Homework

Modes of Caldesmon Binding to Actin: SITES OF CALDESMON CONTACT AND MODULATION OF INTERACTIONS BY PHOSPHORYLATION

Description
Modes of Caldesmon Binding to Actin: SITES OF CALDESMON CONTACT AND MODULATION OF INTERACTIONS BY PHOSPHORYLATION
Categories
Published
of 8
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
Related Documents
Share
Transcript
  Modes of Caldesmon Binding to Actin SITES OF CALDESMON CONTACT AND MODULATION OF INTERACTIONS BY PHOSPHORYLATION* Received for publication, September 2, 2004, and in revised form, September 15, 2004Published, JBC Papers in Press, September 27, 2004, DOI 10.1074/jbc.M410109200 D. Brian Foster‡§ ¶  , Renjian Huang‡ ¶ , Victoria Hatch§, Roger Craig**, Philip Graceffa‡,William Lehman§, and C.-L. Albert Wang‡ ‡‡  From the  ‡  Boston Biomedical Research Institute, Watertown, Massachusetts 02472, the  §  Department of Physiology and Biophysics, Boston University School of Medicine, Boston, Massachusetts 02118, and the  **  Department of Cell Biology,University of Massachusetts Medical School, Worcester, Massachusetts 01655 Smooth muscle caldesmon binds actin and inhibitsactomyosin ATPase activity. Phosphorylation of caldesmon by extracellular signal-regulated kinase(ERK) reverses this inhibitory effect and weakens ac-tin binding. To better understand this function, wehave examined the phosphorylation-dependent con-tact sites of caldesmon on actin by low dose electronmicroscopy and three-dimensional reconstruction of actin filaments decorated with a C-terminal fragment,hH32K, of human caldesmon containing the principalactin-binding domains. Helical reconstruction of neg-atively stained filaments demonstrated that hH32K islocated on the inner portion of actin subdomain 1,traversing its upper surface toward the C-terminalsegment of actin, and forms a bridge to the neighbor-ing actin monomer of the adjacent long pitch helicalstrand by connecting to its subdomain 3. Such lateralbinding was supported by cross-linking experimentsusing a mutant isoform, which was capable of cross-linking actin subunits. Upon ERK phosphorylation,however, the mutant no longer cross-linked actin topolymers. Three-dimensional reconstruction of ERK-phosphorylated hH32K indeed indicated loss of theinterstrand connectivity. These results, together withfluorescence quenching data, are consistent with aphosphorylation-dependent conformational changethat moves the C-terminal end segment of caldesmonnear the phosphorylation site but not the upstreamregion around Cys 595 , away from F-actin, thus neutral-izing its inhibitory effect on actomyosin interactions.The binding pattern of hH32K suggests a mechanismby which unphosphorylated, but not ERK-phosphoryl-ated, caldesmon could stabilize actin filaments andresist F-actin severing or depolymerization in bothsmooth muscle and nonmuscle cells. Caldesmon (CaD) 1 is an actin-binding protein found in bothnonmuscle and smooth muscle cells. In nonmuscle cells it in-fluences contractility by interfering with focal adhesion andstress fiber assembly (1, 2). In smooth muscle, CaD is found onthin filaments within the contractile domain (3) where it sup-presses basal muscle tone by inhibiting active cross-bridgecycling (4), providing fine tuning of the contractility underdiverse physiological conditions. The mechanism by which CaDimpinges on smooth muscle contractility and whether CaDfunction is subject to regulation  in vivo , however, remain con-tentious issues (5–9).Much of the structural information regarding CaD has beengarnered from the study of the smooth muscle isoform, h-CaD,which was srcinally identified as a calmodulin (CaM)- bindingprotein that also binds filamentous actin (F-actin) (10). Innative smooth muscle thin filaments, h-CaD binds lengthwisealong the actin filaments with a periodicity of 38 nm (11),although its length (75 nm) (12) is sufficient to span two actinheptads. This is most likely due to staggered binding of h-CaDto the two actin strands (13). Biochemical studies of purifiedh-CaD demonstrate that it has three functionally distinct do-mains: an N-terminal domain that harbors the major myosin-binding sites (14–17), a rigid   -helical middle domain that isabsent in the nonmuscle isoform, l-CaD (18–20), and a C-terminal domain