The Structural Basis of the Control of Actin Dynamics by the Gelsolin Superfamily Proteins

Full text

(1)Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Medicine 425. The Structural Basis of the Control of Actin Dynamics by the Gelsolin Superfamily Proteins SAKESIT CHUMNARNSILPA. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2009. ISSN 1651-6206 ISBN 978-91-554-7428-7 urn:nbn:se:uu:diva-89218.

(2) .

(3)

(4)

(5) ! " #

(6) $ "

(7)

(8)

(9) %&"#&"#"' % "

(10) " $(() !

(11) $%

(12) !

(13)

(14)

(15) !*$

(16)

(17) $+,

(18) !" - .$

(19) /

(20) 0 %$ $ 1 () .$1 #

(21) ! $

(22)

(23)

(24) !2 $ 3

(25) 1 ! *

(26) 2 .

(27)

(28)

(29)

(30)

(31) 4( 5 )56)454(65 7 %

(32) ! $

(33) 8

(34)

(35)

(36) !

(37)

(38) /

(39) !

(40) 2

(41) % $ $ %

(42) !

(43) +31*-

(44)

(45)

(46) % 9

(47) %! %! % ! 1 $ %/ $ $31* 8

(48)

(49) % $ !

(50) .$ $ % $

(51) $

(52) ! $ $

(53) !31* $

(54) %$ % % $

(55)

(56) ! %

(57) :

(58) % $ /

(59) $

(60) ! ! %

(61) ! $

(62) .$

(63) $ - .$

(64) ! $ !

(65)

(66) ! $ $ !

(67) ! %

(68)

(69)

(70) !

(71)

(72) / $ $ %

(73) %

(74)

(75) 4 +34- !

(76) $ && % +#1-

(77) ; $

(78) $ & /$$

(79) /$ / 34 <

(80) $ $

(81) !

(82)

(83) (-

(84)

(85) $ & &&#1

(86) !%

(87) / $ % / $ $ 0

(88)

(89) !

(90) !

(91) $ #1 %%

(92)

(93) !

(94)

(95) $ % /$$ / $

(96) /

(97)

(98)

(99) ! $ %

(100) !

(101) !

(102) $ ! !

(103)

(104) !%

(105)

(106)

(107) -.$

(108) %$

(109) ! 3= $

(110)

(111) $$

(112) ! !

(113)

(114) =+>=- 4-.$

(115) !

(116)

(117)

(118) ! $ $ !

(119) ! $ .

(120). $

(121) !%

(122) .$ %% $ $ $

(123) !31* !

(124) !

(125) %

(126) / $ .$

(127)

(128) %$ % $$

(129) !

(130) =/$$ 8

(131)

(132) $%

(133)

(134) !

(135)

(136) $ %

(137) ! $

(138) ! $

(139) .$

(140) %% $

(141)

(142) $ &#1

(143)

(144) 4

(145) ! $

(146) !

(147)

(148)

(149) !31*

(150) / $!

(151) % $

(152) / $ $ 9 % $ 31*

(153)

(154) %

(155) !

(156) 3 !%$ & %

(157) 8

(158) ! "

(159) #

(160)

(161) $

(162)

(163) "#

(164) %&'(" ")*&+(, " ?1 8 $ () &11;==(= &1#;)56)454(65 6)(6+$. @@ 8 @

(165) A B 6)(6-.

(166) List of Papers. This thesis is based on the following papers:. I. Kartik Narayan, Sakesit Chumnarnsilpa, Han Choe, Edward Irobi, Dunya Urosev, Uno Lindberg, Clarence E. Schutt, Leslie D. Burtnick, and Robert C. Robinson (2003) Activation in isolation: exposure of the actin-binding site in the Cterminal half of gelsolin does not require actin. FEBS Lett. 552(2-3):82-5.. II. Sakesit Chumnarnsilpa*, Anantasak Loonchanta*, Bo Xue, Han Choe, Dunya Urosev, Hui Wang, Uno Lindberg, Leslie D. Burtnick, and Robert C. Robinson (2006) Calcium ion exchange in crystalline gelsolin. J. Mol. Biol. 357(3):773-82. *Shared first co-authorship. III. Hui Wang, Sakesit Chumnarnsilpa, Anantasak Loonchanta, Qiang Li, Yang-Mei Kwan, Sylvie Robine, Marten Larsson, Ivana Mihalek, Leslie D. Burtnick, and Robert C. Robinson (2009) Helix straightening: an activation mechanism in the gelsolin superfamily. (Submitted). IV. Sakesit Chumnarnsilpa, Qing Ma, Wei Lin Lee, Leslie D. Burtnick, and Robert C. Robinson (2009) The crystal structure of the C-terminus of adseverin: Implications for actin binding. (Submitted). Reprint was made with permission from the publisher..

(167)

(168) Contents. Introduction.....................................................................................................9 Background ...................................................................................................11 The gelsolin superfamily proteins (GSPs)................................................11 Gelsolin................................................................................................11 Adseverin.............................................................................................12 CapG....................................................................................................12 Flightless-I ...........................................................................................13 Villin....................................................................................................13 Advillin................................................................................................14 Supervillin ...........................................................................................14 Functions and regulation of gelsolin ........................................................15 Structural basis of gelsolin-actin interaction ............................................15 Comparative analysis of the GSPs used in this study...............................20 Gelsolin vs villin..................................................................................20 Gelsolin vs adseverin...........................................................................20 Present investigations....................................................................................22 Aims .........................................................................................................22 Results ......................................................................................................23 Activation in isolation: exposure of the actin-binding site in the Cterminal half of gelsolin does not require actin (Paper I) ....................23 Calcium ion exchange in crystalline gelsolin (Paper II)......................23 Helix straightening: an activation mechanism in the gelsolin superfamily (Paper III) ........................................................................25 The crystal structure of the C-terminus of adseverin: Implications for actin binding (Paper IV) ......................................................................26 Discussion ................................................................................................28 Calcium is sufficient to fully induce the conformational changes in G4–G6: ................................................................................................28 Calcium activation of G4–G6 in detail:...............................................28 GLD6 helix straightening as an activation mechanism for GSPs:.......30 Gelsolin and adseverin – common or different:...................................30 Future perspectives .......................................................................................32.

(169) Acknowledgements.......................................................................................33 References.....................................................................................................34.

(170) Abbreviations. ABS ABP ATP CBS cDNA DTT EGTA GLD GSP HP LRR NLS PI PIP PIP2 PEG A1 G1 V1 G1–G3 G4–G6 G1–G3/actin G4–G6/actin G-actin F-actin. actin binding site actin binding protein adenosine triphosphate calcium binding site complementary DNA dithiothreitol ethylene glycol tetraacetic acid gelsolin-like domain gelsolin superfamily of protein head-piece domain leucine-rich repeat nuclear localization signal phosphatidylinositol phosphatidylinositol 4monophosphate phosphatidylinositol 3, 4 or 4, 5bisphosphate polyethylene glycol gelsolin like domain 1 of adseverin gelsolin like domain 1 of gelsolin gelsolin like domain 1 of villin N-terminal half of gelsolin C-terminal half of gelsolin complex of N-terminal half of gelsolin and actin complex of C-terminal half of gelsolin and actin monomeric actin filamentous actin.

