Organization for Scientific

Research ALW open program, ALW-VICI and FOM program grant, Human

Frontier Science Program grant and Erasmus MC grant to A.A, and by

Fundacao para a Ciencia e a Tecnologia fellowship to S.M.G. M.O.S. is sup-

ported by grants from the Swiss National Science Foundation. T.J.G. was

supported by EU FP7 grant Syscilia. N.E.D. was supported by a European

Molecular Biology Laboratory (EMBL) Interdisciplinary Postdoc (EIPOD)

fellowship from EMBL. L.B.P. is supported by grants from the Danish

Natural Science Research Council (272-05-0411 and 09-070398). J.L. was

supported by a Novo Nordisk Scholarship.

Received: April 22, 2012

Revised: July 10, 2012

Accepted: July 23, 2012

Published online: August 9, 2012

References

1. Schuyler, S.C., and Pellman, D. (2001). Microtubule ‘‘plus-end-tracking

proteins’’: The end is just the beginning. Cell 105, 421–424.

2. Akhmanova, A., and Steinmetz, M.O. (2010). Microtubule +TIPs at

a glance. J. Cell Sci. 123, 3415–3419.

3. Honnappa, S., Gouveia, S.M., Weisbrich, A., Damberger, F.F., Bhavesh,

N.S., Jawhari, H., Grigoriev, I., van Rijssel, F.J., Buey, R.M., Lawera, A.,

et al. (2009). An EB1-binding motif acts as a microtubule tip localization

signal. Cell 138, 366–376.

4. Buey, R.M., Sen, I., Kortt, O., Mohan, R., Gfeller, D., Veprintsev, D.,

Kretzschmar, I., Scheuermann, J., Neri, D., Zoete, V., et al. (2012).

Sequence determinants of a microtubule tip localization signal (MtLS).

J. Biol. Chem. Published online June 13, 2012. http://dx.doi.org/10.

1074/jbc.M112.373928.

5. Hsieh, P.C., Chang, J.C., Sun, W.T., Hsieh, S.C., Wang, M.C., andWang,

F.F. (2007). p53 downstream target DDA3 is a novel microtubule-asso-

ciated protein that interacts with end-binding protein EB3 and activates

beta-catenin pathway. Oncogene 26, 4928–4940.

6. Utrera, R., Collavin, L., Lazarevi�c, D., Delia, D., and Schneider, C. (1998).

A novel p53-inducible gene coding for a microtubule-localized protein

with G2-phase-specific expression. EMBO J. 17, 5015–5025.

7. Yang, S., Liu, X., Yin, Y., Fukuda, M.N., and Zhou, J. (2008). Tastin is

required for bipolar spindle assembly and centrosome integrity during

mitosis. FASEB J. 22, 1960–1972.

8. Fukuda, M.N., Sato, T., Nakayama, J., Klier, G., Mikami, M., Aoki, D., and

Nozawa, S. (1995). Trophinin and tastin, a novel cell adhesion molecule

complex with potential involvement in embryo implantation. Genes Dev.

9, 1199–1210.

9. Jang, C.Y., Wong, J., Coppinger, J.A., Seki, A., Yates, J.R., 3rd, and

Fang, G. (2008). DDA3 recruits microtubule depolymerase Kif2a to

spindle poles and controls spindle dynamics and mitotic chromosome

movement. J. Cell Biol. 181, 255–267.

10. Ferenz, N.P., Gable, A., and Wadsworth, P. (2010). Mitotic functions of

kinesin-5. Semin. Cell Dev. Biol. 21, 255–259.

11. Uteng, M., Hentrich, C., Miura, K., Bieling, P., and Surrey, T. (2008).

Poleward transport of Eg5 by dynein-dynactin in Xenopus laevis egg

extract spindles. J. Cell Biol. 182, 715–726.

12. Gardner, M.K., Bouck, D.C., Paliulis, L.V., Meehl, J.B., O’Toole, E.T.,

Haase, J., Soubry, A., Joglekar, A.P., Winey, M., Salmon, E.D., et al.

(2008). Chromosome congression by Kinesin-5 motor-mediated disas-

sembly of longer kinetochore microtubules. Cell 135, 894–906.

13. Jakobsen, L., Vanselow, K., Skogs, M., Toyoda, Y., Lundberg, E., Poser,

I., Falkenby, L.G., Bennetzen, M., Westendorf, J., Nigg, E.A., et al.

(2011). Novel asymmetrically localizing components of human centro-

somes identified by complementary proteomics methods. EMBO J.

30, 1520–1535.

14. Spektor, A., Tsang, W.Y., Khoo, D., and Dynlacht, B.D. (2007). Cep97

and CP110 suppress a cilia assembly program. Cell 130, 678–690.

15. Brito, D.A., Gouveia, S.M., and Bettencourt-Dias, M. (2012).

Deconstructing the centriole: structure and number control. Curr.

Opin. Cell Biol. 24, 4–13.

16. Schrøder, J.M., Larsen, J., Komarova, Y., Akhmanova, A.,

Thorsteinsson, R.I., Grigoriev, I., Manguso, R., Christensen, S.T.,

Pedersen, S.F., Geimer, S., and Pedersen, L.B. (2011). EB1 and EB3

promote cilia biogenesis by several centrosome-related mechanisms.

J. Cell Sci. 124, 2539–2551.

17. Grohmann, A., Tanneberger, K., Alzner, A., Schneikert, J., and Behrens,

J. (2007). AMER1 regulates the distribution of the tumor suppressor

APC between microtubules and the plasma membrane. J. Cell Sci.

120, 3738–3747.

18. Ma, H., Cai, Q., Lu,W., Sheng, Z.H., andMochida, S. (2009). KIF5Bmotor

adaptor syntabulin maintains synaptic transmission in sympathetic

neurons. J. Neurosci. 29, 13019–13029.

19. Drewes, G., Ebneth, A., Preuss, U., Mandelkow, E.M., and Mandelkow,

E. (1997). MARK, a novel family of protein kinases that phosphorylate

microtubule-associated proteins and trigger microtubule disruption.

Cell 89, 297–308.

20. Sato, S., Cerny, R.L., Buescher, J.L., and Ikezu, T. (2006). Tau-tubulin

kinase 1 (TTBK1), a neuron-specific tau kinase candidate, is involved

in tau phosphorylation and aggregation. J. Neurochem. 98, 1573–1584.

21. Houlden, H., Johnson, J., Gardner-Thorpe, C., Lashley, T., Hernandez,

D., Worth, P., Singleton, A.B., Hilton, D.A., Holton, J., Revesz, T., et al.

(2007). Mutations in TTBK2, encoding a kinase implicated in tau

phosphorylation, segregate with spinocerebellar ataxia type 11. Nat.

Genet. 39, 1434–1436.

22. Pezeron, G., Lambert, G., Dickmeis, T., Strahle, U., Rosa, F.M., and

Mourrain, P. (2008). Rasl11b knock down in zebrafish suppresses

one-eyed-pinhead mutant phenotype. PLoS ONE 3, e1434.

Page 8: A Proteome-wide Screen for Mammalian SxIP Motif-Containing Microtubule Plus-End Tracking Proteins

Proteome-wide Screen for +TIPs with an SxIP Motif1807

23. Goriounov, D., Leung, C.L., and Liem, R.K. (2003). Protein products of

human Gas2-related genes on chromosomes 17 and 22 (hGAR17 and

hGAR22) associate with both microfilaments and microtubules. J. Cell

Sci. 116, 1045–1058.

24. Applewhite, D.A., Grode, K.D., Keller, D., Zadeh, A.D., Slep, K.C., and

Rogers, S.L. (2010). The spectraplakin Short stop is an actin-microtu-

bule cross-linker that contributes to organization of the microtubule

network. Mol. Biol. Cell 21, 1714–1724.

25. Kodama, A., Karakesisoglou, I., Wong, E., Vaezi, A., and Fuchs, E.

(2003). ACF7: an essential integrator of microtubule dynamics. Cell

115, 343–354.

26. Nagano, T., Yoneda, T., Hatanaka, Y., Kubota, C., Murakami, F., and

Sato, M. (2002). Filamin A-interacting protein (FILIP) regulates cortical

cell migration out of the ventricular zone. Nat. Cell Biol. 4, 495–501.

27. Ohoka, Y., and Takai, Y. (1998). Isolation and characterization of cortac-

tin isoforms and a novel cortactin-binding protein, CBP90. Genes Cells

3, 603–612.

28. Semba, K., Araki, K., Li, Z., Matsumoto, K., Suzuki, M., Nakagata, N.,

Takagi, K., Takeya, M., Yoshinobu, K., Araki, M., et al. (2006). A novel

murine gene, Sickle tail, linked to the Danforth’s short tail locus, is

required for normal development of the intervertebral disc. Genetics

172, 445–456.

29. Di Stefano, P., Cabodi, S., Boeri Erba, E., Margaria, V., Bergatto, E.,

Giuffrida, M.G., Silengo, L., Tarone, G., Turco, E., and Defilippi, P.

(2004). P130Cas-associated protein (p140Cap) as a new tyrosine-

phosphorylated protein involved in cell spreading. Mol. Biol. Cell 15,

787–800.

30. Jaworski, J., Kapitein, L.C., Gouveia, S.M., Dortland, B.R., Wulf, P.S.,

Grigoriev, I., Camera, P., Spangler, S.A., Di Stefano, P., Demmers, J.,

et al. (2009). Dynamicmicrotubules regulate dendritic spinemorphology

and synaptic plasticity. Neuron 61, 85–100.

31. Bu, W., and Su, L.K. (2003). Characterization of functional domains of

human EB1 family proteins. J. Biol. Chem. 278, 49721–49731.

32. Komarova, Y., Lansbergen, G., Galjart, N., Grosveld, F., Borisy, G.G.,

and Akhmanova, A. (2005). EB1 and EB3 control CLIP dissociation

from the ends of growing microtubules. Mol. Biol. Cell 16, 5334–5345.

33. Lee, T., Langford, K.J., Askham, J.M., Bruning-Richardson, A., and

Morrison, E.E. (2008). MCAK associates with EB1. Oncogene 27,

2494–2500.

34. Komarova, Y., De Groot, C.O., Grigoriev, I., Gouveia, S.M., Munteanu,

E.L., Schober, J.M., Honnappa, S., Buey, R.M., Hoogenraad, C.C.,

Dogterom, M., et al. (2009). Mammalian end binding proteins control

persistent microtubule growth. J. Cell Biol. 184, 691–706.

35. Schwanhausser, B., Busse, D., Li, N., Dittmar, G., Schuchhardt, J., Wolf,

J., Chen, W., and Selbach, M. (2011). Global quantification of mamma-

lian gene expression control. Nature 473, 337–342.

36. Montenegro Gouveia, S., Leslie, K., Kapitein, L.C., Buey, R.M.,

Grigoriev, I., Wagenbach, M., Smal, I., Meijering, E., Hoogenraad,

C.C., Wordeman, L., et al. (2010). In vitro reconstitution of the functional

interplay between MCAK and EB3 at microtubule plus ends. Curr. Biol.

20, 1717–1722.