that houses binding sites for actin (21–24),tropomyosin (Tm) (25, 26), and CaM (27, 28). It is the C-terminal actin-binding domain that blocks the weak binding of myosin and inhibits actomyosin ATPase activity  in vitro  (22–24, 29), as well as force development in Triton-skinned smoothmuscle fibers when added exogenously (29).Regulation of CaD function has been studied extensively  invitro . In the presence of Ca 2  , CaM reverses the binding of CaDto actin (10) and therefore the inhibitory effect of CaD on theactomyosin interaction (30, 31). The affinity between CaD andCaM, however, is only moderate (  10 6 M  1 ) (32). Although ithas been shown that sufficiently high local intracellular con-centrations of CaM do exist in both smooth muscle (33) andnonmuscle cells (2) to allow CaD to interact with CaM  in vivo ,whether such an interaction plays a physiological role stillremains a point of controversy. Alternatively, CaD can be phos-phorylated by a number of kinases, such as protein kinase C(34, 35), CaM-dependent kinase II (36, 37), casein kinase II(38), cAMP-dependent kinase (39), p34 cdc2 (40–42), and mito- * This work was supported in part by Diabetes Endocrinology Re-search Center Grant DK32520 and National Institutes of HealthGrants RO1-HL36153 (to W. L.), RO1-HL62468 (to R. C.), and PO1- AR41637 (to C.-L. A. W.). The costs of publication of this article weredefrayed in part by the payment of page charges. This article musttherefore be hereby marked “ advertisement ” in accordance with 18U.S.C. Section 1734 solely to indicate this fact. ¶  Both authors contributed equally to this work.   Supported by the Boston Biomedical Research Institute ScholarProgram and by an American Heart Association Postdoctoral Fellow-ship (New England Affiliate). Present address: Institute of MolecularCardiobiology, Johns Hopkins School of Medicine, Ross Research Bldg.,Rm. 858, 720 Rutland Ave., Baltimore, MD 21205.‡‡ To whom correspondence should be addressed: Boston BiomedicalResearch Institute, 64 Grove St., Watertown MA, 02472-2829. Tel.:617-658-7803; Fax: 617-972-1753; E-mail: wang@bbri.org. 1 The abbreviations used are: CaD, caldesmon; BPM, benzophenonemaleimide; CaM, calmodulin; DTT, dithiothreitol; ERK, extracellularsignal-regulated kinase; 1,5-IAEDANS, 5-(iodoacetamidoethyl)amino-naphthalene-1-sulfonic acid; MAPK, mitogen-activated protein kinase;NbS 2 , 5,5  -dithio-bis(2-nitrobenzoic acid); Tm, tropomyosin; PIPES, 1,4-piperazinediethanesulfonic acid; Mops, 4-morpholinepropanesulfonicacid. T HE  J OURNAL OF  B IOLOGICAL  C HEMISTRY   Vol. 279, No. 51, Issue of December 17, pp. 53387–53394, 2004 © 2004 by The American Society for Biochemistry and Molecular Biology, Inc.  Printed in U.S.A. This paper is available on line at http://www.jbc.org  53387   a t   J  oh n s H  o pk i  n s  U ni  v  er  s i   t   y  , onM ar  c h 1  3  ,2  0 1  3 www. j   b  c . or  gD  ownl   o a d  e d f  r  om   gen-activated protein kinase (MAPK or ERK) (43). Phosphoryl-ation, at sites primarily in the C-terminal domain of CaD,mitigates its ability to inhibit actin  Tm-activated myosin ATPase activity (44, 45), thus providing another mechanism toregulate the function of CaD.Evidence for regulation of CaD by phosphorylation  in vivo has come from work on nonmuscle cells and differentiatedsmooth muscle. Matsumura and colleagues (40) showed thatphosphorylation by cdc2 kinase during mitosis caused l-CaD todissociate from microfilaments in proliferating fibroblasts.Working with differentiated smooth muscle, Adam  et al.  (46)demonstrated that  32 P-labeled h-CaD, purified from phorbol12,13-dibutyrate-stimulated canine aortic smooth muscle, wasphosphorylated at sites VTS*PTKV and S*PAPK within its Cterminus (Ser 759 and Ser 789 by the mammalian numberingscheme). Subsequent work has shown that Ser 789 is the pre-ponderate site of h-CaD phosphorylation in porcine carotidartery strips (47). These sequences conform to the consensusmotif S(T)P  X  P that constitutes the preferred target site for thefamily of “proline-directed” kinases, of which cdc2 kinase andERK are prototypes. ERK has been purified from smooth mus-cle (48), and its activation in smooth muscle has been studied(49, 50). Furthermore, it has been shown that Ca 2  -free