(171)

(172) Introduction. Actin is one of most abundant proteins in eukaryotic cells. It can occur in monomeric (G) and polymeric (F) states. Transition between these two states is a dynamic process controlled by the innate properties of actin and by actin-binding proteins (ABPs). A number of ABPs control actin dynamics via one or more activities: capping, severing, bundling and nucleation of actin filaments. Gelsolin superfamily proteins (GSPs) are known as ABPs which have all the above actin modifying activities. Gelsolin is the founding member of GSPs, which contains six other members in humans: villin, adseverin, capG, advillin, supervillin, flightless-I. In addition to their respective roles in actin filament remodeling, these proteins have some specific and apparently nonoverlapping roles in several diverse cellular processes, including cell motility, control of apoptosis, phagocytosis, and gene expression. GSPs evolved from gene replication so that they are believed to share several common properties. The activation mechanisms of many of the GSPs are currently unclear. Gelsolin has been the most extensively studied of the GSPs. Gelsolin is regulated by calcium ions, intracellular pH and phosphatidylinositol 3, 4 or 4, 5-bisphosphate (PIP2) isomers. It has been shown to require calcium ions for both the severing and nucleating activities. Protons and calcium ions show a synergistic effect in activation of gelsolin. However, severing activity of gelsolin can be observed at acidic pH in complete absence of calcium (Lamb et al., 1993). The structure of gelsolin adopts a closed conformation in the inactive state (Burtnick et al., 1997) which cannot bind to actin. All the actin binding sites of gelsolin are buried inside the molecule, in this conformation, and are not accessible to bind to actin. The structure of the C-terminal half of gelsolin/actin complex in the presence of calcium ions (Robinson et al., 1999) proves that the conformation of the molecule must be changed in order to interact with actin. The function of gelsolin is clearly controlled by its ability to adopt different conformations. The active structures of the N-terminal (Burtnick et al., 2004) and C-terminal (Robinson et al., 1999) halves of gelsolin have only been reported as complexes with actin and in the presence of calcium ions. These have provided a structural view of the role of gelsolin in controlling actin dynamics. Together, these data have led to a model of actin-filament capping which is an important step in actin dynamics. These structures detail the large domain movements in the C-terminus needed to 9.

(173) convert the inactive conformation into the active actin-bound conformation. While calcium is critical in initiating the conformation change, evidence to prove that calcium ions can accomplish the entire conformation change in the absence of actin is missing. This thesis was aimed at understanding the activation mechanism of the C-terminal half of the GSPs by investigating the atomic structure of the three closest related members: gelsolin, adseverin and villin. X-ray crystallography was used to determine the structure of portions of the C-terminal halves of these proteins. The results showed that the structures of the calcium-bound C-terminal halves of gelsolin and adseverin in the absence of actin are similar and type II calcium ions are crucial for activation of these molecules. The structure of isolated calcium-free V6, adopts an active conformation, revealing that calcium ion is not necessary for the activation of this domain. Our results proved that the CBSs in the C-terminal half of gelsolin are genuine metal ion binding sites and metal ions at these CBSs are exchangeble. The local conformation changes at the type II CBSs in the Cterminal half of gelsolin, revealed by the calcium-extracted G4–G6 structures, suggest the initials steps that reverse the molecule into the inactive conformation. We suggested that the activation of C-terminal half of GSPs is induced by type II calcium ions and that the activation of GLD6 through straightening of the helix is a key component in the global conformation changes of C-terminal halves of these proteins.. 10.

(174) Background. The gelsolin superfamily proteins (GSPs) The gelsolin superfamily is a conserved family of proteins present in mammalian as well as in non-mammalian organisms. All gelsolin superfamily members contain either three or six homologous repeats of a domain named the gelsolin-like domain (GLD). However, some of gelsolin superfamily members; villin, advillin, supervillin, flightless-I, protovillin and EhABPH have additional domains beyond the six fold repeat at either N-terminus or C-terminus or at both termini (Fig. 1).. Figure 1. Schematic representation of domain structure of the 7 human GSPs: Gelsolin, adseverin, CapG, flightless-I, villin, advillin and supervillin and 2 nonmammalian GSPs EhABPH (Entamoeba histolytica) and protovillin (Dictyostelium amoeba).. Gelsolin Gelsolin, a six-domain protein, is expressed in variety of cell types. In humans, this protein exists as three distinct isoforms. The different isoforms are. 11.

(175) encoded by a single gene on chromosome 9 (Kwiatkowski et al., 1988; Yin, 1987; Yin et al., 1984) through alternative splicing of mRNA. Plasma gelsolin, gelsolin-1, is the longest isoform (755 amino acids), and is an extracellular protein. It contains an additional 24-amino acid signaling peptide and the presence of a disulfide bond between cysteine residues at positions 188 and 201 (Wen et al., 1996; Yin et al., 1984). Plasma gelsolin plays an important role in clearing actin from the blood circulation (Lind et al., 1986). A mutation at position 187 of the gelsolin gene causes disease, familial amyloidosis of the Finnish type (FAF) (Maury et al., 1990; Maury and Baumann, 1990; Ghiso et al., 1990; de la Chapelle et al., 1992a; de la Chapelle et al., 1992b). Gelsolin-2, the shortest isoform (731 amino acids), is present in the cytoplasm and plays roles in controlling several cellular processes such as phagocytosis (Arora et al., 2004), motility (Kwiatkowski, 1999), apoptosis (Kwiatkowski, 1999), and platelet formation and activation (Casella et al., 1981). Gelsolin-3 (742 amino acids), is characterized by 11 additional residues at the N-terminus and is also found in the cytoplasm. It is expressed in oligodendrocytes mainly in the brain, lungs and testis, but its specific function is still unknown (Sun et al., 1999).. Adseverin Adseverin or scinderin is the most similar GSP to gelsolin and was first discovered in platelets, megakaryocytic and chromaffin cells (Rodriguez Del Castillo et al., 1992a). Adseverin shares 63 and 53% homology with gelsolin and villin, respectively (Marcu et al., 1994). In comparison to gelsolin, adseverin has a more restricted expression. Adseverin is present in all secretory cells and is involved in actin cytoskeleton remodeling occurring during exocytosis (Marcu et al., 1996; Rodriguez Del Castillo et al., 1992b; Tchakarov et al., 1990) and in platelet activation (Casella et al., 1981). Adseverin is a calcium-dependent actin filament severing protein that controls cortical actin network dynamics during secretion (Trifaro et al., 2000). Adseverin is also found localize in the nucleus (Onoda et al., 1993; Prendergast and Ziff, 1991).. CapG CapG, originally isolated from the cytoplasm of alveolar macrophages (Southwick and DiNubile, 1986), is involved in the control of actin-based cell motility (Silacci et al., 2004) and phagocytosis (Witke et al., 2001) of non-muscle cells. It is relatively small compared to the other members in gelsolin superfamily, containing three GLDs (Southwick and DiNubile, 1986). In the cytoplasm, CapG stops polymerization of actin filaments by capping and is unable to sever filaments. The activity of CapG is regulated by intracellular calcium and PIP2. CapG has been found to accumulate in the 12.

(176) nucleus (Onoda et al., 1993). Its sequence contains several regions which are responsible for nuclear import (Gettemans et al., 2005) and lacks a known nuclear export sequence. The function of CapG in the nucleus is unknown (Van Impe et al., 2003). Knockout studies of CapG shows disruption in membrane ruffling and phagocytosis of macrophages, and causes motility defects in neutrophil granulocytes and dendritic cells (Parikh et al., 2003). In contrast, overexpression of CapG triggers an increase in cell motility of various benign cells (Sun et al., 1995), endothelial cells (Parikh et al., 2003) and kidney cells (De Corte et al., 2004).. Flightless-I Flightless-I, originally identified from a Drosophila melanogaster mutant that was unable to fly (Campbell et al., 1993), contains six GLDs (Silacci et al., 2004). Although there is evidence that flightless-I regulates the actin cytoskeleton, its GLDs are far more divergent from other gelsolin family members, which suggests that the GLDs of flightless-I may have other unique regulatory functions. Importantly, flightless-I has a N-terminal leucine-rich repeat (LRR) protein-protein interaction domain that has many identified binding partners, including FLAP1 (flightless-I-interacting protein) which binds to the LRR (Fong and de Couet, 1999; Liu and Yin, 1998). Flightless-I is the only member of gelsolin superfamily that has been shown to be essential for mouse development; flightless-I deficient mice are embryonic lethal (Campbell et al., 2002), whereas mice deficient in several other GSPs such as gelsolin, villin, and capG survive (Pinson et al., 1998; Witke et al., 1995). In humans, abnormality of flightless-I gene results in developmental and behavioral abnormalities (Chen et al., 1995).. Villin Villin, first isolated from chicken intestinal epithelial cells and later from mammalian species, is an acidic polypeptide with a molecular mass of 92.5 kDa. It is a tissue-specific actin-binding protein expressed in the brush border of enterocytes and proximal kidney cells (Khurana et al., 1997). Villin contains 6 GLDs and C-terminal head-piece (HP). The HP is unrelated to the GLDs and is involved in the bundling activity of villin, a property that is absent from gelsolin. The activities of villin, bundling, capping, severing and nucleating are regulated through calcium-binding, PIP2-binding and phosphorylation (Khurana and George, 2008). Villin has been demonstrated to participate in cytoskeletal remodeling in response to various stimuli in the intestine (Ferrary et al., 1999). Villin is also found in invertebrates and its expression tends to be limited to the brush border. Surprisingly, villin or villin-like proteins also are found in protists and plants but do not seem to be associated with microvilli-like structures (Klahre et al., 2000). 13.

(177) Advillin Advillin is encoded by the AVIL gene and was initially identified by screening an adult murine brain cDNA library with a probe for bovine adseverin. The predicted amino acid sequence of the 92 kDa murine protein p92 (advillin) is 75% homologous to villin, in comparison; gelsolin is 65% identical to adseverin. Advillin shares the six-domain structure with other GSPs and has a C-terminal headpiece, similar to, yet distinct from, villin (Marks et al., 1998). The AVIL gene has been evaluated as a candidate susceptibility gene for inflammatory bowel disease, IBD2 (Tumer et al., 2002). Advillin is involved in neurite growth and morphogenesis via interaction with cytoplasmic domain of Type F scavenger receptor, SREC-I (Shibata et al., 2004). Advillin is also involved in axon regeneration (Hasegawa et al., 2007).. Supervillin Supervillin is a 205-kDa F-actin binding protein originally isolated from bovine neutrophils (Pestonjamasp et al., 1997). Supervillin forms a highaffinity link between the actin cytoskeleton and membranes. Murine supervillin has at least 2 isoforms. Supervillin isoform 1 is expressed in many tissues, being most abundant in muscle, bone marrow, thyroid gland and salivary gland with a comparatively lower expression level in brain. Isoform 2 (archvillin) is muscle specific and is among the first costameric proteins to assemble during myogenesis, contributing to the myogenic membrane structure and to differentiation. It is tightly associated with both actin filaments and membranes (Chen et al., 2003). The human supervillin gene (SVIL), encodes for 7.5 kb mRNA, is localized to a single chromosomal locus at 10p11.2, a region that is deleted in some prostate tumors. The two isoforms are generated via alternative splicing of mRNA. Supervillin isoform 1 is an 1788-amino acid protein and isoform 2 is an 2214-amino acid protein with insertions at positions 275–669 and 749–781 in the amino acid sequence (Pope et al., 1998). The supervillin cDNAs cloned from normal human kidney and from the cervical carcinoma HeLa S3 predict a bipartite structure with three potential nuclear localization signals in the N-terminus and three potential actin binding sequences in the C-terminus. The C-terminal half of supervillin is similar to segments 2–6 plus the C-terminal headpiece of villin. Comparison of the bovine and human sequences indicates that supervillin is highly conserved at the amino acid level, with 79% identity of the Nterminus and conservation of three of the four nuclear localization signals found in bovine supervillin. The C-terminus is even more highly conserved, with 95.1% amino acid identity overall and 100% conservation of the villinlike headpiece. (Pope et al., 1998).. 14.

(178) Functions and regulation of gelsolin Gelsolin was first described as a protein able to bind and sever actin filaments, and to control polymerization of barbed ends (Sun et al., 1999). In vitro, this protein also initiates formation of actin filaments by binding two monomeric actin molecules. In the absence of calcium ions, gelsolin exists in a globular conformation (Sun et al., 1999). Crystal structure analysis has revealed the existence of a C-terminal tail (‘latch helix’) which is in close contact with the actin binding region of the G2 domain (Burtnick et al., 1997). The importance of this latch helix in calcium-regulation of gelsolin-actin interactions was first revealed by the demonstration that the lack of 20 residues in the C-terminal tail abrogates the calcium regulation of actin binding (Kwiatkowski et al., 1989; Way et al., 1990). The latch helix hypothesis suggests that calcium ion binding to the G6 domain induces an initial conformation change in the gelsolin structure, releasing latch helix inhibition of G2 binding to actin. G2 binding to actin directs the G1 domain to its actin binding site. In the absence of calcium ions, G4 and G6 domains are close together (Fig. 2). Calcium ions open this structure, and the G6 domain forms new contacts with the G5 domain, releasing the G4 domain, which now can contact actin through an interface mediated by calcium. The actin binding sites in the N- and C-terminal halves bind to two adjacent actin protomers in a filament. Gelsolin severs the actin filament, remaining bound to the newly formed actin plus end (barbed end). Uncapping of actin filaments requires binding of gelsolin to phosphatidylinositol lipids. This process exposes the barbed end for polymerization (Liepina et al., 2003; Yu et al., 1992). PIP2 isomers are the known intracellular agents inhibiting gelsolin severing and inducing disassociation of gelsolin from actin (Lamb et al., 1993). It has also been demonstrated that the calcium ion requirements for both severing and nucleating activities decrease with lower pH. Half-maximal activities require 10 M calcium at a pH of 7.4 and decrease to 3 M at pH 6.5. At lower pH, direct activation of gelsolin severing activity can also be observed in a complete absence of calcium. The low pH-activated gelsolin is also inhibited by PIP2. This feature has not been observed for the other gelsolin superfamily members (Lamb et al., 1993).. Structural basis of gelsolin-actin interaction Structure of the C-terminal half of gelsolin (G4–G6) in complex with actin derived from crystals of the complex in the presence of calcium ions is shown in Figure 2B. The structure shows the dramatic global conformation change of G4–G6 from the inactive, closed conformation (Fig. 2A, C) to the 15.

(179) active, open conformation (Fig. 2D). The straightening of the helix of G6 during activation shows that the conformation changes are not limited to domain-domain rearrangements (Fig. 3).. Figure 2. Activation of the C-terminal half of gelsolin; (A) Structure of the inactive C-terminal half of gelsolin adopted from PDB 1D0N (Burtnick et al., 1997). (B) Structure of the active C-terminal half of gelsolin in the presence of actin and calcium ions: PDB 1H1V (Choe et al., 2002). (C) Structure of the inactive C-terminal half of gelsolin adopted from PDB 1D0N (Burtnick et al., 1997). (D) Structure of the active C-terminal half of gelsolin in the presence of calcium adopted from PDB 1DB0 (Robinson et al., 1999). 16.

(180) Figure 3. Structural changes within G6 induced by calcium activation; (A) Structure of the inactive G6 (PDB 1D0N) (Burtnick et al., 1997). (B) The active form of G6 in a similar orientation (PDB 1DB0) (Robinson et al., 1999).. An initial model of gelsolin-capped F-actin had been proposed (Fig. 4) (Robinson et al., 1999). This model was based on the structural similarity between N- and C-terminal halves of gelsolin harvesting from the structure of inactive gelsolin 1D0N (Burtnick et al., 1997), both halves of gelsolin are assumed to bind to actin in the same fashion. The model was built by overlaying the structure of active C-terminal half and predicted N-terminal half of gelsolin onto the Holmes model of the actin filament (Holmes et al., 1990). The later structure of N-terminal half of gelsolin in complex with actin (Burtnick et al., 2004) showed a different interaction in comparison to the Cterminal half of gelsolin. This new information led to the proposal of new capping model of F-actin (Fig. 5C). The N-terminal half of gelsolin interacts with subdomains 1, 2, and 3 of actin (Fig 5A) whereas C-terminal half of gelsolin interacts with only subdomains 1 and 3 of actin (Fig 5B).. 17.