stim-ulation of ferret aortic smooth muscle cells, with phenyleph-rine, resulted in the recruitment of ERK to the plasma mem-brane, phosphorylation of tyrosine (thereby activating thekinase), and redistribution to CaD-decorated thin filaments(51). Taken together these studies implicate ERK as an endog-enous CaD kinase.Further understanding of the mechanism of action of CaDhas been afforded by electron microscopy and three-dimen-sional helical image reconstruction. Addition of a 150-residueC-terminal CaD fragment, 606C, to reconstituted actin  Tm fil-aments caused Tm to move from its position on the inner aspectof the outer domain of actin, toward the inner domain of actin(52). This indicates that CaD affects the conformation of actin  Tm differently than the striated muscle regulatory pro-tein, troponin. Subsequent studies of optimally negativelystained native chicken gizzard thin filaments revealed densityon the outer domain of actin on subdomains 1 and 2 that wasattributed to CaD (53). However, image density was weak,likely because of incomplete saturation of actin filaments,which may have resulted from partial dissociation of CaD dur-ing the purification process. Thus although the difference den-sity map was generated between CaD-bound and CaD-freefilaments (after incubation with Ca 2   /CaM), ambiguity re-mains with regard to the assignment of the binding positionof CaD.To determine the conformation of the C-terminal domain of CaD on purified actin, such that contact regions on F-actincould be assigned unambiguously, and to test whether phos-phorylation of this region by ERK, a physiologically relevantevent, alters its conformation on actin, we have undertaken thepresent study. Our data reveal actin-CaD contacts that havenot been detected previously and demonstrate that phospho-rylation affects the conformation of actin-bound CaD. Thesereconstructed images, corroborated by results from fluores-cence quenching and cross-linking experiments, support amodel where the C-terminal region of CaD interacts withactin via two clusters of contact points, one of which dissoci-ates from actin upon phosphorylation, resulting in the loss of inhibition on actomyosin interaction (44). Because this C-terminal domain is shared by both CaD isoforms, the ob-served conformational change may serve as a common mech-anism for regulating the function of CaD in smooth muscleand nonmuscle cells. MATERIALS AND METHODS Cloning and Expression of C-terminal Fragments of CaD— The His 6 -tagged C-terminal region of chicken gizzard CaD (H32K, residuesMet 563 –Pro 771 , using the corrected numbering system; see Ref. 54) andits variant, H32Kqc (with Gln 766 mutated to Cys) were prepared asdescribed previously (44). Gln 766 was chosen for mutagenesis because itis in the region of, yet not too close to, the ERK phosphorylation site(Ser 717 ). Another mutant, H32Kqc/ca, in which Cys 595 and GLn 766 aresimultaneously mutated to Ala and Cys, respectively, was prepared bythe same procedure. Thus the wild-type H32K and the double mutantH32Kqc/ca each contain a single cysteine, whereas H32Kqc has twocysteine residues. The mammalian homolog (hH32K) corresponding toresidues Leu 604 –Val 793 of human CaD with a His 6  tag at the N terminuswasexpressedinHigh-FivecellsandpurifiedonaNi 2  columnfollowedby a CaM affinity column (44). As in the previous work, mutagenesis inhH32K was not attempted, because Gln 766 does not exist in the mam-malian sequence, and there is no suitable mutation site near the phos-phorylation sites.  ERK Phosphorylation of C-terminal Fragments of CaD— Phosphoryl-ation of both chicken and human CaD fragments was carried out usingpurified proteins and recombinant ERK2 (New England BioLabs, Inc.)in the manufacturer-supplied 1   MAPK buffer (50 m M  Tris-HCl, pH7.5, 10 m M  MgCl 2 , 1 m M  EDTA, 2 m M  DTT, and 0.01% Brij35), andascertained by mass spectrometric analysis as described previously(44). Although ERK phosphorylates hH32K at both Ser 759 and Ser 789 ,H32K is only phosphorylated at Ser 717 (which corresponds to Ser 759 inthe mammalian sequence), because the other site is absent in thechicken sequence.  