(181) Figure 4. Proposed model of the gelsolin-two-actin complex (Robinson et al., 1999); Schematic representations of the two copies of the G4–G6/actin complex are oriented so that the actin protomers are related according to the Holmes model of the actin filament (Holmes et al., 1990). Gelsolin domains G4, G5, and G6 bind to actin protomer 1 are labeled G1, G2, and G3, respectively. The actin filament is oriented to run vertically in the plane of the paper, indicated by the arrow (cyan). The Cterminus of G3 (labeled C) and the N-terminus of G4 (labeled N) are 66 Å apart in this model.. 18.

(182) Figure 5. Second model of a gelsolin-capped filament; (A) A schematic representation of the G1–G3/actin complex. The gelsolin domains are colored: G1, red; G2, green; G3, yellow. Actin, with subdomains 1, 3 and 4 indicated, is colored gray. ATP is shown as a ball-and-stick representation with its associated calcium in purple. Type I and type II calcium ions are depicted as gold and black spheres, respectively. (B) The structure of the G4–G6/actin complex (PDB 1H1V) (Choe et al., 2002) for comparison. Gelsolin domains are colored: G4, pink; G5, dark green; G6, orange. (C) The gelsolin-capped filament model (four protomers are drawn in blue and gray) with G1–G3/actin (PDB 1RGI) (Burtnick et al., 2004) and G4–G6/actin (G4, pink; G5, dark green; G6, orange; PDB 1H1V; (Choe et al., 2002) overlaid onto the barbed end. Purple spheres mark Asp371 of G3 and Met412 of G4, a gap of 63.1 Å to be bridged by the G3–G4 linker, which is modeled in purple.. 19.

(183) Comparative analysis of the GSPs used in this study Pair wise comparison of reported data for the three closely related proteins in this study (gelsolin, villin and adseverin) provides the basis to interpret the data presented in this thesis.. Gelsolin vs villin Gelsolin and villin share 50% amino acid sequence identity and show similar proteolytic cleavage patterns (Janmey and Matsudaira, 1988). Both proteins comprise six GLDs. The largest difference between them is an additional domain at C-terminus of villin, called head piece domain (HP). The HP folds into a compact structure that introduces a second F-actin binding site into villin. Villin uses it to achieve bundling of F-actin (George et al., 2007). In the absence of calcium ions, gelsolin and villin both exist in the inactive conformations. After association with calcium ions, villin and gelsolin undergo conformation rearrangements that expose the F-actin binding sites (Hesterberg and Weber, 1983). Sequential mutagenesis of the CBSs in villin has identified six functional CBSs (Kumar and Khurana, 2004). V1 contains one of each type I and II CBSs whereas each of V2–V6 contain one of type I CBS. F-actin capping and severing activities are regulated by the type I CBS in V1 whereas the lower affinity type II site in V1 only affects severing (Northrop et al., 1986). The type II CBSs in V4–V6 are involved in stabilizing the active villin conformation. NMR studies of a fragment of villin that consists of V6 and the HP domain have implicated V6 residues Asn647, Asp648 and Glu670 in coordinating calcium, and also revealed the first 80 residues of V6 to undergo significant conformation changes as a result of calcium binding (Smirnov et al., 2007).. Gelsolin vs adseverin Gelsolin and adseverin are each composed of six gelsolin-like domains sharing 60% homology. Again, the largest difference between these two proteins lies in their C-termini. Gelsolin contains an additional 20 amino acids at Cterminus which form an α-helix (latch helix). In the absence of calcium, the latch helix covers the F-actin binding site on G2. Two CBSs, Kd = 0.6 and 3 M of adseverin have been determined by equilibrium dialysis experiments (Rodriguez Del Castillo et al., 1990). These two CBSs regulate severing activity of adseverin in a single rate-limiting step (Lueck et al., 2000). By the same taken, equilibrium dialysis experiments also show 2 CBSs for gelsolin, Kd = 0.3 and 1.2 M (Lin et al., 2000) whereas the activation is translated into two rate-limiting steps with respect to severing. The second rate-limiting. 20.

(184) step has been attributed to the unlatching of the C-terminal helix (Lueck et al., 2000). The calcium activation of gelsolin is much more complicated than simply the two rate-limiting steps in removing the latch helix. This was revealed by crystallographic identification of the number of bound calcium ions, eight in total to gelsolin. Biochemical studies of CBSs of gelsolin have identified CBSs in G1 (Kd = 0.6 mM) (Zapun et al., 2000), G2 (Kd = 0.7 M) (Chen et al., 2001), G4 (Kd = 2 M) (Pope et al., 1995), and G6 (Kd = 0.2 M) (Pope et al., 1995). Moreover, radiolytic footprinting and small angle X-ray scattering experiments have revealed that calcium-induced conformation changes occur at 0.1–5 M and at 10 M to 1 mM calcium (Ashish et al., 2007; Kiselar et al., 2003). These data suggest that each CBS will have a functional role. At acidic pH both gelsolin and adseverin can sever F-actin in the absence of calcium (Lueck et al., 2000). Adseverin is deactivated by broader range of phospholipids, phosphatidylinositol 4, 5-bisphosphate (PIP2), phosphatidylinositol 4-monophosphate (PIP), phosphatidylinositol (PI) and phosphatidylserine (PS), whereas only PIP and PIP2 dissociate gelsolin from actin filaments (Janmey and Stossel, 1987; Maekawa and Sakai, 1990). Each half of gelsolin and adseverin is able to sequester 1 actin monomer (Trifaro et al., 1992). Surprisingly, severing and sequestering activities of the N-terminal half of adseverin (A1–A3) are regulated by calcium. In contrast, the N-terminus of gelsolin (G1–G3) actin-severing and -sequestering activities do not require calcium (Sakurai et al., 1991). The C-terminal halves of adseverin (A4–A6) and gelsolin (G4–G6) contain calcium-dependent actin monomer-sequestering and filament-capping activities (Sakurai et al., 1991). Nucleating activity of the isolated A5 have been reported (Marcu et al., 1998). In contrast, no such nucleating activity has been reported for G5.. 21.

(185) Present investigations. Aims The overall objective of these studies was to understand the control of actin dynamics by the gelsolin superfamily of proteins. The specific aims of the studies included in this thesis were 1. To answer the question: Does the open conformation of the C-terminal half of gelsolin need to be formed before forming a complex with actin, or does actin binding contribute to the conformation change? 2. To investigate the calcium-binding sites of the N-terminal and C-terminal halves of gelsolin by varying the free metal ion environment surrounding protein crystals. 3. To determine the structures of active state of domain 6 of villin (V6) to compare it to those of gelsolin domain 6 (G6) 4. To determine the structure of active state of C-terminal half of adseverin and compare to that of gelsolin. 22.

(186) Results. Activation in isolation: exposure of the actin-binding site in the C-terminal half of gelsolin does not require actin (Paper I) In order to understand how the activation and actin binding of the C-terminal half of gelsolin occurs, crystallographic studies on the molecule were performed. We found that the structure of G4–G6, in the presence of high levels of calcium ions, displays an open conformation which is essentially identical to the structure of G4–G6 in complex with actin (Robinson et al., 1999). The structure also shows that all the type II CBSs in G4, G5, and G6 are occupied by calcium ions, in agreement with the G4–G6/actin complex structure (Choe et al., 2002). In contrast, the type I calcium ion, which binds between G4 and actin, is not present in isolated G4–G6. Hence, in high levels of calcium ions, the actin-bound conformation of the C-terminal half of gelsolin is solely induced by 3 type II calcium ions. These data lead to a two step model for the mechanism of interaction between C-terminal half of gelsolin and actin in high calcium environments. Firstly, three type II calcium ions induce rearrangement of G4–G6 to facilitate interaction between the molecule and actin by exposing actin-binding sites on G4 and G6, which are hidden in the inactive state. Secondly, active G4–G6 binds to actin. Simultaneous incorporation of the type I calcium ion at the G4:actin interface stabilizes the complex.. Calcium ion exchange in crystalline gelsolin (Paper II) Paper I details the mechanism of the activation of the C-terminal half of gelsolin. The mechanism of reversing this process, dissociating actin and calcium and returning G4–G6 to its inactive state, remains unresolved. The three type II calcium ions are able to activate the C-terminal half of gelsolin but it is still unclear about their role in the reversion of the molecule into inactive state. In order to investigate the roles of these calcium ions, we performed soaking experiments designed to either to extract or exchange calcium ions from the active G4–G6 crystals. We also investigated the structure at different pHs as protons are a second activator of gelsolin. Analysis of the structure of active G4–G6 derived from G4–G6 crystals grown at pH 4.5 in the presence of calcium ions demonstrated that the structure is indistinguishable to that determined at pH 7.5 (paper I). This result suggests that pH does not have any activating effects beyond those induced by calcium in altering conformation of G4–G6. We found that washing the crystals with EGTA solutions enabled extraction of calcium ions from the structure. However, washing with calcium-free 23.