Sample Preparation for Electron Microscopy— Filamentous rabbitskeletal actin (5   l of 1   M ; prepared as described in Ref. 55) in 5 m M PIPES, pH 7.5, 50 m M  KCl, 3.5 m M  MgCl 2 , 0.1 m M  EGTA, 0.02% NaN 3 and 0.5 m M  DTT was applied to carbon-coated microscope grids. Theactin solution was wicked down to a volume of   0.5   l, allowing F-actinto adsorb weakly to the grid surface before a solution of hH32K (5   l of   5   M ; in 20 m M  Tris-HCl, pH 7.5, 50 m M  NaCl, 1 m M  DTT, 1 m M phenylmethanesulfonyl fluoride, and 5   M  leupeptin) was added to thegrid. The grids were then allowed to stand for 5–15 min at roomtemperature (22 °C) in a chamber, maintained at a relative humidity of 70–80% to minimize sample evaporation prior to staining. The sampleswere stained with 1% uranyl acetate. This method, which involvespartial adsorption of F-actin to the grid surface, thus restricting free-dom of F-actin movement, was used to circumvent bundling of actin bythe CaD fragment that occurred when the proteins were simply mixedtogether and applied to the grids. Inclusion of 0.5–1.0 m M  DTT in thebuffers had no effect on actin bundling, and attempts to minimizebundling by increasing the ionic strength binding weakened binding of hH32K concomitantly, confounding the search for decorated filaments.  Electron Microscopy and Image Reconstruction— Electron micro-graph images of decorated filaments were recorded on a Philips CM120electron microscope at 60,000   magnification under low dose condi-tions (12 e   /Å  2 ). The micrographs were digitized using a SCAI scannerat a pixel size corresponding to 0.7 nm in the filaments (56). In thecurrent study, filaments were chosen for analysis if the stain surround-ing them was well spread and even and if the filaments lacked distor-tions, discontinuities, or overlying contaminants. Areas displayingastigmatism or specimen drift were not processed, and curved filamentswere straightened by applying spline-fitting algorithms (57). Helicalreconstruction was carried out using standard methods (58–60) asdescribed previously (61, 62). Layer line data extended to a resolution of   25–30 Å, and no data were collected beyond 23 Å. The maps of actin-hH32K and actin-phospho-hH32K filaments were each generatedby calculating the average amplitudes and phases along layer lines of Fourier transforms determined for 19 filaments from two hH32K andtwo phospho-hH32K preparations. Maps of individual filaments wereaveraged after aligning them to each other by iterative rotation andtranslation in reciprocal space to attain a common phase srcin (63).  Photo-cross-linking Experiments— Cross-linking between CaD andactin was achieved by using a photo-cross-linker, benzophenone male-imide (BPM). To protect the photo-sensitive reagent, all of the photo-cross-linking experiments were performed in the dark. Both the phos-phorylated (by ERK2 for   4 h at room temperature) andunphosphorylated H32K fragments were first reduced with 10 m M  DTTfor 1 h at room temperature and extensively dialyzed to remove DTTagainst20m M Tris-HClbuffer,pH7.5, 50m M  NaCl,1m M  EDTA.Tothesample 5-fold molar excess of BPM was added from a 20 m M  stocksolution in dimethylformamide, and the mixture was rotated for 5 h atroom temperature. The reaction was quenched with 5 m M  DTT, and the  Modulation of Caldesmon Binding to Actin by ERK  53388   a t   J  oh n s H  o pk i  n s  U ni  v  er  s i   t   y  , onM ar  c h 1  3  ,2  0 1  3 www. j   b  c . or  gD  ownl   o a d  e d f  r  om   reaction mixture was dialyzed against 20 m M  Tris-HCl buffer, pH 7.5,50 m M  NaCl.BPM-labeled H32K fragments and actin were mixed typically in a 1:5ratio in F-buffer (50 m M  NaCl 0.2 m M  CaCl 2 , 0.4 m M  ATP, 2 m M  MgCl 2 ,2 m M  DTT, 2 m M  HEPES, pH 7.5). Ultraviolet irradiation was carriedout in a Rayonet RPR-100 photochemical reactor equipped with sixteen3500 lamps (Southern New England Ultraviolet, Hamden, CT) at 4 °Cfor 15 min, and the thin filaments were centrifuged at 85,000 rpm for 30min at 4 °C. The cross-linked products in both pellet and supernatantfractions were analyzed with 10% or 4–20% gradient SDS-polyacryl-amide gels (Bio-Rad). The apparent molecular mass of the gel bandswas calculated using the mobility of the molecular mass markers (Bio-Rad) on the same gel as standards.  Disulfide Cross-linking Experiments— To disulfide cross-link H32K mutants to actin, we have made use of the ability to cross-link actinCys 374 to CaD Cys 595 with the reagent NbS 2  (64). NbS 2  can catalyzedisulfide bond formation between two nearby thiol groups by means of disulfide exchange. G-actin Cys 374 was first activated by reacting withNbS 2  as described previously (64, 65), except that G-actin monomer wasused in place of filamentous F-actin. The resulting NbS-G-actin wasthen polymerized to F-actin by adding NaCl to 40 m M  and MgCl 2  to 2m M  (F-buffer). Unphosphorylated or ERK2-phosphorylated CaD frag-ments (H32K, H32Kqc, and H32Kqc/ca) were reduced with 10 m M  DTTandthenexhaustivelydialyzedagainstabuffercontaining40m M NaCl,5 m M  Mops, pH 7.5, 0.2 m M  EDTA, and 0.01% NaN 3.  The disulfidereaction between NbS-F-actin (  14   M ) and CaD fragments, togetherwith gizzard smooth muscle Tm, was carried out at room temperaturein F-buffer with a molar ratio of 1:2:14 CaD fragment:Tm:actin. Thereaction was quenched at specific times with 2 m M  N  -ethylmaleimide toblock all available cysteine residues. The reaction products were sepa-rated on SDS-PAGE with the running gel containing 2 m M  CaCl 2 , whichresults in the resolution of the   Tm band from actin (66, 67). The bandsof the cross-linked products were excised from the gel, incubated with100 m M  DTT, and reapplied to SDS-PAGE. A reaction mixture withoutCaD was used as a control. Quenching Experiments— Unphosphorylated and ERK-phosphoryl-ated H32K fragments were first treated with 10 m M  DTT for 1 h at roomtemperature and extensively dialyzed to remove DTT against 20 m M Tris-HCl buffer, pH 7.5, 50 m M  NaCl, 1 m M  EDTA. A 5-fold molar excessof 1,5-IAEDANS was added from a 20 m M  stock solution in dimethyl-formamide, and the samples were rotated for 4 h at room temperature.The reaction was quenched with 5 m M  DTT, and the samples weredialyzed against 20 m M  Tris-HCl buffer, pH 7.5, 50 m M  NaCl. Afterlabeling, the H32K fragments were mixed with F-actin in F-buffer. Aliquots of acrylamide solution were then added to the mixture, andthe fluorescence intensity was measured in a 1-cm-path length cuvette(  exc  335 nm;   em  494 nm). Analysis was done with KaleidaGraghsoftware. RESULTS  Electron Microscopy of F-actin-hH32K Complexes— F-actinwas complexed with a polypeptide containing the C-terminal189 residues of human h-CaD (hH32K), under conditions tomaximize saturation of F-actin filaments with the protein.Electron micrographs of negatively stained filaments showedthat hH32K caused F-actin to form tight bundles. Bundlingwas minimized, but not eliminated, by applying F-actin to thesample grids prior to incubation with hH32K or phospho-hH32K (see “Materials and Methods”). Only unbundled fila-ments were analyzed. Actin substructure, although evident,was frequently obscured by the binding of the hH32K on thesurface of filaments (Fig. 1,  b  and  c ), which also caused them toappear wider than pure F-actin. Globular structures were oc-casionally seen projecting from filaments, but details of theshape, orientation, and periodicity of the hH32K were notdiscernable. To detect the hH32K binding and determine itsposition on F-actin, image processing and three-dimensionalreconstruction were therefore necessary. Three-dimensional Reconstructions of Reconstituted Thin Filaments— Filaments bearing hH32K, from two preparations,were negatively stained as described under “Materials andMethods. ” The data arising from different hH32K preparationswere pooled because they were highly similar. Density maps of reconstituted filaments were calculated from the averages of the Fourier transform layer line data (not shown). All of themaps obtained showed typical two-domain actin monomersthat could be further divided into identifiable subdomains 1, 2,3, and 4 (see labeling in Fig. 2 a ). When compared with themaps generated from pure F-actin, each separately calculatedreconstruction of F-actin-hH32K showed obvious extra densitylying on subdomain 1, reaching around the back of the subdo-main and ultimately spanning to the inner domain of theneighboring monomer ( n  1) down in the adjacent long pitchhelical strand of F-actin. Inspection of the surface views aver-aged from 19 hH32K-bearing actin filaments (Fig. 2 b ) showedthat hH32K makes broad contact with subdomain 1, and to aless degree with subdomain 2, with a protuberance of densityon the top edge of subdomain 1. The hH32K density also ex-tends from the backside of subdomain 1 and spans the “inter-strand” gap (Fig. 2,  a  and  c ,  green arrows ) to make contact withsubdomain 3 of the previous actin monomer on the other longpitch helix (Fig. 2 b ,  red ellipse ). hH32K therefore appears tobridge the two strands of the right-handed long pitch helices of F-actin, acting as a “molecular staple.”  