(187) buffers in the absence of EGTA resulted in little extraction. The EGTAextracted G4–G6 crystals at pH 8.0 retained the global active conformation of G4–G6. However, local conformation changes were observed at the type II CBSs. Loss of the calcium ion from the G4 type II CBS results in the release of the Thr524 carbonyl group from the calcium ion-binding site and the disordering of the proximal loop at G5 (residues 526–528). The calcium coordinating residues at the G5 CBS show little change after calcium ion is removed. Finally, removal of calcium from G6 shows reorientation of residues that previously coordinated calcium. Asp670 withdraws from the CBS by approximately 1.5 Å and a water molecule moves in to replace the calcium ion, forming hydrogen bonds with Glu692 and Asp669. At pH 4.5, EGTA was less efficient in removing calcium ions. It was only able to extract one calcium ion from G6 of molecule C, one of three molecules A, B and C in asymmetric unit. The conformation of this calcium free G6 showed the same features as calcium extracted G6 at pH 8.0. We studied the exchangeability of metal ions at CBSs of G4–G6. The G4– G6 crystals were washed in solutions lacking calcium ions and EGTA, but containing 1.0 mM terbium ions. Anomalous scattering data were used to determine the absence or presence of terbium ions. Peaks greater than 15 σ were observed at all of the type II CBSs at both pH 4.5 and pH 8.0 confirming the exchange. Moreover, the G4 type I CBS also showed partial occupancy in two of the three molecules (18.17 σ and 6.31 σ, respectively, in molecules A and B at pH 4.5). This finding indicates that all of the calcium ions of G4–G6 can be replaced by terbium ions. Rewashing of terbiumtreated crystals in 1 mM calcium demonstrated that the terbium ions bound to G4–G6 could be competed away by excess calcium ions. These results demonstrate that metal ions at both types of CBS of G4–G6 are exchangeable and identify that the type I CBS on G4 is genuine metal ion-binding site, even when lacking the coordination of Glu167 from actin. Similar calcium-terbium ion exchangeability experiments were performed with G1–G3/actin crystals. These crystals also accepted terbium ions into the 3 type II sites, in G1, G2 and G3, and one type I site in G1. However, residual terbium ion binding was observed at the G2 type II site and G1 type I site after back soaking with calcium. In summary: this study shows that calcium ions can be removed from G4–G6 while it is maintained in a globally activated conformation. Loss of the G4 type I site calcium, which lies at the G4:actin interface, and the G6 type II site calcium, which orders the G6:actin interaction, would both contribute to loss of affinity of G4–G6 for actin. The loss of calcium and local conformation changes observed at the G4 and G5 CBSs occurs at the interfaces between adjacent domains, G4–G5 and G5–G6, respectively. Hence, loss of these calcium ions and the ensuing local conformational rearrangements around the CBSs will lead destabilization of these interfaces. Together. 24.

(188) these changes drive actin dissociation and return of G4–G6 to the inactive conformation.. Helix straightening: an activation mechanism in the gelsolin superfamily (Paper III) The structure of inactive gelsolin showed that G6 is the central domain of gelsolin, forming interactions with all other domains (Burtnick et al., 1997). The results presented in papers I and II demonstrate that G6 changes its interactions with G4 and G5 on binding calcium. Hence, we hypothesized that G6 plays a crucial role in the activation mechanism of whole gelsolin. During attempts to crystallize the C-terminal half of villin (V4–V6), the sixth domain crystallized in the absence of V4–V5. Therefore, we performed the crystallographic analysis on isolated villin domain 6 (V6). The overall structure of isolated V6 in an inactive buffer environment, low calcium ion concentration, no PIP2, and no phosphorylation, displays the common fold of the gelsolin-like domain (GLD). We compared structure of V6 to the available structures of the six domains of gelsolin, the structure of V6 is most similar to active G6 (calcium-bound form). In the calcium-free environment, the long helix of V6, instead of being kinked like inactive G6 (calcium-free form), is straight like the long helix of active G6. Similarity in the arrangement of residues in V6 (Ser625, Asn626, Asp648 and Glu670) to homologous residues in G6 that are involved in calcium binding (Ser646, Asp647, Asp670 and Glu692) confirmed that V6 is in an active state despite the absence of bound calcium. V6 interaction interfaces: The structure of inactive gelsolin (Burtnick et al., 1997) showed that G6 interacts most extensively with G2 and G4. The βstrand (490-504) from G4 and side chains of Arg168 and Arg169 from G2 are conserved in villin suggesting similar intramolecular interactions. Models of active G6 and isolated V6 superimposed onto G6, in inactive G4–G6, showed that a loop connecting two strands of G4 forms a steric clash in both models. This result suggests that G4, and hence V4, cannot form interaction with these two conformations, adding weight to the view that isolated V6 represents an activated state. Analysis of the interaction interface at G2 was performed by superimposing structures of active G6 and isolated V6 onto inactive G6, with respect to G2. We found different conformations of the AB loop (loop connecting strands A and B) between the active and inactive forms of G6. This results from calcium binding to the G6 CBS and straightening long α-helix of G6. These conformation changes in G6, we predict, will disrupt electrostatic interactions between Asp670 of G6 and Arg168 and Arg169 of G2 and lead to separation of G6 and G2. These residues are conserved in villin and the straightened long helix conserved in the isolated structure of V6. Thus we 25.

(189) propose that helix straightening in domain 6, through calcium binding, leading to separation of domains 2 from 6 is a common mechanism in gelsolin and villin activation. An interaction between the AB loop of G6 and an α-helix (residues 307– 321) in actin subdomain 3 was observed in the G4–G6/actin complex structure (Choe et al., 2002). Superimposition of the coordinates of the inactive G6, and isolated V6, onto active G6 in the G4–G6/actin structure reveals that inactive G6 does not fit to the α-helix of actin while V6 adopts a reasonable conformation. This model implies that the activation of domain 6 of GSPs, by the straightening of the long α-helix, may allow domain 6 to bind actin and induce severing at suboptimal calcium concentrations. Molecular dynamics calculations were performed as a collaboration to investigate the apparent anomaly that isolated calcium-free V6 should more closely resemble the calcium-bound G6 rather than the calcium-free one. The simulations confirmed that the straight form of the α-helix in G6 is more stable, even in the absence of a bound calcium ion, indicating the reliability of the structure of isolated V6. In the absence of V4, it is not necessary for the helix to be held in the less energetically favored kinked form. In gelsolin, the kink of the long α-helix of G6 is needed for packing G6 into the inactive conformation to avoid sterically clashing with the long α-helix of G4. This we predict to be the same in the villin core (V1–V6) due to the high sequence identity. This result suggests how gelsolin and villin balance energy during activation. Straightening the long α-helix of G6 may provide a portion of the energy needed to tear apart the interaction between extended β-sheet between G4 and G6 and the interaction between G2 and G6. The remainder of the energy to complete the activation process may be recouped from the binding of calcium ions and the formation of new domain contacts, such as between G5 and G6 (Choe et al., 2002; Robinson et al., 1999).. The crystal structure of the C-terminus of adseverin: Implications for actin binding (Paper IV) The preceding papers detail common activation mechanisms shared by gelsolin and villin. We further investigated the generality of these mechanisms through comparing the structural and biochemical details of the C-terminal half of gelsolin with that of its closest relative, adseverin. The structure of C-terminal half of adseverin (A4–A6) in an active state, in the presence of calcium ions comprises three typical GLDs each bound to a single calcium ion. Domain arrangement and calcium binding of A4–A6 closely resemble active G4–G6 rather than the inactive form. The C-terminal residues of adseverin are clearly defined whereas the longer C-terminal tail of gelsolin is defined up to the equivalent last residue of adseverin. The gelsolin C-terminal extension residues were disordered. Inspection of the details 26.