Effect of ERK Phosphorylation on Reconstructed Images— Inthe reconstruction of phospho-hH32K-decorated F-actin (Fig.2 c ) averaged over 19 filaments, one sees a number of differ-ences when compared with that of the unphosphorylated sam-ple (Fig. 2 b ). The mass density over subdomain 1 and subdo-main 3   (of the  n    1 actin monomer) shifts more towardsubdomain 3   in such a manner that the “molecular bridge”between adjacent long pitch F-actin strands (Fig. 2 b ) is nolonger visible (Fig. 2, compare  e  with  f  ). Subdomain 1 stillretains some density that is not observed on F-actin alone (Fig.2, compare  d  with  f  ), although it is more diffuse than thatobserved for F-actin-hH32K (Fig. 2 b ), and the protrusion at thetop edge of subdomain 1 disappears. The weaker or more dif-fuse density observed in reconstructions of F-actin-phospho-hH32K filaments may reflect both lower saturation of the fil- F IG . 1.  Electron micrographs of negatively stained filaments. a , rabbit skeletal muscle F-actin alone (two examples).  b , skeletalmuscle F-actin-hH32K (four examples).  c , skeletal muscle F-actin-phos-pho-hH32K (three examples). Note the increased diameter of the deco-rated F-actin.  Bar , 50 nm.  Modulation of Caldesmon Binding to Actin by ERK   53389   a t   J  oh n s H  o pk i  n s  U ni  v  er  s i   t   y  , onM ar  c h 1  3  ,2  0 1  3 www. j   b  c . or  gD  ownl   o a d  e d f  r  om   aments because of weakened binding of phospho-hH32K toF-actin and/or greater flexibility of F-actin-bound phospho-hH32K. Greater flexibility would suggest that part of the phos-phorylated CaD fragment is no longer strongly bound to actinfilaments. Concomitantly, we have observed that phospho-hH32K caused less actin bundling (not shown), a phenomenonconsistent with weakened binding. Cross-linking between CaD and Actin— To test the “staple-like” binding mode of CaD fragment on F-actin biochemically,we have performed cross-linking experiments. The photo-cross-linking results (Fig. 3) showed that H32Kqc (a mutant of thechicken isoform of hH32K with Gln 766 converted to Cys), whichhas two Cys residues at positions 595 and 766, cross-linkedmore than one actin subunit and formed higher order products.In addition to the H32Kqc dimer (  70 kDa) and the 1:1 adduct(at   80 kDa) of H32Kqc and actin, there were also proteinbands, albeit weak, on the gel that could be attributed to suchspecies as H32Kqc 2  actin (  110 kDa), H32Kqc  actin 2  (  120kDa), etc. Notably, the 80-kDa band is a doublet. These twobands may correspond to the cross-linked products through thetwo Cys, or simply two different sites on actin being hit by thecross-linker. The cross-linking yield was only moderate, espe-cially for the high molecular mass products. Photo-cross-link-ers are known to form intramolecular cross-linking; they canalso be quenched by water molecules. Such alternative reactionpathways may explain the observed low yield for intermolecu-lar cross-linking. When the reaction was allowed to last for anhour or longer, a smear of high molecular mass species devel-oped with a concomitant decrease in the 80-kDa species (datanot shown), indicating that H32Kqc acts as a cross-linker tocovalently polymerize actin subunits.The ability of H32Kqc to cross-link two actin monomers wasfurther demonstrated by disulfide cross-linking. When Cys 374 of actin was activated by NbS 2 , cross-linking between H32Kqcand actin occurred instantaneously and almost quantitatively,resulting in two cross-linked species (Fig. 4). When these twoproducts, at 80 kDa ( band A ) and 120 kDa, ( band B ), wereexcised from the gel, reduced with DTT, and applied to SDS-PAGE again, they were resolved into two, and only two, proteinspecies, H32Kqc and actin, the molar ratios between whichbeing close to 1:1 and 1:2, respectively. This clearly and un-equivocally showed that cross-linking did occur and occurredonly between the CaD fragment and actin, in a more completethin filament with Tm also present. The stoichiometry of thereduced protein bands demonstrated that the 120-kDa speciesindeed contained one H32Kqc and two actin monomers. Bothwild-type H32K and the double mutant H32Kqc/ca also formeddisulfide-cross-linked products with actin, each giving rise to asingle species of 80 kDa (similar to  band A  in Fig. 4; data notshown). Although these results are consistent with the ex-pected binding mode, it is somewhat surprising that bothCys 595 and Cys 766 were able to form disulfide linkages withCys 374 of actin, despite the fact that each binds to a separateactin monomer. Clearly, our results indicated that the two F IG . 2.  Surface views of thin filament reconstructions showing the position of hH32K and phospho-hH32K on F-actin, andtransverse sections (z-sections) through maps of three-dimensional reconstructions.  All of the reconstructions were aligned relative toeach other and are directly comparable.  a–c , surface views of F-actin or decorated F-actin.  a , F-actin ( subdomains 1 ,  2 ,  3 , and  4  are labeled).  b ,F-actin-hH32K.  c , F-actin-phospho-hH32K. Note the extra density contributed by hH32K associated with subdomains 1 and 2 of actin in  b  and  c ( open bold arrows ). Also note, in  b , the density that spans from the back of subdomain 1 to subdomain 3 of the previous actin monomer of the genetichelix ( red ellipse ). This interstrand density is present in  b  ( red arrow ) and is absent from  a  and  c  (  green arrow ).  d–f  , transverse sections of F-actinor decorated F-actin. Because adjacent actin monomers on either side of the filament axis are staggered by half a subunit, sectioning through thecenter of subdomains 1 and 3 of one monomer will result in sectioning through subdomains 2 and 4 of the other monomer.  d , F-actin. e,F-actin-hH32K.  f  , F-actin-phospho-hH32K. The  open bold arrows  in  e  indicate regions of significant hH32K density, and the  red arrow  points tothe interstrand density.  Modulation of Caldesmon Binding to Actin by ERK  53390   a t   J  oh n s H  o pk i  n s  U ni  v  er  s i   t   y  , onM ar  c h 1  3  ,2  0 1  3 www. j   b  c . or  gD  ownl   o a d  e d f  r  om   actin-binding clusters of H32K target individual regions on theactin surface that are both close to Cys 374 (see “Discussion”).  Effect of ERK Phosphorylation on Cross-linking between CaDand Actin—  After H32Kqc was phosphorylated by ERK2, thephoto-cross-linked bands of 80 kDa and greater were dimin-ished (Fig. 3), indicating a weakened ability of this CaD frag-ment to bridge two actin monomers. Thus phosphorylationinduces a conformational change in H32Kqc such that one of the two cysteines moves farther away from actin, as depicted inour previously proposed model (44). To determine which cys-teine is affected, we have carried out cross-linking experimentsusing wild-type H32K and the double mutant H32Kqc/ca.The H32K-actin photo-cross-linking results showed thatthere was no obvious difference between phosphorylated andunphosphorylated H32K-BPM (Fig. 5  A ), indicating that phos-phorylation does not significantly affect the environment nearposition 595, which is the only Cys residue in the wild-typeH32K and is relatively far away from the phosphorylation site(Ser 717 ) of CaD. On the other hand, H32Kqc/ca, which containsa single Cys at position 766, showed that the H32Kqc/ca  actincross-linking diminishes after H32Kqc/ca was phosphorylatedby ERK2 (Fig. 5  B ), indicating that ERK phosphorylation af-fects the conformation of the region around Cys 766 .The differential effect of ERK phosphorylation on the sulf-hydryls located in the two halves of the CaD fragment was alsotested by disulfide cross-linking. In good agreement with theresults obtained in the photo-cross-linking experiments, bothH32Kqc and H32Kqc/ca resulted in less cross-linking productswhen they were treated with ERK2 prior to exposure to NbS 2 -activated F-actin, whereas the cross-linking efficiency of thewild-type H32K with actin was not affected by phosphorylation(data not shown). Thus the physical separation of Cys 766 , butnot Cys 595 , from actin is sensitive to phosphorylation of Ser 717 ,reflecting a structural change in the C-terminal end of CaD.  