(190) of the CBSs in A4–A6 demonstrated they are structurally conserved with those of G4–G6. The only difference is seen at the A4 CBS where the A5 loop is disordered as a result of crystal packing. In gelsolin the G5 loop partly coordinates the calcium ion in the G4 CBS. Biochemical assays were carried out as a collaboration to determine the similarities in calcium and actin binding by gelsolin and adseverin. Thermal transitions of adseverin during calcium activation was monitored by a thermoshift assay in comparison to those measured for gelsolin in the same assay (Ma et al, 2009). Both proteins demonstrated three states: a stable calcium-free state, an intermediate lower stability state, and a stable calciumbound state. For adseverin, these states were detected at calcium concentrations < 10 nM, 10–25 nM, and > 25 nM, respectively. These studies revealed that adseverin reaches the minimum of stability at a stoichiometric ratio of two calcium ions to one adseverin molecule. In gelsolin, the initial transition is slightly later, at between 16–25 nM. Similar to adseverin, gelsolin needs two calcium ions to reach the minimum stability. Similar assays were carried out on the two halves of adseverin (A1–A3 and A4–A6). A1–A3 showed similar transition patterns to adseverin and the transition reached the minimum of stability by one or two calcium ions per molecule whereas A4–A6 showed that the initial transition occurs at higher calcium concentrations, 25–45 nM with one calcium ion required per molecule. The transition pattern of G4–G6 qualitatively resembled that for A4–A6 but the transition occurred earlier at between 4–25 nM calcium concentration (Ma et al., 2009). The role of calcium ions in actin disassembly by adseverin in comparison with gelsolin (Ma et al., 2009) was monitored by pyrene-actin depolymerization assays. These experiments revealed that one molecule of adseverin required one calcium ion for complete actin disassembly while one molecule of gelsolin needed two calcium ions. A molecule of A1–A3 needed one calcium ion to disassemble actin. A molecule of A4–A6 needed two calcium ions to completely sequester the actin. Taken together these data suggest that calcium activation of gelsolin and adseverin are similar, however, only one calcium ion is sufficient to activate adseverin whereas 2 are required for gelsolin. The structures of G4–G6 and A4–A6 also suggest a high level of similarity in actin binding. Superimposition of the structure of A4–A6 onto G4–G6 in the G4–G6/actin structure (Choe et al., 2002) revealed homology in the actin binding interface. The distribution of residues in the actin-binding long α-helix of G4 and A4 are similar in character. The centers of these α-helices display a hydrophobic property while the ends are hydrophilic. Importantly, Asp487 of G4, the type I calcium ion coordinating residue and Asp461 of A4 are conserved. Moreover, the surface charge distribution of the actinbinding portion of G6 is remarkably similar to that of A6. These data strongly suggest that A4–A6 and G4–G6 will bind to actin in the same way. 27.

(191) Discussion Calcium is sufficient to fully induce the conformational changes in G4–G6: Exposing and hiding the actin-binding sites are the main mechanisms for the GSPs in controlling actin dynamics. In paper I, the active conformation of G4–G6 is clearly shown to be achieved in the absence of actin. The presence of type II calcium ions in the active G4–G6 structure and in G4–G6/actin structure (Choe et al., 2002) indicates that these type II calcium ions are crucial for the activation of G4–G6 and its interaction with actin. Biochemical studies also show that G4–G6 requires calcium ions in order to bind to actin (Pope et al., 1995) and for the activation of G4–G6 (Lin et al., 2000). Taken data together, we strongly suggest that type II calcium ions induce conformation changes within G4–G6 in order to expose the actin binding sites on G4 and G6 prior binding to actin. At high levels of calcium, actin is not involved in the activation of G4–G6.. Calcium activation of G4–G6 in detail: The presence of a type II calcium ion in each CBS of G4, G5 and G6 suggests that all calcium ions are important in activation and stabilization of active G4–G6. In contrast to the structure of actin-free, active G4–G6 (Kolappan et al., 2003), the structure presented in this thesis shows that type II calcium ion at G4 is crucial. This calcium ion bridges the calcium binding residues in G4 to the proximal loop in G5 (residues 526–528). In paper II, the role of this type II calcium bound to G4 is revealed. The structure of calcium-extracted G4–G6 demonstrates that the G5 loop moves away from the G4 type II CBS coordination, resulting in the loss of the interaction between the G4 type II CBS and the G5 loop. The G4 type II CBS calcium-dependent binding of the G5 loop suggests an important role for this calcium ion in G4–G6 activation. Structural (Burtnick et al., 1997) and biochemical data (Pope et al., 1995), reveal that the type II CBSs of G4 and G6 are accessible for calcium ion binding however the affinity of the type II CBS at G4 (2 μM) is higher than that of G6 (0.2 μM). We suggest that the activation of G4–G6 is initiated by both G4 and G6 type II calcium ions. Close inspection shows that there is a transition of G4 relative to G5 in comparison to inactive state. We speculate that in moving the G5 loop to form interactions with G4 during activation drags along the whole domain of G5 28.

(192) from its inactive position. This movement of G5 involves rotating and twisting which would effect the orientation of the type II CBS in G5 and place stress on the G4 and G6 interaction. We speculate that calcium binding to G5 and dissociation of G4 and G6 would occur simultaneously after G4 activation. The calcium extraction experiments suggest that the calcium ions bound at the type II CBS of G6 are easiest to remove whereas that of G5 is the hardest to remove. These data are in the reverse order in comparison to the reported affinities at these sites. Clearly, conditions within the crystal are not similar to those within solution. We speculate that by locking G4–G6 in an active state, within the crystal, the Kd for the binding the G5 calcium ion is lowered. Extrapolation of this hypothesis to solution leads to the suggestion that the calcium-binding affinities of G4–G6 may be different in the presence and absence of actin. The reverse process of G4–G6 deactivation is unknown. However, the calcium-extracted structures suggest a couple of key points in the reversal process. Firstly, the disorder in the G5 loop at the type II CBS predicts a destabilization of the G4:G5 interface. Secondly, loss of the calcium ion at the G5 CBS aborts the electrostatic interaction between Arg542 of G5 and Glu731 and Trp732 of G6, leading to the release of G6 from G5. Once released G6 presumably returns to its lowest energy state in which the three domains are packed together in an inactive conformation. Terbium-calcium exchange experiments demonstrate that calcium binding by gelsolin is a reversible process even when it is locked in its active state. The CBS’s can clearly bind to other types of metal ions. This opens a possibility that GSPs may bind to other types of metal ions inside the cell, although magnesium is not a candidate. GSPs localize within different parts of cells and tissues which possess a variety of environments such as pH, oxidation-reduction conditions, calcium, and other metal concentrations. Several members of GSPs have been shown to be activated by calcium ions, which, seem to be the most powerful activator of these proteins. Supervillin and flightless possess mutated calcium-binding sites and hence their response to calcium may be expected to differ (Choe et al., 2002). However, it is unclear how calcium ions may regulate different activities of the protein from the same family? In particular, how do different members that share cellular localization co-operate their functions? And how are some members activated when they localize to areas where the calcium concentrations are very low? Our studies reveal that adseverin and gelsolin have different sensitivities to calcium ions and that both proteins are able to interact with actin at calcium levels below those needed for full occupancy of all sites. We predict that the other GSPs will possess the different calcium sensitivities in order to maintain their functional differences.. 29.

(193) GLD6 helix straightening as an activation mechanism for GSPs: The structure of gelsolin (Burtnick et al., 1997) demonstrated that the G6 is the central domain within gelsolin which interacts with all other domains. Molecular dynamics calculations suggest that G6 prefers an active conformation in an isolated environment in the absence of the other domains. The structure of isolated calcium-free V6 (paper III) adopts an active conformation consistent with the molecular dynamics simulations carried out on G6. This study suggests that the release of GLD6 from other domains arising from any circumstance will allow it to adopt an active conformation. Calcium ions are not required for the activation of GLD6. The conformation change of GLD6 appears a spontaneous process, using energy trapped in the kinked -helix which is a result from domains packing within the inactive conformation. We speculate that the straightening of the long -helix of GLD6, and by analogy the helix in GLD3, may be driving forces during the activation of gelsolin, leading to the domain-domain rearrangements needed to bind to actin. This mechanism may go partway to explain why protons and calcium ions may have similar effects on gelsolin activation.. Gelsolin and adseverin – common or different: Adseverin is the closest related member of the GSPs to gelsolin, sharing 60% sequence homology. However, these proteins play different roles in cellular function. So despite their similarity, these proteins must be fundamentally different. We sought to discover whether there is a structural or biochemical basis to the functional variation. The structure of active A4–A6 presented in Paper IV is similar to active G4–G6 (paper I). The structure suggests that A4–A6 will be able to bind actin in the same manner as G4–G6 and also predicts that C-terminal half of other GSPs will interact to actin in the same fashion. The A4–A6 structure shows that A5 is not accessible for actin binding. Residues that have previously been reported as the actin-binding region, identified through biochemical assays (Dumitrescu Pene et al., 2005) are buried between the G4 and G5 interaction interface. This contrast between the structural and biochemical data suggests that there may be a range of adseverin activities, and perhaps gelsolin activities, under different circumstances. The biochemical analysis of adseverin in this study reveals that adseverin is slightly more sensitive to calcium than gelsolin. The major sequence difference between adseverin and gelsolin lies within the C-terminal extension of gelsolin, the latch helix. Gelsolin needs at least 2 calcium ions to be activated and bind to actin, believed to involve the release of the latch helix. In contrast, adseverin needs only one calcium ion. Even though adseverin does not have a latch helix, modeling studies predict that adseverin should break open its conformation in order to bind to actin. Hence, we 30.

(194) speculate that, although both proteins become activated in the similar manner, the gelsolin latch helix adds a requirement for an extra calcium ion. Taken together, these studies predict that in the cell where both adseverin and gelsolin are present, adseverin will be more active than gelsolin and dominate in controlling of actin dynamics in low calcium circumstances.. 31.

(195) Future perspectives. The increased understanding of activation mechanism of C-terminal half of the GSPs uncovered in this work provides a small step in discerning the complete mechanisms of whole GSPs activation. Since this protein superfamily is very diverse in both structure and function, it will be informative to more thoroughly understand the structures and other properties of the entire family. Therefore the crucial experiments will be to investigate the activation mechanisms of N-terminal half of GSPs and also to determine the full length structures of the other members in order to gain a more detailed knowledge and understanding about their mechanisms.. 32.

(196) Acknowledgements. I would like to thank the Department of Medical Biochemistry and Microbiology (IMBIM) Uppsala University, Sweden and Institute of Molecular and Cell Biology (IMCB), Singapore for funding my study. I gratefully thank my supervisor, Dr. Robert Robinson, for the opportunity and for his mentorship. Many thanks to Prof. Kristofer Rubin, Prof. Göran Magnusson and Barbro Lowisin in supporting my registration at IMBIM. Thanks to the crystallization lab at BMC, special thanks to Terese Bergfors for very useful comments and techniques. Thanks to my friends, Wimal and his family, Igor, Xiaomin, Marten, Kajsa and Lena for being the best part of my PhD study. Thanks to Bob/BR-lab people who help me run my experiments smoothly. Big thanks to Prof. Leslie D. Burtnick, Dr. Kartik Narayan and Dr. Bo Xue for very useful information with regards to publication. Importantly, thanks to Anantasak Loonchanta (Tacky) for being the greatest partner. Most importantly, thanks my family for always being with me, especially during the difficult times.. 33.

(197) References. Arora, P. D., Glogauer, M., Kapus, A., Kwiatkowski, D. J., and McCulloch, C. A. (2004) Mol Biol Cell 15, 588-599. Ashish, Paine, M. S., Perryman, P. B., Yang, L., Yin, H. L., and Krueger, J. K. (2007) J. Biol. Chem. 282, 25884-25892. Burtnick, L. D., Koepf, E. K., Grimes, J., Jones, E. Y., Stuart, D. I., McLaughlin, P. J., and Robinson, R. C. (1997) Cell 90, 661-670. Burtnick, L. D., Urosev, D., Irobi, E., Narayan, K., and Robinson, R. C. (2004) Embo J. 23, 2713-2722. Casella, J. F., Flanagan, M. D., and Lin, S. (1981) Nature 293, 302-305. Chen, C. D., Huff, M. E., Matteson, J., Page, L., Phillips, R., Kelly, J. W., and Balch, W. E. (2001) Embo J. 20, 6277-6287. Chen, K. S., Gunaratne, P. H., Hoheisel, J. D., Young, I. G., Miklos, G. L., Greenberg, F., Shaffer, L. G., Campbell, H. D., and Lupski, J. R. (1995) Am. J. Hum. Gene.t 56, 175-182. Chen, Y., Takizawa, N., Crowley, J. L., Oh, S. W., Gatto, C. L., Kambara, T., Sato, O., Li, X. D., Ikebe, M., and Luna, E. J. (2003) J. Biol. Chem. 278, 46094-46106. Choe, H., Burtnick, L. D., Mejillano, M., Yin, H. L., Robinson, R. C., and Choe, S. (2002) J. Mol. Biol. 324, 691-702. de la Chapelle, A., Kere, J., Sack, G. H., Jr., Tolvanen, R., and Maury, C. P. (1992a) Genomics 13, 898-901. de la Chapelle, A., Tolvanen, R., Boysen, G., Santavy, J., BleekerWagemakers, L., Maury, C. P., and Kere, J. (1992b) Nat. Genet. 2, 157160. Dumitrescu Pene, T., Rose, S. D., Lejen, T., Marcu, M. G., and Trifaro, J. M. (2005) J. Neurochem. 92, 780-789. Ferrary, E., Cohen-Tannoudji, M., Pehau-Arnaudet, G., Lapillonne, A., Athman, R., Ruiz, T., Boulouha, L., El Marjou, F., Doye, A., Fontaine, J. J., et al. (1999) J. Cell Biol. 146, 819-830. George, S. P., Wang, Y., Mathew, S., Srinivasan, K., and Khurana, S. (2007) J. Biol. Chem. 282, 26528-26541. Gettemans, J., Van Impe, K., Delanote, V., Hubert, T., Vandekerckhove, J., and De Corte, V. (2005) Traffic 6, 847-857. Ghiso, J., Haltia, M., Prelli, F., Novello, J., and Frangione, B. (1990) Bio chem. J. 272, 827-830. Hasegawa, H., Abbott, S., Han, B. X., Qi, Y., and Wang, F. (2007) J. Neurosci. 27, 14404-14414. Hesterberg, L. K., and Weber, K. (1983) J. Biol. Chem. 258, 365-369. Holmes, K. C., Popp, D., Gebhard, W., and Kabsch, W. (1990) Nature 347, 34.

(198) 44-49. Janmey, P. A., and Matsudaira, P. T. (1988) J .Biol. Chem. 263, 1673816743. Janmey, P. A., and Stossel, T. P. (1987) Nature 325, 362-364. Khurana, S., Arpin, M., Patterson, R., and Donowitz, M. (1997) J. Biol. Chem. 272, 30115-30121. Khurana, S., and George, S. P. (2008) FEBS Lett. 582, 2128-2139. Kiselar, J. G., Janmey, P. A., Almo, S. C., and Chance, M. R. (2003) Proc. Natl. Acad. Sci. U S A 100, 3942-3947. Klahre, U., Friederich, E., Kost, B., Louvard, D., and Chua, N. H. (2000) Plant Physiol. 122, 35-48. Kolappan, S., Gooch, J. T., Weeds, A. G., and McLaughlin, P. J. (2003) J. Mol. Biol. 329, 85-92. Kumar, N., and Khurana, S. (2004) J. Biol. Chem. 279, 24915-24918. Kwiatkowski, D. J. (1999) Curr. Opin. Cell Biol. 11, 103-108. Kwiatkowski, D. J., Janmey, P. A., and Yin, H. L. (1989) J. Cell Biol. 108, 1717-1726. Lamb, J. A., Allen, P. G., Tuan, B. Y., and Janmey, P. A. (1993) J. Biol. Chem. 268, 8999-9004. Liepina, I., Czaplewski, C., Janmey, P., and Liwo, A. (2003) Biopolymers 71, 49-70. Lin, K. M., Mejillano, M., and Yin, H. L. (2000) J. Biol. Chem. 275, 2774627752. Lind, S. E., Smith, D. B., Janmey, P. A., and Stossel, T. P. (1986) J. Clin. Invest. 78, 736-742. Lueck, A., Yin, H. L., Kwiatkowski, D. J., and Allen, P. G. (2000) Biochemistry 39, 5274-5279. Ma, Q., Wang, H., Nag S., Chumnanrsilpa S., Lee W. L., HernandezValladares Maria, Burtnick, L., and Robinson, R.C. (2009) Manuscript submitted to PNAS. Maekawa, S., and Sakai, H. (1990) J Biol Chem 265, 10940-10942. Marcu, M. G., Rodriguez del Castillo, A., Vitale, M. L., and Trifaro, J. M. (1994) Mol. Cell Biochem. 141, 153-165. Marcu, M. G., Zhang, L., Elzagallaai, A., and Trifaro, J. M. (1998) J. Biol. Chem. 273, 3661-3668. Marcu, M. G., Zhang, L., Nau-Staudt, K., and Trifaro, J. M. (1996) Blood 87, 20-24. Marks, P. W., Arai, M., Bandura, J. L., and Kwiatkowski, D. J. (1998) J. Cell Sci. 111 (Pt 15), 2129-2136. Maury, C. P., Alli, K., and Baumann, M. (1990) FEBS Lett. 260, 85-87. Maury, C. P., and Baumann, M. (1990) Biochim. Biophys. Acta. 1096, 8486. Northrop, J., Weber, A., Mooseker, M. S., Franzini-Armstrong, C., Bishop, M. F., Dubyak, G. R., Tucker, M., and Walsh, T. P. (1986) J. Biol. Chem. 261, 9274-9281. Onoda, K., Yu, F. X., and Yin, H. L. (1993) Cell Motil. Cytoskeleton 26, 227-238. 35.

(199) Parikh, S. S., Litherland, S. A., Clare-Salzler, M. J., Li, W., Gulig, P. A., and Southwick, F. S. (2003) Infect. Immun. 71, 6582-6590. Pestonjamasp, K. N., Pope, R. K., Wulfkuhle, J. D., and Luna, E. J. (1997) J. Cell Biol. 139, 1255-1269. Pinson, K. I., Dunbar, L., Samuelson, L., and Gumucio, D. L. (1998) Dev. Dyn. 211, 109-121. Pope, B., Maciver, S., and Weeds, A. (1995) Biochemistry 34, 1583-1588. Pope, R. K., Pestonjamasp, K. N., Smith, K. P., Wulfkuhle, J. D., Strassel, C. P., Lawrence, J. B., and Luna, E. J. (1998) Genomics 52, 342-351. Robinson, R. C., Mejillano, M., Le, V. P., Burtnick, L. D., Yin, H. L., and Choe, S. (1999) Science 286, 1939-1942. Rodriguez Del Castillo, A., Lemaire, S., Tchakarov, L., Jeyapragasan, M., Doucet, J. P., Vitale, M. L., and Trifaro, J. M. (1990) Embo J. 9, 43-52. Rodriguez Del Castillo, A., Vitale, M. L., Tchakarov, L., and Trifaro, J. M. (1992a) Thromb. Haemost. 67, 248-251. Rodriguez Del Castillo, A., Vitale, M. L., and Trifaro, J. M. (1992b) J. Cell Biol. 119, 797-810. Sakurai, T., Kurokawa, H., and Nonomura, Y. (1991) J. Biol. Chem. 266, 4581-4585. Shibata, M., Ishii, J., Koizumi, H., Shibata, N., Dohmae, N., Takio, K., Ada chi, H., Tsujimoto, M., and Arai, H. (2004) J. Biol. Chem. 279, 4008440090. Silacci, P., Mazzolai, L., Gauci, C., Stergiopulos, N., Yin, H. L., and Hayoz, D. (2004) Cell Mol. Life Sci. 61, 2614-2623. Smirnov, S. L., Isern, N. G., Jiang, Z. G., Hoyt, D. W., and McKnight, C. J. (2007) Biochemistry 46, 7488-7496. Southwick, F. S., and DiNubile, M. J. (1986) J. Biol. Chem. 261, 1419114195. Sun, H. Q., Kwiatkowska, K., Wooten, D. C., and Yin, H. L. (1995) J. Cell Biol. 129, 147-156. Sun, H. Q., Yamamoto, M., Mejillano, M., and Yin, H. L. (1999) J. Biol. Chem. 274, 33179-33182. Tchakarov, L., Vitale, M. L., Jeyapragasan, M., Rodriguez Del Castillo, A., and Trifaro, J. M. (1990) FEBS Lett. 268, 209-212. Trifaro, J. M., Rodriguez del Castillo, A., and Vitale, M. L. (1992) Mol. Neurobiol. 6, 339-358. Trifaro, J. M., Rose, S. D., and Marcu, M. G. (2000) Neurochem. Res. 25, 133-144. Tumer, Z., Croucher, P. J., Jensen, L. R., Hampe, J., Hansen, C., Kalscheuer, V., Ropers, H. H., Tommerup, N., and Schreiber, S. (2002) Gene 288, 179-185. Way, M., Pope, B., Gooch, J., Hawkins, M., and Weeds, A. G. (1990) Embo J. 9, 4103-4109. Wen, D., Corina, K., Chow, E. P., Miller, S., Janmey, P. A., and Pepinsky, R. B. (1996) Biochemistry 35, 9700-9709. Witke, W., Li, W., Kwiatkowski, D. J., and Southwick, F. S. (2001) J. Cell Biol. 154, 775-784. 36.

(200) Witke, W., Sharpe, A. H., Hartwig, J. H., Azuma, T., Stossel, T. P., and Kwiatkowski, D. J. (1995) Cell 81, 41-51. Yin, H. L., Kwiatkowski, D. J., Mole, J. E., and Cole, F. S. (1984) J. Biol. Chem. 259, 5271-5276. Yu, F. X., Sun, H. Q., Janmey, P. A., and Yin, H. L. (1992) J. Biol. Chem. 267, 14616-14621. Zapun, A., Grammatyka, S., Deral, G., and Vernet, T. (2000) Biochem. J. 350 Pt 3, 873-881.. 37.

(201)

(202) %2

(203)

(204) % .

(205)

(206)

(207)

(208)

(209) "1

(210) 0

(211) 3

(212) & 1. 1

(213) 1

(214)

(215)

(216) 3

(217) & 1 4% %2

(218) 4 5 6

(219)

(220) 1

(221)

(222) 7

(223)

(224) 8 6 1. 46

(225) 1

(226)

(227) 1

(228)

(229)

(230) 9

No results found