Solvent Accessibility Assessed by Fluorescence Quenching— If ERK treatment indeed causes a conformational change inH32K sufficient to differentially affect the proximity betweenactin and the two cysteine residues, one might expect that theenvironment of these two residues is also changed. To test thiswe have used fluorescence quenching to probe the solvent ac-cessibility of labels attached at these two positions. The twosingle-Cys fragments, H32K and H32Kqc/ca, were labeled with1,5-IAEDANS for this purpose. When the quencher, acrylam-ide, was added to a solution containing F-actin and labeledH32K, the AEDANS fluorescence intensity decreased becauseof collisional quenching. The slope of the Stern-Volmer plot (thereciprocal of fluorescence intensity plotted as a function of thequencher concentration; Fig. 6) reflects the solvent accessibilityof the probe at this position (68). ERK phosphorylated H32K yielded essentially the same slope as that of the unphosphoryl-ated fragment, indicating that the solvent accessibility of Cys 595 is not affected by phosphorylation at Ser 717 . The exper-iment with H32Kqc/ca, however, showed that after ERK2 phos-phorylation, the AEDANS label at Cys 766 became more exposed(with a greater slope in the Stern-Volmer plot; Fig. 6), whereasthe Cys 766 accessibility of the unphosphorylated H32Kqc/ca ismuch more restricted. Thus the region harboring Cys 766 ismore sensitive to ERK2-mediated phosphorylation than thataround Cys 595 . In the unphosphorylated state both Cys 595 andCys 766 are situated in similar environments, but the latterdissociates from F-actin and becomes more exposed to thesolvent after phosphorylation. These results are again consist-ent with the cross-linking results and also agree well with thephosphorylation-induced flexibility observed in three-dimen-sional image reconstruction. F IG . 3.  Photo-cross-linking of H32Kqc-BPM with F-actin. ERK2-phosphorylated ( lanes 3  and  4 ) and unphosphorylated ( lanes 5 and  6 ) H32Kqc labeled with photo-cross-linker BPM was mixed withF-actin and irradiated with UV light.  Lane M  , molecular mass markers; lane 1 , H32Kqc-BPM alone;  lane 2 , actin alone;  lanes 3  and  5 , super-natant fractions of the reaction mixture after cross-linking;  lanes 4  and 6 , pellet fractions of the reaction mixture after cross-linking.F IG . 4.  Disulfide cross-linking of H32Kqc with F-actin  Tm.  Un-labeled H32Kqc was mixed with NbS 2 -activated actin  Tm (see “Materi-als and Methods”).  Lane 1 , H32Kqc alone;  lanes 2–5 , reaction productsof the H32Kqc-F-actin  Tm cross-linking at  t    2, 10, 25, and 60 min,respectively;  lane 6 ,  band A  reduced with DTT;  lane 7  ,  band B  reducedwith DTT;  lane 8 , mixture of H32Kqc and F-actin  Tm plus DTT. Theprotein bands corresponding to actin and H32Kqc in  lanes 6  and  7   werescanned, and the integrated areas for  bands A  and  B  yielded actin/ H32Kqc ratios of 1.19 and 1.96, respectively. Note that there are twospecies in the H32Kqc preparation ( lane 1 ). The faster migrating spe-cies, which did not react with actin ( lanes 2–5 ) and disappeared uponreduction ( lane 8 ), could be an internally oxidized fragment.F IG . 5.  Photo-cross-linking of H32K-BPM (  A ) and H32Kqc/ca-BPM (  B ) with F-actin.  H32K or H32Kqc/ca labeled with BPM wasmixed with F-actin and irradiated with UV light.  Panel A ,  lane M  ,molecular mass markers;  lane 1 , phospho-H32K-BPM with F-actin; lane 2 , H32K-BPM with F-actin;  lane 3 , F-actin alone.  B ,  lane M  ,molecular mass markers;  lane 1 , phospho-H32Kqc/ca-BPM with F-ac-tin;  lane 2 , H32Kqc/ca-BPM with F-actin;  lane 3 , H32Kqc/ca alone;  lane 4 , F-actin alone. Only pellet fractions (except  lane 3  in  B ) are shown.The  arrows  indicate the cross-linked product of CaD fragmentand actin.  Modulation of Caldesmon Binding to Actin by ERK   53391   a t   J  oh n s H  o pk i  n s  U ni  v  er  s i   t   y  , onM ar  c h 1  3  ,2  0 1  3 www. j   b  c . or  gD  ownl   o a d  e d f  r  om 
Search
Similar documents
View more...
Tags
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks