|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 SYSTEMATIC STUDIES OF SIGNAL PATHWAYS IN AXONAL INJURY AND REGENERATION By ZHIQUN ZHANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2007
2 2007 Zhiqun Zhang
3 To my father Daxin Zhang, to my mother Xi jun Li for their support, encouragement and blessings during my academic journey
4 ACKNOWLEDGMENTS First I would like to th ank my parents and sisters for thei r love, support and patience that I needed during my academic career to achieve th is doctoral degree. I w ould like to thank my advisor, Dr. Kevin K. W. Wang, whose support, patience and guidance have allowed me to finish this academic challenge. It is due to Dr Wangs enlightened vision in research that we pursued areas in calmodulin and system biology th at allowed me to reach frontiers of knowledge in ways that neither one of us could ha ve expected when I began this journey. I would also like to thank th e members of my committee, Pr ofessors Andrew K. Ottens, Dena Howland and Lungji Chang for their precious time, effort and guidance. I acknowledge Dr. Stephen Larner for his tutorials and help in preparing this document and many other publications. I am grateful to th e people in the Departments of Psychiatry and Neuroscience, in particular Professor Susan Semple-Rowland, Dr. Ma rk Good and B.J. Streetman for their efforts in taking care of things that made my life so simple. I am gracious to my dearest friend and colleague Dr. Firas Kobeissy for his unequalled a ssistance in the laborator y. Also, I would like to thank all the members of Dr. Wangs laborator y for their assistance and friendship throughout my 5 years pursuing my Ph.D. I would like to especially thank Dr. Mi ng Cheng Liu and his wonderful wife Dr. Wenrong Zhe ng, Dr. Stanislav Svetlov, Dr. Ronald Hayes, Mr. Shankar Sadasivan and Mrs. Barbara Osteen, for their effo rt and support. Very special thanks go to my hushand, Jianghui Chao, who has patiently stood by me and supported me for more than a decade and my son, Ryan Baichuan Chao who has been given me lots of happiness.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........8 LIST OF FIGURES................................................................................................................ .........9 ABSTRACT....................................................................................................................... ............11 CHAPTER 1 INTRODUCTION..................................................................................................................13 Traumatic Axonal Injury: A General Overview.....................................................................13 Pathology and Calmodulin Mediated Path ways in Traumatic Axonal Injury........................14 Axons Rarely Regenerate after Injury in the Adult CNS.......................................................17 Molecular Approach to Promote Axonal Regeneration Converge at Rho-ROCK Pathway........................................................................................................................ .......22 2 USING CALMODULIN-AFFINITY CAPT URE TO STUDY THE RAT BRAIN CALMODULIN BINDING PROTEOME AND ITS VULNERABILITY TO CALPAIN AND CASPASE PROTEOLYSIS.......................................................................30 Introduction................................................................................................................... ..........30 Materials and Methods.......................................................................................................... .32 Brain Tissue Collection and Protein Extraction..............................................................32 In Vitro Calpain-2 and Caspase3 Digestion of Brain Lysate.........................................33 CaM Affinity Capture and Elution..................................................................................33 In Gel Digestion..............................................................................................................34 Capillary RPLC-MSMS Based Protein Identification....................................................34 Immunoblot Analysis......................................................................................................35 Results........................................................................................................................ .............35 Calmodulin Binding Proteomic Profili ng by Calmodulin-affinity Capture....................35 Identification of CaM Binding Proteome and Calpain/caspase Mediated Breakdown Products by RPLC-MSMS Based Proteomics.............................................................36 Functional Analysis of Putativ ely Novel CaM-Binding Proteins...................................37 Immunoblot Analysis of Select CaMB Ps Identified by RPLC-MSMS Following CaM-Affinity Purification...........................................................................................38 Discussion..................................................................................................................... ..........39
6 3 CALPAIN-MEDIATED COLLAPSIN RESP ONSE MEDIATOR PROTEIN-1, 2 AND 4 PROTEOLYSIS FOLLOWING NEUROTOXIC AND TRAUMATIC BRAIN INJURY......................................................................................................................... .........55 Introduction................................................................................................................... ..........55 Materials and Methods.......................................................................................................... .56 Primary Cortical Neuron Culture....................................................................................56 Neurotoxic Challenges and Pharmacologic Intervention................................................57 Lactate Dehydrogenase Releas e Assay of Cell Death.....................................................57 Cell Lysate Collection and Preparation...........................................................................58 Immunocytochemistry.....................................................................................................58 Rat TBI Model.................................................................................................................58 Brain Tissue Collection and Preparation.........................................................................59 In Vitro Calpain-2 and Caspase3 Digestion of Brain Lysate.........................................60 SDS-PAGE Electrotransfer and Immunoblot Analysis...................................................60 Statistical Analysis..........................................................................................................61 Results........................................................................................................................ .............61 Proteolysis of CRMP-2 Following NMDA a nd MTX Induction in Primary Cortical Neurons........................................................................................................................61 Calpain Inhibition Blocked the Proteolysis of CRMP-2.................................................62 Calpain Inhibition Attenuates NMDA Induced Neuronal Cell Injury and Neurite Damage and Prevents CRMP-2 Redistribution Following NMDA Treatment...........63 CRMP-2 Integrity After TBI...........................................................................................64 Discussion..................................................................................................................... ..........65 4 CRMP-2 IS A NEW CALMODULIN-BINDING PROTEIN...............................................78 Introduction................................................................................................................... ..........78 Materials and Methods.......................................................................................................... .80 Results........................................................................................................................ .............84 Discussion..................................................................................................................... ..........88 5 DIRECT RHO-ASSOCIATED KINASE INHIBITION INDUCES COFILIN DEPHOSPHORYLATION AND NEURI TE OUTGROWTH IN PC-12 CELLS..............100 Introduction................................................................................................................... ........100 Materials and Methods.........................................................................................................103 Chemicals and Antibodies.............................................................................................103 Cell Culture...................................................................................................................103 Quantification of Neurite Outgrowth............................................................................104 Immunoblotting.............................................................................................................104 Immunocytochemistry...................................................................................................105 Results........................................................................................................................ ...........105 Dose-Dependent Neurite Outgrowth Indu ced by ROCK Inhibition in PC-12 Cells.....105 Dynamics of Neurite Outgrowth in ROCK Inhibitor Treated PC-12 Cells..................106 Remodeling of Cytoskeletal Architecture in ROCK Inhibition Mediated Neurite Outgrowth..................................................................................................................107
7 ROCK Inhibition Induces Transi ent Cofilin Dephosphorylation..................................107 Discussion..................................................................................................................... ........108 6 SYSTEMS BIOLOGY APPROACH TO DE CIPHER NEURITOGENESIS: ROCK PATHWAYS IN MEDIATING NEURI TE OUTGROWTH IN PC-12 CELLS.................119 Introduction................................................................................................................... ........119 Experimental and Computational Methods..........................................................................121 Results........................................................................................................................ ...........124 Discussion..................................................................................................................... ........128 7 CONCLUSIONS AND FUTURE DIRECTIONS...............................................................136 LIST OF REFERENCES.............................................................................................................139 BIOGRAPHICAL SKETCH.......................................................................................................158
8 LIST OF TABLES Table page 2-1 Known CaMBPs and potential breakdow n products identified by CaM affinity capture/RPLC-MSMS........................................................................................................43 2-2 Putative CaMBPs identified by CaM-affinity capture/ RPLC-MSMS..............................45 2-3 Functional grouping of putative novel CaMBPs...............................................................48 6-1 The up-regulated genes upon Y-27632 treatment............................................................134 6-2 The down-regulated genes upon Y-27632 treatment.......................................................134 6-3 Proteins with Increased Abundance post Y-27632 treatment..........................................135 6-4 Proteins with decreased Abundance post Y-27632 treatment.........................................135
9 LIST OF FIGURES Figure page 1-1 Ca2+/Calmodulin involved TAI........................................................................................28 1-2 Pathways involved Axonal injury/r egeneration converge at Rho-ROCK.........................29 2-1 Calmodulin binding proteome studies...............................................................................51 2-2 Confirmation of CaM-affinity captu re for 2 known CaM binding proteins ( IIspectrin and calcineurin)....................................................................................................52 2-3 Immuoblot of dynamin, betaII-spectrin as examples of putative CaMBP proteins and their breakdown products...................................................................................................53 2-4 Quality control CaM-affinity purification immunoblot.....................................................54 3-1 Effect of neurotoxins on the integrity of CRMP-1, 2, 4 and 5 in primary cortical neurons........................................................................................................................ .......71 3-2 Effects of calpain and caspase-3 inhi bition on neurotoxin induced proteolysis of CRMP-2 in primary cortical neurons.................................................................................72 3-3 Calpain inhibition attenuated NMDA i nduced neuronal death and prevented CRMP2 redistribution and neurite da mage following NMDA induction.....................................73 3-4 Immunoblot analysis of integrity of CRMPs in rat brain after TBI...................................74 3-5 Time-course of CRMP-2 proteolysis in rat cortex and hippocampus following TBI........75 3-6 Fragmentation patterns of CRMP-2 afte r TBI matches calpain-2 digested CRMP-2 in vitro ............................................................................................................................... .....76 3-7 Sequence analysis of CRMP-2 and pote ntial calpain cleavage sites assignment..............77 4-1 CRMP-2 is a putative CaM binding prot ein. (A) Identification of CRMP-2 as a putative calmodulin binding protein by mass spectrometry..............................................92 4-2 CRMP-2 binds to CaM in a direct and specific Ca2+-dependent manner.........................93 4-3 CRMP-2 co-localizes with F-actin and CaM antagonist W7 altered F-actin organization................................................................................................................... .....94 4-4 CaM antagonist W7 induces filopod ia retraction in CRMP-2 overexpressing 293 cells.......................................................................................................................... ..........95 4-5 CaM modulates calpain-mediated CRMP-2 proteolysis in vitro.......................................96
10 4-6 Elevation of intracellular Ca2+ results in CRMP-2 proteolysis and dephosphorylation..............................................................................................................97 4-7 Okadaic acid enhances CMRP-2 phos phorylation and CaM does not affect the dynamics of CRMP-2 phosphoryaltion.............................................................................98 4-8 Proposed model of the role of calmodulin in CRMP-2 post-translation modification under pathological conditions............................................................................................99 5-1 Neurite outgrowth of PC-12 cells in response to ROCK inhibitor Y-27632 in a dosedependent manner............................................................................................................112 5-2 ROCK inhibitor Y-27632 induced neurit e outgrowth in PC-12 cells in a timedependent manner............................................................................................................113 5-3 Quantification of neurite outgrowth post Y-27632 treatment in PC-12 cells..................114 5-4 Reorganization of cytoskeletal architect ure in ROCK inhibition mediated neurite outgrowth...................................................................................................................... ...115 5-5 Immunoblot analysis of cofilin phosphorylation following Y-27632 treatment.............116 5-6 Effect of various protei n kinase inhibitiors on cofilin rephosphorylation in the presence of Y-27632........................................................................................................117 5-7 Persistent cofilin phosphorylation fo llowing ROCK inhibitor H-1152 treatment...........118 6-1 Correspondence analysis bipot for PC-12 microarray dataset.........................................132 6-2 One representative proteome of RO CK inhibition induced neurite outgrowth visualized on 1D-PAGE following CAX Fractionation..................................................133
11 Abstract of Dissertation Pres ented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy SYSTEMATIC STUDIES OF SIGNAL PA THWAYS IN AXONAL/NEURONA INJURY AND REGENERATION By Zhiqun Zhang December 2007 Chair: Kevin K W. Wang Major: Medical Sciences--Neuroscience It is well known that axons fa il to regenerate in the adult central nervous system (CNS) after injury. Both inhibitory and permissi ve pathways converge at the Rho/ROCK (Rhoassociated kinase) pathway. A systems biology approach to decipher Rho/ROCK inhibition induced neuritogenesis signaling transduction pathways would help significantly in understanding the mechanisms of axonal regenerati on. To achieve this goal first we utilized a pharmacological agent Y-27632 (ROCK inhibitor), which has demonstrated the ability to promote neuritogenesis in various model systems. We found that the ROCK inhibitor can induce robust neurite outgrowth in PC12 cells and it initiated neurit es through dephosphorylation of cofilin. Besides the cofilin change, differentia l gene transcriptome and protein expression influenced by ROCK inhibition we re identified by affymetrix Mi croarray and proteomic studies. More than 200 genes and 20 proteins, many of which are known to be associated with neuritogenesis or cell growth, are potentially involved in RO CK inhibition-induced neurite outgrowth in PC-12 cells. As one of the down-str eam targets of ROCK, brain enriched collapsin response mediator protein-2 (CRMP-2) is of mo st intereste for further study. CRMP-2 plays a crucial role in neurite outgrowth and axon form ation. This is the first time calpain mediated degradation of CRMP-2 after in vivo TBI and in vitro glutamate excitotoxicitic injury has been
12 demonstrated. Moreover, proteolytic CRMP-2 appear s to correlate well with neuronal cell injury and neurite damage. Our subsequent data showed that calmodulin binds to CRMP-2 in a calcium dependent manner, and thereby significan tly retards proteolysis of CRMP-2 both in vivo and in vitro Taken together, data from our studies provide clues regarding the changing factors that are important in the axonal injury/regeneration. The results of these studies lay the foundation for future experiments aimed at fi nding potential biomarkers and th erapeutic targets following axon injury.
13 CHAPTER 1 INTRODUCTION Traumatic Axonal Injury: A General Overview Traumatic axonal injury (TA I) is one of the most comm on and important pathologic features associated with traumatic brain injury (TBI) (Thienpont et al., 2005). Data from Centers of Disease Control and Prevention and other re search groups inferred that there were 26,000 trauma deaths per year due to TAI, with another 20,000 to 45,000 patients suffering with longterm disabilities. Moreover, TAI clinically links to coma and ons et of low Glasgow Coma Scale (GCS) scores. Therefore, TAI has recently been recognized as a key pr edictor of outcome in head and spinal cord trauma (Medana and Esiri, 2003). TAI was initially described by a pathologist as being with microscopic changes due to wallerian-type axonal degeneration. The axonal se gments in the white matter display elongated varicose swelling and axonal bulbs at the termin al stumps of the disconnected axons. These swelling and bulbs are filled with mitochondria, ne urofilaments, and other organelles that have been transported to the axon tip. TAI was used to discribe diffuse axonal in jury. However, it is not a diffuse injury to the whole brain. Rather it commonly occurs in corticomedullary junctions, located in frontal and temporal regions as well as the corpus callosum (CC), brainstem, and deep gray matter (Farkas and Povlishock, 2007). Although damages to axon are universally found in cases of severe, moderate and mild head trauma, the ability to diagnosis of TAI is a big challenge for neurosurgeons. The changes in the axon due to injury as noted on conventi onal brain images scan, such as computed tomography (CT) or standard standard magnetic resonance imaging (MRI) are almost invisible. Fortunately, with the continued development in image technology, the diffusion tensor imaging (DTI), the MRI of diffusion weighted imaging, and magnetization transfer imaging techniques
14 showed promise in revealing TAI by taking adva ntage of the molecular disarrangement of the white matter tracts with diffuse axonal pathol ogy (Hergan et al., 2002; Gallagher et al., 2007; Newcombe et al., 2007). However, it is still difficult to detect mild to moderate degree of TAI, even though structural damages to axons that ca n be seen under microscopy. Furthermore, due to the hostile environment surrounding CNS lesions and the decrease of intrinsic regeneration ability, axons rarely regenerate after injury in the adult cent ral nervous system (CNS). The clinical challenge is to promote compensatory sprouting of the re maining intact axons or the regrowth of several axons across th e injury site (Geddes et al., 2 000). Usually, TAI is associated with a broad pattern of secondary damage cascades that occur in the neuron cell body. Therefore, a systematic approach to deci phering the signaling transduction pathways involved in axonal injury would help significantly assess and deve lop therapeutic strategi es to traumatic axonal injury. In the remaining portion of this chapter, I will focus on discussing TAI and recent progress that has been made in the unde rstanding the mechanisms underly ing TAI and the development of therapies to promote axon regeneration. Specifi c topics will include pathology and calmodulin mediated pathways in TAI, inability of axons to regenerate in the adult CNS, and the therapeutic target offered by the Rho-ROCK path way to enhance axonal regeneration. Pathology and Calmodulin Mediated Path ways in Traumatic Axonal Injury Due to intensive research in both the clinical settings and with experimental animal models of TAI, it is now believed that TAI is a progressive event grad ually evolving from focal axonal alteration to delayed axonal disconnection. Typically, mechanical forces induce direct axon disconnection (primary axotomy). This is a relatively low occurrence with the exception of tissue tearing in the white matter in severe brain injury. The most common cause of white matte r loss is secondary biochemical events that
15 may ultimately result in axon di sconnection with the significant pathologic feature of a bulb formation at the terminal end of the axon (S mith et al., 2003; Buki and Povlishock, 2006). In most of cases the injury process happens not only in the neuron cell bodies, axons and dendrites but also in glial cells. However, axons often extend for long distances from their cell bodies, and therefore are more vul nerable to injury even without related somatic or dendritic alterations. Regardless of injury sites, TAI is associated with a broa dly similar pattern of secondary deleterious cascades in the neuron cell bodies, including disr uption of sodium and calcium channel actvities, abnormal myelin ation, disturbed mito chondrion functions, excitotoxicity, ischemia, oxidative stress and ne uroinfalmmation responses (Povlishock et al., 1999; Farkas and Povlishock, 2007). Physical damages have been thought to be due to direct cause of changes in the activity of Na+ channels leading to patholog ical influx of axonal sodium ion with resultant swelling. This, together with activation of voltage sensitive Ca2+ channels and activation of Na+/ Ca2+ exchangers, has been demonstrated to lead to increased intracellular Ca2+. As a Ca2+ sensor, calmodulin binds to a diverse group of calm odulin binding proteins including enzymes, cytoskeletal proteins, r eceptors and ion channels, and thereby regulates neuronal/axonal response upon stimulus. Calmoldin binding to the pore form ing subunit of voltage-gated calcium channel induces a rapid inactivation of Ca2+ channels which limits the amount of Ca2+ influx in response to injury (Peterson et al., 1999). Furthermore, studies have demonstrated that calmodulin binding to N-Methyl-Aspartate Receptor (NMDA) subunit 1 faci litates calcium-dependent in activation of NMDA receptor, which in turn serves as a negative feedback to fine tune Ca2+ influx after injury (Zhang et al., 1998).
16 As noted, there is immediate microscopic axonal damage after trauma with reduced spacing or compaction of neurofila ments (NFC) within 6 hours. It has been demonstrated that this compaction is coming from the loss or co llapse of neurofilament sidearms, which may mediated by proteolytic or phosphorylation modifica tion of the neurofilament (NF) after calcium influx. Different aspects showed th at calmodulin also mediates the pathology of NFC. It is well established that calmodulin activat es and binds to calcineurin upon Ca2+ influx, altering NF phosphorylation, thereby modifying the repelling forces of the side-arms, leading to formation of NFC (Hashimoto et al., 2000). Recent studies re ported that a group of GMC family proteins including the cytoplasmic protein GAP43, MARC KS and cytoskeleton-associated protein 23 kDa (CAP23) has been found to mo dulate the formation of filopodia and microspikes, as well as neurite outgrowth in the spinal cord regene ration model by regulating actin polymerization, origination and disassemble. These GMC prot eins bind to acid phospholipids PI-(4,5)-P2, calmodulin, protein kinase C and actin filaments in a mutually exclusive manner, modifing the raft-recruitment of signaling molecules, such as src, and in turn are i nvolved in the dynamic interactions that occu r between cell surface and the cytoskel eton core molecules of the growing axon during axon regeneration. Addi tionally, calmodulin binding to these GMC proteins with or without Ca2+ may contribute an important link in tr anslating receptor-mediated calcium fluxes into signals that may be guiding growth cones. This may be done by modifing Ca2+ homeostasis (Gerendasy, 1999; Krucker et al., 2002; van Dale n et al., 2003). The calmodilin involvement in traumatic axonal injury is summarized in Figure 1-1. Excessive calcium influx can cause growth cone collapse. However, with the fall of initial hi gh level intracellular calcium level, growth cone formation is allowed after axonal injury.
17 Given the importance of calmodulin involvement in gene regulation, protein synthesis, axonal transportation and cell motilit y, it is imparative to study the calmodulin signal pathway in a systematic manner after traumatic injury. At the same time excess Ca2+ influx results in the activati on of the non-lysosomal cysteine protease, such as -calpain (calpain-1) and m-calpain (calpa in-2). The result of the activation of calpains is an irreversible cytoskeletal structural and functional proteolysis that invariably leads to cell death (Hayes et al., 1998; Pi ke et al., 1998; Yamashima, 2004). Ca2+ overloading may also open the mitochondrial membrane permeability transi tion (MPT) pore, and subsequently leads to mitochondrial swelling and ultimaely mitochondrial rupture. In turn, cytochrome C release from those abnormal/ruptured mitochondria, with concom itant caspase activation in the axon (Buki et al., 2000). The activation of calpain and caspases c ontinue to devastate in tra-axonal cytoskeletal and organelles leading to the ultimate demise of the axons with a signature pathologic feature of disconnection. To data, no systematic calm odulin binding proteins pathways study after traumatic brain/axonal injury has taken place. Ther efore, one of our goals is to profile all the calmodulin binding proteins and their changes in the brain under pa thological injury perturbations, and further explore calmodulin involved mechanisms underlying the traumatic neuronal/axonal injury. Axons Rarely Regenerate aft er Injury in the Adult CNS It is well known that axons ra rely regenerate after injury in the adult central nervous system. This is the main reason why permanent f unctional damages, such as paralysis and loss of sensation exist in traumatic brain injury or sp inal cord injury patients. During the last two decades, numerous studies showed multiple concurre nt factors contribute to restrict the growth potential of maturing neurons by acting at different levels. Among those factors, the best
18 characterized are the intrinsic re generation abilitys of the neur ons and the hostile environment surrounding CNS lesions (Tang, 2003; Hata et al., 2006). The poor axonal growth outcome of injured mature CNS neurons was first observed by Cajal in the 1927. Subsequently, breakthrough experiments conducted by Aguayo and his colleagues demonstrated that injured neurons in the adult spinal cord ca n regenerate over long distances when the peripheral nervous system (PNS ) grafts were introduced into the lesion site (Aguayo et al., 1981). However, ne urons stopped extending their axons when they were exposed to the CNS environment immediately after leaving the PNS graft. This indicated that the hostile environment around the CNS lesion restricts the a xonal regeneration after injury. A series of subsequent experiments identifi ed some of the environmental inhibitors released from CNS myelin and astroglial scar, which are responsible for the restricti ons of CNS axonal plasticity and regeneration. The characterized inhibitors include Nogo, myelin -associated glycoprotein (MAG), oligodendrocyte-myelin glycoprot ein (OMgp), ephrinB3, and chondr oitin sulphate proteoglycans (CSPGs) (Smith et al., 2003; Xie and Zheng, 2007). Schwab and their colleagues developed a monoc lonal antibody that when it targets Nogo it can neutralize its inhibitory properties in vitro. Furthermore, in vivo application of the antibody strikingly enhances the axonal regeneration and is associated with functional performance improvement in the adult CNS after spinal cord injury. Three different isoforms of Nogo have been found and are due to alternative transcri ption of the Nogo genes. Nogo-A is highly expressed by oligodendrocytes, while Nogo-B and Nogo-C are widely present within and outside the CNS. Nogo-A has two domains with inhibi tory activity on neurite outgrowth. One is extracellular loop of sixty-six am ino acids (Nogo-66), which also exists in two other isoforms. The other is Nogo-A with a unique domain, located in the N-terminal region. Unfortunately, the
19 actual inhibitory determinant has not yet defined. Fortunately, the identification of a receptor for the Nogo-66 (NgR) delineates wh ich residues of Nogo-66 are speci fically responsible for the inhibitory activity (Walmsley and Mir, 2007). It is interesting that the application of an antagonist of Nogo-A effectively promotes axonal regeneration, however the eff ect is incomplete. Independent studies on myelin-associated inhibition discovered other inhibitors, such as MAG and OMgp. MAG is an immunoglobulin superfamily protein and expressed by both CNS and PNS glial cells. One interest feature of MAG is the developmentally regulated function. It is initially promote neurite outgrowth and only becomes inhibitory beyond a specific developmental time point. Since the MAG protein remains unaltered during development, its expre ssion level does not correl ate with axon growth. Furthermore, the distribution of MAG in both CN S and PNS suggests that it may not be as potent as Nogo-A. OMgp is a 120 kDa glycosylphosphotidylinositol (GPI)-anchored protein. This protein is present in the membrane surrounding the Nodes of Ranvier made by oligodendrocyte. Recent in vitro studies suggested the roles for this protei n include growth cone collapse and inhibition of neurite outgrowth. Furthermore, OMgp kock ut mice show elevated collateral sprouting from the CNS nodes of Ranvier, suggesting a potent role for OMgp in rest ricting axonal sprouting under development and physiology condictions. However, the precise role for OMgp in adult CNS axon regeneration after injury is yet to be determined in vivo. Quite remarkably, all three myelin-associa ted inhibitors, Nogo-A, MAG and OMgp, exert their inhibitory effects by binding the GPI-linke d neuronal Nogo-66 receptor (NgR) despite lack of sequence similarity (Yamashita et al., 2005). NgR does not have an intracellular signaling domain; therefore, it needs the transmembran e co-receptors LINGO and p75NTR or TROY (also
20 known as TAJ) to transduce neur ite outgrowth inhibitory signal s. Owing to the observation that NgR can bind to multiple myelin proteins, the desi gn of the specific receptor targets can lead to interventions that can overcome the failure of CNS regeneration. At the site of the CNS injury, a complex cellu lar reaction referred as theglila scar forms through the activation of astrocytes, microglial cells, macrophages and fibroblasts. CSPGs, a group of extracellular matrix inhibitors, are de posited in the dense complicated glial scar structure. Recent studies have identified th e inhibitory role of CSPGs play on axon growth/regeneration both in vitro and in vivo Furthermore, enzymatic degradation of CSPGs by chondroitinase ABC both enhanced axon elongation a nd played a part in significant behavioral improvements after spinal cord injury. Although a great deal attention has been placed on CSPGs, no definitive CSPG receptors have been identified. It has been shown that CSPGs interacts with growth factors, sequestering them away from their receptors (Del Rio and Soriano, 2007). Protein kinase C and the Rh o-Rho kinase (ROCK) pathway ma y also been involved in the barrier effect of CSPGs on neur ite outgrowth (Monnier et al., 2003b). And aside from CSPGs, several other putative axona l growth/regeneration inhibitors have been ascribed to the glial scar, such as tenascin, NG2, neurocan and semaphorin 3. However, the relevant potential of these molecules as inhibitors in axonal grow th need to be further verified. Although the hostile extracellu lar environment around the le sion in the CNS has been thought of as being a dominant inhibitory fa ctor on axon regeneration, the intrinsic growth capacity of the injured neurons is also critical It is not surprising th at embryonic CNS neurons are able to extend long axons in the adult CN S environment. Moreover, isolated adult PNS neurons can grow long axons in the white matter tracts of the adult CNS (Davies et al., 1997; Davies and Silver, 1998). Compared to the embryonic CNS or PNS neurons, this response
21 suggests that mature CNS neurons have signif icantly reduced intrinsi c capacity for axonal regeneration after injury. In a ddition, neutralizat ion of environment inhi bition is not significant enough per se to induce robust axon regeneration. Axons still regenerate poorly in the knockout mice that lack all three isoforms of Nogo or the NgR that mediates the effects of all three myelinassociated inhibitors, Nogo-A, MAG and OM gp (Zheng et al., 2003; Zheng et al., 2006). In order to initiate and sustain axon regene ration a neuron require s the upregulation of specific transcription factors, cy toskeletal elements, growth cone components, and mediators of signal transduction. Comparing the mature axon to developing a xon, some notable changes have been observed: 1) the ability to activate gr owth genes, such as GAP43 and CAP23, and form growth cones declines in adult; 2) age-de pendent cAMP decreases; 3) sensitivity to the neurotrophic growth factors, such as BDNF, NGF and GDNF decr eases; and 4) the capacity for sustained extensive structural remodeling is also reduced. Nume rous observations have reported that induction of neuronal growth genes such as GAP43, allows neuritic elongation overriding myelin-derived inhibition. However, overexpression of GAP43 or CAP23 in transgenetic mice induces spontaneous sprouting but not regeneration. There are a number of factors that switch effects. For example the decline of cAMP during development switches the response of retinal axon s to netrin-1 from attraction to repulsion (Shewan et al., 2002), and the effect of MAG on cerebellar neurons switchs from growthpromoting to growth inhibitory (Cai et al., 2002). There is also ev idence showing that the injection of disbuteryl cAMP into DRG neurons promotes rege neration. Cyclic AMP regulates the transcription of growth rela ted genes through CREB, thereby re sulting in the upregulation of genes such as arginase 1, IL-6 and polyam ine synthesis, which directly promote axon regeneration. Although sensitivity to the neurotrophic factors decrea ses, there is some evidence
22 to suggest that locally or dyna mically applying NGF or BDNF can partially restore the axonal damage. Moreover, priming neurons with neurotrophic factor is ye t another means of stimulating cAMP production to overcome inhibition of mye lin. The term priming is used to describe experiments where neurons are treated overnight w ith neurotrophic factors and are, subsequently, transferred to an inhibitory substrate (Cai et al., 1999). Even though selective strategies targeting a single factor yiel ds some effects, concomitant interventions that enhance intr insic growth capability coupled with block of environmental inhibition might produce more significant regeneration. Molecular Approach to Promote Axonal Rege neration Converge at Rho-ROCK Pathway Several critical signals of convergence with in the developing and regenerating axon for targeting axon outgrowth have been identified. One of such critical convergence point is the Rho-ROCK pathway. RhoA, one of the best characterized of the small GTPases, exists in two states: a biochemically inactive GDP-bound state and an active GTP-bound state. Rho kinase (ROCK) was the first identified RhoA downstream effecter. Two isoforms of ROCK have been described: ROCK I and ROCK II. ROCK II is hi ghly expressed in the muscel and brain, while ROCK I is more ubiquitously dist ributed in all tissues. So far, no funcational difference has been found between these two isoforms. Activated RhoA GTPase binds to ROCK and thereby regulates cell motility, neuronal morphogenesi s, axon guidance and outgrowth, dendrite development, and synapse formation (N g and Luo, 2004; Mueller et al., 2005). With regard to the numerous inhibitors that have been identified as potential obstacles to axonal regeneration, there is accumulating eviden ce that their intracellu lar signaling converge by means of Rho. It has been demonstrated that MAG, NogoA and Omgp can bind to NgR, Lingo and p75 or TROY to form a trimeric receptors complex, and thus activate the Rho-ROCK pathway thereby exhibiting an inhibitory e ffect on axon renegeration (Wang et al., 2002;
23 Yamashita et al., 2002; Lingor et al., 2007; Zhao et al., 2007). Fu rthermore, other environmental inhibitors around lesion, such as CSPG, repulsive guidance molecule (RGM) and members of the semaphorin and ephrin families also stimulat e the RhoAROCK pathway (Lingor et al., 2007). Several laboratories have reported that by bl ocking RhoA activity with either the dominantnegative RhoA or bacterial toxi n Clostridium botulinum exoenzyme C3 transferase (C3), or by inhibiting ROCK with Y27632, prom otes axonal elongation, overcoming the inhibitory effects of myelin, as well as CSPG, semaphorin 3A and EhpA receptor signaling in vitro (Wahl et al., 2000; Monnier et al., 2003a; Yukawa et al., 2005). Most importantly, there is associated other evid ence indicates there is a TBI or spinal cord injury (SPI) induced activation of RhoA and RhoB at the lesion site in human brains (Brabeck et al., 2003; Brabeck et al., 2004) and of ROCKI and ROCKII in rat, respecti vely (Aimone et al., 2004). Intriguingly, the observed of upregulation of RhoA and RhoB was still detectable months after TBI in human (Brabeck et al., 2004) or ove r a period of 4 weeks afte r spinal cord injury (SCI) in rats (Conrad et al., 2005). Moreover, the persistent of the activation of Rho-ROCK pathway around the lesion site makes Rho-ROCK inhibition an attractive therapy not only for acute and sub-acute treatment, but also fo r delayed intervention after CNS injury. While the intrinsic capabilities of neurons to regenerate axons in the mature CNS decrease, there has been renewed interest in improving the neurons intrinsic ab ility to promote axon regeneration. Cyclic AMP activates transcript factor CREB, and this in turn leads to growth gene expression. As an interesting ad junct to the role of cAMP signaling in promoting axonal regeneration, RhoA activity was inactivated by elevated cAMP by PKA phosphorylation. Separate studies also point to the crosstal k that occurs between cAMP and Rho/ROCK by integrating MAG signaling events. A direct correlation between neuronal cAMP levels and axon
24 outgrowth mediated by overcoming the inhibiti on of MAG was found in DRG, cerebellar, cortical and hippocampal neurons. MAG bindi ng to the NgR and p75NTR complex, with subsequent activation of Rho, ex ecutes inhibition of axonal outgr owth (Christensen et al., 2003; Hannila and Filbin, 2007). Locally or dynamically applying neurotrophic factors such as NGF, BDNF and NT3, or priming neurons with neurotrophic f actors enhances axonal regeneration both in vitro and in vivo (Blesch et al., 1998; Blesch and Tuszynski, 2007). It was reported that neurotrophin binding to the p75 low affinity neurotrophin receptor abolishes Rho activit y, as well as in creases cAMP level (Yamashita et al., 2002). Thus, boosting intr insic growth capability of neurotrophin or cAMP may be mediated, at least in pa rt by inhibition of Rho/ROCK pathway. Since both inhibitory and perm issive pathways point to the Rho/ROCK pathway, it makes this pathway an attractive target for the appli cation of an Rho/ROCK inhibitor in order to enhance the regeneration of injured CNS axons (Figure 1-2) Fortunately, a number of specific Rho/ROCK inhibitor co mpounds have been developed, in cluding analogues BA-210, fasudi, HA-1077, Y-27632 (Davies et al., 2000) and wf -536 (Nakajima et al., 2003). Among them, fasudi has already been used in patients with cerebral vasospsm after aneurismal subarachnoid hemorrhage. Remarkably, the local application of C3 exoenzyme or Y-27632 improved functional recovery in mice and rats with transection spinal-cord injuries in different studies. In these studies, ROCK inhibition not only enhanced ax onal growth beyond the lesion site, but also decreased tissue damage and cavity formati on (Lehmann et al., 1999; Dergham et al., 2002; Chan et al., 2005).
25 However, the complete mechanism of ROCK-med iated axonal suppressi on is still unclear. What has been well established is that Rho-ROCK pathway played a central role in cytoskeletal dynamic rearrangement. ROCKs, when activat ed by the small GTPase RhoA, phosphorylate various actin associated protei ns, such as myosin light chain (MLC), myosin light chain phosphatase (MLCP), LIM kinases, calponin, MA RCKs, ERM (ezrin/radixin/moesin), adducin, and profilin IIa. These in turn stimulate ac tin polymerization, which seems lead to cell contraction, stress fiber assembly and growth co ne collapse (Amano et al., 2000; Ohashi et al., 2000; Fournier et al., 2003; Riento and Ridley, 2003). Additionally, ROCKs also phosphoryl ate a group of intermediate f ilament proteins, such as vimentin, glial fibrillary acidic protein (GFAP) and neurofilament L (Kosako et al., 1997; Goto et al., 1998; Hashimoto et al., 1998; Matsuzaw a et al., 1998). Although recent observation suggest it might contribute to the cytokinesis, the exact phosphorylation induced regulation step is not clear. Other downstream targets of ROCKs include microtubule associated protein 2 (MAP2), tau and collapsing response medi ator protein 2 (CRMP-2) (Arimu ra et al., 2000; Amano et al., 2003). These proteins are either es pecially expressed or highly en riched in the brain and spinal cord to regulate microtubule dynamics. Most in terestingly, CRMP-2, whose activity is regulated by phosphorylation, is known to interact with tubulin and Numb and promote microtubule assembly and Numb mediated endocytosis (M imura et al., 2006). P hosphorylated CRMP-2 by ROCK decreases its functions and in turn co ntributes to growth cone collapse and MAGmediated neurite retraction (Yamashita et al., 2002; Arimura et al., 2005). Therefore, Rho-ROCK pathway seems to be an integration point for various signaling pathways restricting axon growth, particularly those regulating cytosk eletal rearrangements
26 (Figure 1-2). From the standpoint of understanding axon growth/regen eration at the system level, it is very important to identif y all components in the ROCK-pro tein pathways linked to axon growth/regeneration. The rat pheochromocytoma cell line, PC-12, has been widely used as an important model system for axon growth. An important feature of PC-12 cells is that they differentiate into a neuronal phenotype in response to various neur otrophins. For instance, nerve growth factor (NGF) treated PC-12 cells exhibit proliferation arrest, neuritogenesis and elec trical excitability (Greene and Tischler, 1976). Moreover, PC-12 ce lls elicit neuritogenesis via Rho inhibition by Clostridium C-3 exoenzyme treatment (Lehmann et al., 1999; Sebok et al., 1999; Fournier et al., 2003), making it an invaluable model system for studying Rho-ROCK transduction pathways in neuritogenesis. As mentioned previously, one major objective of my dissertat ion research is to explore signal transduction pa thways involved in the ax on injury. In order to perform that research the following steps were undertaken. First, the calmodulin signal pathways were ex amined at a systematic level after traumatic injury. A novel CaM-affinity capture coupled reversed liquid chromatography tandem mass spectrometry (PRLC-MS/MS) method needed to be developed to identify the CaM binding proteome and its vulnerability to calpain a nd caspase proteolysis. CRMP-2, a putative CaM binding protein, was further characterized. Next, a systems biology approach was establ ished to decipher the Rho/ROCK signal pathways in neuritogenesis. Syst ems biology is a latest domain in biology that aims at systemlevel understanding of complex biological processes. The scope of systems biology combines multiple advanced research areas, such as bi osciences, data mining, control theory and other
27 engineering fields. A therapeuti c approach by ROCK inhibitor was applied to PC-12 cells to induce neuritogenesis. Differential gene transc riptome and protein expression influenced by ROCK inhibition were iden tified by Affymetrix mciroarray and proteomic studies.
28 Figure 1-1. Ca2+/Calmodulin involved tr aumatic axonal injury.
29 Figure 1-2. Pathways involved in axonal injury/regen eration converge at Rho-ROCK pathway.
30 CHAPTER 2 USING CALMODULIN-AFFINITY CAPT URE TO STUDY THE RAT BRAIN CALMODULIN BINDING PROTEOME AND IT S VULNERABILITY TO CALPAIN AND CASPASE PROTEOLYSIS Introduction Calmodulin (CaM) is one of major calcium (Ca2+) sensors that are cen tral in regulating multiple intracellular events. Ca M exerts its regulatory functi ons through the modulation of a diverse number of CaM-binding proteins (CaMBPs) in a Ca2+ dependent manner (Benaim and Villalobo, 2002; Haeseleer and Palc zewski, 2002; Bouche et al., 2004). With the extraordinarily high concentration of CaM (10 to 100 M) found in neuron of the central nervous system (CNS) (Xia and Storm, 2005), there is si gnificant interest in the downstream pathways involving CaM regulation. Several CaMBPs have already been ch aracterized including adenyl cylases (AC1 and AC8), calcineurin A, CaM-dependent protein kinase I, II, IV, neur onal nitric oxide synthase, and various calcium-ion channels. Known CaMBPs ar e involved in synaptic plasticity, learning, memory (Xia et al., 1993; Xia and Storm, 2005); however; it is necessary to identify all unknown CaMBPs in order to fully construct a calcium-C aM mediated protein network in the brain. In traumatic brain injury (TBI), increased calc ium influx leads to a rapid rise in cytosolic calcium levels and activation of the cysteine proteases calpain and caspase (Wang, 2000b; Pineda et al., 2004). As a result, activated calpai ns and caspases produce irre versible proteolysis, and ultimately leading to necrotic and apoptotic cell deat h (Pike et al., 2001; Pineda et al., 2004). Recently, disruption of calcium signaling was al so recognized as the underlying cause of neuronal dysfunction and associated apoptosis in stroke and Al zheimers disease (AD)(O'Day and Myre, 2004). Previous studies showed that CaMBPs as a group are potential substrates for calpains and caspases, and are vul nerable to brain injury induced proteolysis and deregulation (Wang et al., 1989b; Barnes and Gomes, 1995; McGinnis et al., 1998; M ukerjee et al., 2000;
31 Mukerjee et al., 2001). Fo r example, non-erythroid II-spectrin, an identified Ca2+ dependent CaMBP, is degraded to 150, 145 and 120 kDa fragme nts in TBI. Further studies showed that calpain-2 mediated II-spectrin proteolysis produces a 150 and a 145 kDa fragment (SBDP150 and SBDP145) that are present in neuronal necrosis and apoptosis, whereas the caspase-3 proteolysis produces a sub-150 kDa and a 120 kD a fragment (SBDP150i and SBDP120) that appear exclusively in neuronal a poptosis. Therefore, specific spectrin fragments can be used as markers for the activation of necrosis and apoptosis (Wang, 2000b). The vulnerability of IIspectrin to proteolysis in both necrosis and apoptosis may be just one example of common proteolytic and functional ch anges occurring among CaMBPs after brain injury and in neurodegenerative disease. Therefore, the globa l identification and characterization of CaMBPs and their vulnerability to calpa in and caspase proteolysis may improve our understanding of the pathophysiological mechanisms asso ciated with neuronal injury. In the past, traditional molecular biology t echniques, such as the calmodulin overlay technique (CaMBOT), have been used to isolate and identify i ndividual CaMBPs (O'Day, 2003). A genetic and proteomic approach to screen protein-protein intera ction based expression libraries with CaM has proven to be a powerful way in identifying novel CaMBPs in Arabidopsis and yeast (Zhu et al., 2001; Reddy et al ., 2002). Those techniques have provided remarkable results; however, there are notable limitations. The labo rious procedures of radioisotope labeling, isolation and sequencing of cD NA encoding for CaMBPs, and complicated proteome chip preparation hindered use of those techniques to profile CaMBPs in animals. More recently, a novel mRNA display method was introduced to id entify and characterize human CaMBPs in a comprehensive manner (Shen et al., 2005). The mR NA display method provides a powerful way to read and amplify CaMBPs selected from a large human proteome library. This technique
32 offered improved ability to analyze the lowabundance proteins. Furt hermore, binding CaMmotifs are easily mapped by locating overlapping re gions of selected fragments of the parent protein of interest. However, as an in vitro amplification process this technique requires efficiency and specificity of many steps (transcr iption, translation, fusion etc), thus limits the effective coverage of the CaMbindng proteome. Moreover, this method is a labor intensive process with hundreds of in vitro translation, libr ary amplification, cloning and sequencing steps (Gold, 2001; Takahashi et al., 2003; Shen et al., 2005). Recent developments in mass spectrometry technology, coupled to advances in the fields of bioinformatics, have allowed us to investigate CaMBPs in a high throughput manner. In this study we have exploited Ca2+-dependent Ca MBP interaction with CaM agarose and the subsequent contrary calcium-chelating dissocia tion of CaMBPs from the resin to purify a CaM binding sub-proteome. Following resolution by one dimension sodium dodecyl sulfatepolyacrylamide gel electrophoresis (1D-SDS-PAGE), the purified CaMBPs and their calpain-2 and caspase-3 proteolytic products were profiled by RPLC-MSMS analysis. Materials and Methods Brain Tissue Collection and Protein Extraction Animal surgery procedures were conducted in compliance with the Animal Welfare Act and the University of Florida Institutional An imal Care and Use Comm ittee and the National Institutes of Health guidelines detailed in the Guide for the Care and us e of Laboratory Animals. As described previously (Pik e et al., 2001), adult male (280 300g) Sprague-Dawley rats (Harlan; Indianapolis, IN) were initially anesthetized w ith 4% isoflurane in a carrier gas of 1:1 O2/N2O (4 minutes) followed by maintenance anesthesia of 2.5% isoflurane in the same carrier gas until the animals were sacrificed by decapitation. Nave co rtex tissue was removed, rinsed with ice-cold PBS, and halved. Brain tissues we re rapidly dissected, rinsed in ice-cold PBS, snap-frozen in
33 liquid nitrogen, and stored at -80C until used. The brain samples were pulverized with a small mortar and pestle set over dry ice to a fine powder. The pulverized brain tissue powder was then lysed for 90 minutes at 4C with 50 mM Tris -HCl (pH 7.4), 5 mM EDTA, 5 mM EGTA, 1% Triton X-100, and 1 mM DTT (added fresh). Brain cortex lysates were then centrifuged at 100,000 g for 10 minutes at 4C. The supernatant wa s retained and a DC pr otein assay (Bio-Rad; Hercules, CA) was performed to determine protein concentration. In Vitro Calpain-2 and Caspase-3 Digestion of Brain Lysate Nave cortex lysate was prepared as above. In vitro digestion of rat ly sate (5 mg) with the purified proteases, porcine calpa in-2 (Calbiochem, San Diego, CA) and recombinant human caspase-3 (Chemicon, Temecula, CA) was performed in a buffer containing 100 mM Tris-HCl (pH 7.4) and 20 mM DTT. For calpain-2, 2 mM Ca Cl2 was also added, and then incubated at room temperature for 30 minutes. For caspase-3, sample was incubated at 37C for 4 hours. The protease reaction was stopped by the addition of 30 M calpain inhibitor (MDL 28170) (Calbiochem, San Diego, CA) or 100 M pan-caspase inhibitor (Z -D-DCB) (Calbiochem, San Diego, CA) and a protease inhib itor cocktail solu tion (Roche Biochemicals, Indianapolis, IN). CaM Affinity Capture and Elution Nave brain (control) an d digested lysates (900 g each) were premixed with 600 L (50% slurry) CaM-Sepharose (CaM-agarose) (Sigma, St Louis, MO) in 20 mM Tris-HCl (pH 7.4) and 150 mM NaCl (TBS). EDTA (5 mM) was added to the calpain-2 digest prior to mixing with CaM-agarose. A negative control was also prepar ed using plain lysis buffer with CaM-agarose. After premixing the lysate with CaM-agarose, 10 mM CaCl2 was added and the reacting mixture was continuously incubated for 90 minutes at room temperature to ensure efficient binding. The CaM-agarose bound CaMBPs were collected by cen trifugation, were washed eight times with 1.5 mL TBS with 1 mM CaCl2, and eluted from the resin in 300 L TBS containing 15 mM
34 EDTA. Purified CaMBPs were concentrated via a Millipore YM-10 filter unit (Millopore, Bellerica, MA), suspended in 2x sample buffer (Invitrogen, Carlsbad, CA), and resolved by 1020% sodium dodecyl sulfate-polyacrylamide ge l electrophoresis (SDS-PAGE) visualized by Coomassie blue (Bio-Rad R250) staining. In Gel Digestion Visible bands were excised, cut into four 1 mm cubes, and washed with HPLC water followed by 50% 100 mM ammonium bicarbonate/ 50 % acetonitrile. Pieces were dehydrated with 100% acetonitrile followed by speed vacu um. Cubes were re-hydrated with 10 mM dithiothreitol for 30 minutes at 56 C. Proteins were alkylated with 55 mM iodoacetamide in 50 mM ammonium bicarbonate for 30 minutes in the da rk at room temperature. Gel cubes were then washed with water and dehydrated with 100% acetonitrile. To each tube was added 15 L of a 12.5 ng/ L trypsin solution for 30 minutes at 4 C. Then 20 L of 50 mM ammonium bicarbonate was added and the gel cubes were incubated overnight at 37 C. Peptides were extracted first in water then in 50% water/50% a cetonitrile sequentially. Extracted peptides were dried by speed vacuum and re-suspended in m obile phase solution for capillary RPLC-MSMS. Capillary RPLC-MSMS Based Protein Identification Capillary reversed phase liquid chromat ography tandem mass spectrometry protein identification was performed as described prev iously (Ottens et al., 2005). Briefly, sample digests (2 L) were loaded via an autosampler onto a 100m x 5-cm c-18 reversed phase capillary column at 1.5 L/min. Peptide elution was performe d by linear gradient: 5% to 60% methanol in 0.4% acetic acid over 30 minutes at 200 nL/min Tandem mass spectra were collected in data-dependant mode (3-most in tense peaks) on a Thermo Electron LCQ Deca XP plus ion trap mass spectrometer. Tandem mass spectra were searched against an NCBI rat indexed RefSeq protein database using Sequest Filtering and sorting was performed with
35 DTAselect software by peptide number and Seques t cross correlation values (Xcorr values of 1.8, 2.5, 3.5 for +1, +2, +3 charge states) (Tabb et al., 2002). Peptides filtered and sorted by DTAselect were assigned to specific prot ein accession numbers (National Center for Biotechnology Information [NCBI]). Immunoblot Analysis Purified calmodulin binding proteins and c ontrol samples were separated by SDS-PAGE gel, and transferred to a polyvinylidene difl uoride (PVDF) membrane by the semi-dry method. Membranes were blotted either with antiII spectrin (Biomol Affinity, Exeter, UK), anticalcineurin A (BD-transduc tion, San Jose, CA), antiII spectrin (BD-transduction, San Jose, CA), anti-dynamin (BD-transduction, San Jose, CA ) or anti-calmodulin (upstate, Charlottesville, VA) antibodies, and developed w ith biotin and avidin-conjugat ed alkaline phosphatase. Blots were developed using nitroblue tetrazoliu m and 5-bromo-4chloro-3-indolyl phosphate. Results Calmodulin Binding Proteomic Profil ing by Calmodulin-affinity Capture To explore the rat brain calmodulin binding proteome, we employed a calmodulin-affinity capture method combined with RPLC-MSMS based proteomics. The specificity of the proteome methodology (Figure 1A) is derived from the Ca2+ dependent CaM-CaMBPs interaction. It is also important to note that the presence of endogenous CaM in samp les reduces the efficiency of affinity capture brain CaMBPs to CaM agaros e. To allow efficien t CaM-agarose binding, endogenous CaM-CaMBP complexes we re first dissociated in the presence of EDTA (5 mM). The lysate was then mixed with excess amount of CaM-agarose follo wed by the additional excess CaCl2 (10 mM) to allow association of CaMBPs to CaM-agarose. The optimal excess amount of CaM-agarose for CaM-affi nity capture appears to be 100 g CaM immobilized agarose per 300 g rat brain protein lysate (data not shown). The majority of CaMBPs bound to
36 the CaM-agarose were specifically released in the presence of excess EDTA producing an enriched CaMBP sub-proteome. To assess the vulnerability of those purified Ca MBPs to proteolysis, we performed in vitro calpain-2 and caspase-3 digestions on the same amount of nave brain protein lysate prior to applying the CaM-affinity captu re described above. In all cases, rat brain CaMBPs (and fragments) were selectively el uted from CaM-agarose resin with an EDTA-containing buffer (Figure 2-1A). The nave control CaM binding pr oteome is shown next to the calpain-2 or caspase-3 digested CaM binding proteome in Figure 2-1B by Coomassie blue staining. Sepharose-4B was used as a control to rule out non-specific binding to the resin alone. No detectable proteins were eluted from Sepharo se-4B group compared with the CaM-agarose group (Figure 1C). Identification of CaM Binding Proteome and Calpain/caspase Mediated Breakdown Products by RPLC-MSMS Based Proteomics As shown in Tables 2-1 and 2-2, a total of 69 proteins were identif ied between the three samples separated on the gel in Figure 21B. After checking the CaMBPs database (http://calcium.uhnres.utoronto.c a/ctdb (Yap et al., 2000)) and Medline data mining, we summarized 26 identified known CaMBPs in Tabl e 2-1 with hyperlinked accession number in the NCBI (National Center for Biotechnology In formation) Database and their respective molecular mass and approximate molecular mass on gel, as well as the number of peptides identified and the associated protein sequence coverage found by RPLC-MSMS in nave, calpain and caspase digested samples. Table 2-2 list s 43 putatively novel CaMBPs which include the brain abundant proteins collapsin response medi ator protein 2 (CRMP2), GTPase dynamin 1 and creatine kinase. In analyzing data from Tables 2-1 and 2-2, an apparent reduction in protein mass on gel in relation to the native protein mass was indicative of pr oteolysis in the calpain and
37 caspase digests. This was exemplified by the identification of II-spectrin (natively 280 kDa) at 150 kDa on-gel after calpain-2 and casp ase-3 digestion, the mass of known II-spectrin breakdown products (Pike et al., 2004). Myelin basic protein was also shown vulnerable to calpain-2 digestion while calmodulin dependent phosphodiesterase 1A was found vulnerable to caspase-3 digestion (Table 21). Of 43 putative CaMBPs (Tab le 2-2), pyruvate kinase-3, glutamate oxaloacetate transaminase-1, enolase, albumin precursor pr otein, peptidylprolyl isomerase A, brain myelin proteolipid protei n and alpha H+/K+ transport ATPase were all shown vulnerable to ca lpain-2 proteolysis. II-spectrin, pyruvate kinase-3, glutamate oxaloacetate transaminase-1 and glutamate oxaloacetate transaminase-2 were all shown vulnerable to caspase-3. The proteins II-spectrin, pyruvate kinase-3 and glutamate oxaloacetate transaminase-1 were found sensitive to calpain and caspase digestion in our CaM-binding subproteome. Functional Analysis of Putative ly Novel CaM-Binding Proteins In total, we identified 43 putatively novel CaMB Ps implicated in a wide range of cellular activities, while only a few of these proteins ar e potential background protei ns (marked in Tables 2-1 and 2-2 by *). All putative CaMBPs were searched against the Gene Ontology Consortium (http://www.geneontology.org (Harri s et al., 2004)) and the Human protein Reference Database (http://www.hprd.org (Peri et al., 2003)) public databases, and grouped into categories that define their biological and molecu lar functions, as indicated in Table 2-3. The largest group of putative CaMBPs is involved in metabolism and energy pathways (60%). In the cellular communication and signal transduction category, we identified nine putative CaMBPs (21%), which include four 14-3-3 isoforms, GTPase dynamin-1, CRMP2 and th e G-protein coupled receptor P2Y4. There was also a significant numbe r of CaMBPs implicated in cell growth and maintenance (14%), such as two cytoskeletal a ssociated protein and four structural proteins.
38 Interestingly, we found four mitochondrial proteins that seldom appear in the literature, three of which were susceptible to calpain/caspase di gestion. Thus, the presented methodology may be useful in the study of mitochondrial Ca MBPs and their role in cell death. Immunoblot Analysis of Select CaMBPs Identified by RPLC-MSMS Following CaMAffinity Purification Traditional immunoblots were performed on II-spectrin and calcineurin (Figure 2-2) to validate their proteolytic fragments identified by RPLC-MSMS. Figures 22A and 2-2B showed immunoblots from crude brain lysate before a nd after CaM-affinity pur ification. Figure 2-2A confirmed the extent of proteo lysis of samples in caplain-2 digestion produced SBDP150 and SBDP145 while caspase-3 digestion produced SBDP150i and SBDP120 (Wang, 2000b) (Wang, 2000b). Data in Figure 2-2B correlated with RPLC -MSMS data (Table 2-1) to show intact IIspectrin and the 150 kDa breakdown products after Ca M-affinity binding, an indication that the CaM-binding domain is retained in this fragme nt. Figure 2-2B also show s that calcineurin was proteolysis to 55 and 49 kDa fr agments upon calpain-2 digestion, and that both retained the CaM-binding domain. In contra st, caspase-3 proteolysis of calcineurin produced a 45 kDa fragment that did not contain a CaM binding domain. The putatively novel CaMBP dynamin was examined by Immunoblot in Figure 2-3A. Interes tingly, dynamin was digested by calpain-2 into fragments at 65 and 40 kDa, and by caspase-3 into 42, 40 and 32 kDa, none of which were observed by RPLC-MSMS. This is because only the 65 kDa fragment appeared to retain the CaM-binding domain, which was likely missed due to its low abundance. Figure 2-3D illustrates that intact and caspase-3 digested II-spectrin were also found to elute from the CaM-agarose, validating the RPLC-MSM S data in Table 2. II-spectrin has a putat ive CaM binding domain (Boivin and Galand, 1984; Bignone and Baines 2003) supporting the observation that IIspectrin elutes from CaM-agar ose; however, the possible prot ein-protein interaction with II-
39 spectrin (Boivin and Galand, 1984; Bignone and Baines, 2003) can also explain the presence of II-spectrin, as will be discussed later. Unexpectedly, calmodulin (17 kDa) was observed in the CaMBP elution fraction. Dissociation of CaM from a protein complex, such as with delta subunit of phosphorylase kinase (Dasgupta et al., 1989), was one possibi lity, or that CaM could be an artifact as leaking from the CaM-agarose resin. To verify the source of the Ca M, we used plain lysis buffer (without protein) as a negative control. Figure 2-4 demonstrates that the CaM was also detected in the negative control, which indicates that small amount CaM was leaking from the CaM-agarose resin during the CaM-affinity purification procedure. This is likely due to disruption of non-covalently bound CaM to the agarose beads. To avoid this in th e future, the CaM-agarose should be extensively pre-washed with a concentrated EDTA solution. Discussion To date, this is the first proteomic study of the CaM binding proteome and its vulnerability to calpain and caspase proteolysis in the rat br ain. Most of the known CaMBPs were previously identified by the CaM overlay coupled genom e encoding technique, which is a low throughput method due to the limitations associated with radioisotope labeli ng and DNA sequencing (O'Day, 2003). We present a simple, easy to scale up, high-throughput method for discovery and study CaMBPs which we used to identify 69 known and putative CaMBPs in a single experiment. Of the 26 known CaMBPs (Table 2-1), including II-spectrin, calcine urin, synapsin 1, -synuclein, CaM kinase II, myelin basic protei n and synaptotagmin-1 among others, most are cytoplasmic proteins; however, of the 43 putat ive CaMBPs identified (Table 2-2), a large number were proteins normally localized in mitochondria, such as ubi quinol cytochrome c reductase core protein 2, mito chondrial creatine kinase, mitoc hondrial glutamate oxaloacetate transaminase 2, and mitochondrial malate dehydr ogenase (Table 2-3). Interestingly, only
40 mitochondrial malate dehydrogenase was f ound at nave conditions, while the other mitochondrial proteins were found in the calpain-2 or caspase-3 digests. Mitochondria have been established as a Ca2+ sink during cytosolic Ca 2+ overload that precedes apoptic and necrotic cell death (Carafoli, 2003 ; Saris and Carafoli, 2005). Prior st udies suggested that CaM can be found in mitochondria (Ruben et al., 1980; Ita no et al., 1986; Babcock et al., 1997); however, this is the first study to demonstrate the pres ence of mitochondrial asso ciated CaMBPs in rat brain that are potentially involve d in CaM regulated Ca2+ signali ng cascades during apoptosis or necrosis. Within 12 substrates found in CaMBPs, nine of them are vulnerable to calpain-2. This is consistent with our previous observation th at many CaMBPs are vulnerable to calpain proteolysis (Wang et al., 1989b). Fu rthermore, the immunoblot anal yses in Figure 3 demonstrate that the CaM-affinity / RPLC-MSMS method can efficiently capture and enrich CaMBPs along with their proteolyzed fragments that still co ntain calmodulin binding sites. In addition, our calcineurin results were consistent with a pr evious calcineurin prot eolytic degradation study (Lakshmikuttyamma et al., 2004). Dynamin 1, on e of the GTPase proteins, is generally considered essential for intracellular membrane trafficking events such as synaptic vesicle endocytosis in nerve terminals (Takei et al., 2005; Verma and Hong, 2005). Recent reports indicated the involvement of calcium in dyna min mediated synaptic vesicle endocytosis; however, this is the first report to show that dynamin interacts with CaM (Cousin, 2000; Smillie and Cousin, 2005), providing the basis for inve stigating possible dynamin-CaM modulation of vesicle endocytosis. Previous studies showed that II-spectrin can be eluted from the CaM-column with very weak affinity (Boivin and Galand, 1984; Berglund et al., 1986), whereas a nother study indicated
41 that II-spectrin passed through the CaM-affinity column (Glenne y and Weber, 1985). In our study, immunoblot analysis demonstrated that II-spectrin and its proteolyzed fragments were captured by the CaM-affinity resin. Since II-spectrin and II-spectrin form a high affinity tetramer, II-spectrin and its proteolyzed fragment s elute possibly due to protein-protein interactions with II-spectrin (Bignone and Baines, 2003). In a similar fashion, clathrin heavy chain subunits also bind to a light chain subuni t(Ybe et al., 1999) that is a known CaMBP. Additionally, using a yeast twohybrid screen and in vitro Ca M overlay assay, human adaptor protein 14-3-3 epsilon protein was found to inte ract with human calmodulin (Luk et al., 1999). Whereas, seven 14-3-3 isoforms, including zeta, eta, gamma, were also sh own to interact with each other through direct or indirect associati on (Jin et al., 2004). More stringent conditions might be needed to rule out some indirect inte ractions. Taken together, these would suggest that some putative CaMBPs listed in Table 2-2 ma y bind to CaM through similar protein-protein interactions, presenting the possibility of a comp lex CaM binding network that can be elucidated from the presented results. Though CaM-affinity purification is an effi cient method to enrich Ca2+ dependent CaMBPs, some highly abundant brain proteins might be non-selectively retained by the CaM resin, such as M2 pyruvate kinas, pyruva te kinase-3, glyceraldehyde-3-phosphate dehydrogenase, phosphoglycerate mutase 1, and -actin (Shevchenko et al., 2002). In the future, proteins that non-specific bind could be minimized by incubating the samples first with plain agarose beads and then applying the supe rnatant for CaM-agarose purification. To address these issues, the putative CaMBPs we identified need to be further confirmed by an independent method such as biotinylated CaM overlay technique (O'Day, 2003).
42 In conclusion, the combined CaM-affinity capture / RPLC-MSMS method is a simple and efficient way to explore the Ca M-binding proteome and its vulne rability to proteolysis. The putative CaMBPs and their calpain/caspase substrat es identified in this study will be helpful in constructing a protein-protein in teraction map that may help to explain how brain cells respond in terms of initiating proteases to Ca2+ stimuli. This approach can easily be applied to other biological samples and may provide promisi ng potential in finding novel biomarkers and therapeutic targets in TBI, stroke and other neurodegenerative diseas e which involve calcium dyshomeostasis(Lopez-Otin and Overall, 2002; O'Day, 2003; O'Day and Myre, 2004; Kurz et al., 2005; Liu et al., 2005).
43Table 2-1. Known CaMBPs and potential breakdown produc ts identified by CaM affinity capture/RPLC-MSMS. Nave Calpain-2 Caspase-3 Protein Name Accession no. MW of Protein(kDa) MW from gel(kDa) Matched peptides Sequence Coverage(%) MW from gel(kDa) Matched peptides Sequence Coverage(%) MW from gel(kDa) Matched peptides Sequence Coverage(%) # Alpha-II spectrin P16086 272 280 3 1.2 280 30 17 150 6 2.9 150 5 2.9 Clathrin, light polypeptide NP_446287 30 2 9.8 Plasma membrane calcium ATPase 4 NP_001005871 129 140 3 3.4 Heat shock protein 90 AAT99569 83.3 90 2 3 90 4 6.5 Synapsin 1 NP_062006 74.1 75 2 2.3 Heat shock protein 8 NP_077327 70.8 70 5 6.3 70 2 4.5 70 2 4.5 Heat shock protein 70 CAA54424 70 70 7 8.7 70 5 6.9 70 1 2.5 Calcineurin, catalytic Subunit, alpha isoform NP_058737 58.7 55 4 10.7 55 3 6.7 55 4 6.7 CaM kinase II beta subunit S68470 72.7 55 1 2 55 1 2 CaM kinase II gamma Subunit S43845 56 55 1 2 55 1 2 CaM-kinase II delta chain NP_036651 56.4 200 1 2.4 50 2 4.6 CaM-kinase II alpha chain NP_037052 55.3 200 1 2.7 50 1 2.7 50 3 8 50 1 2.7 # Phosphodiesterase 1A, calmodulin-dependent NP_110498 62.3 55 1 1.7 Synaptotagmin 1 S09595 47.4 55 3 8.1 Tubulin, alpha 1 NP_071634 49.9 50 12 24.2 50 20 26.7 52 1 4 Tubulin, beta 5 NP_775125 49.9 50 1 2.2 Tubulin, beta2-like NP_954525 49.9 50 14 41 52 4 14.2 Neuron-specific class III beta-tubulin AAM28438 49.9 50 5 15 50 4 12.4 52 1 2.2
44Table 2-1 (continued) Myristoylated alanine rich Protein kinase C substrate A39773 30 75 1 4.9 Neuronal axonal membrane protein Nap-22 Q05175 21.8 22 2 18.6 Sirtuin 2 AAH86545 43.3 34 2 10 Guanine nucleotidebinding protein, beta-2 NP_112299 37 32 2 6.2 Guanine nucleotidebinding protein, beta-3 NP_068630 37 32 1 2.9 Guanine nucleotidebinding protein, beta 1 AAH78809 37 32 2 5.9 Synuclein, alpha NP_062042 14.5 17 2 18.6 17 2 19.3 # Myelin basic protein NP_058722 17 17 6 23 13 3 14.4 Synaptobrevin 2 NP_036795 12.6 13 1 14.7 # undergoing proteolysis; potential bac kground proteins (Shevchenko et al., 2002)
45Table 2-2. Putative CaMBPs identified by CaM-affinity capture/ RPLC-MSMS. Nave Calpain-2 Caspase-3 Protein Name Accession no. MW of protein(kDa) MW from gel (kDa) Matched peptides Sequence Coverage (%) MW from gel(kDa) Matched peptides Sequence Coverage (%) MW from gel(kDa) Matched peptides Sequence Coverage (%) # Spectrin beta 2 isoform 1 NP_00101314 8 274 280 3 1.4 180 9 3.8 Clathrin, heavy polypeptide NP_062172. 192 200 5 4.1 200 2 1.6 200 2 1.4 Collapsin response mediator protein 2 P47942. 62.3 180 2 4.7 180 1 2.8 62 2 3.7 62 3 6.1 Na+/K+ -ATPase alpha 3 subunit NP_036638 116 100 4 6.3 100 5 7.1 Na,K-ATPase alpha-1 AAA41671 112 100 4 6.3 100 6 8.7 100 1 1.8 Na,K-ATPase alpha-2 NP_036637 112 100 1 1.8 Dynamin 1 NP_542420 97 95 1 1.2 Microtubule-associ ated protein 6 NP_058900 86.5 150 4 6.4 Aconitase 2, mitochondrial NP_077374 85.5 85 3 6.7 Syntaxin binding protein 1 NP_037170 67.7 65 6 9.4 65 5 9.4 M2 pyruvate kinase AAB93667 57.8 55 5 13 Pyruvate kinase-3 NP_445749 57.8 55 5 13. 8 55 1 3.4 50 1 2.7 ATPase, H+ transporting, V1 subunit B, isoform 2 AAH85714 56.6 50 2 5.1 ATP synthase alpha chain P15999 59.8 50 2 4.7 52 2 4.7 ATP synthase beta subunit AAB02288 56.3 50 5 14 50 5 11. 5 #* Enolase NP_036686 47.2 46 5 19.1 33 1 3 G-protein coupled receptor P2Y4 CAA75007 46 44 1 4.2 44 2 18.6
46Table 2-2 (continued) Glucose phosphate isomerase NP_997475 62.8 50 1 3 Glutamine synthase AAH87131 42 38 1 2.1 40 1 2.1 Glutamate dehydrogenase 1 NP_036702 2',3'-cyclic nucleotide 3'Phosphodiesterase AAA64429 47.3 45 2 5.5 45 4 10. 7 45 3 8.5 Creatine kinase, mitochondrial 1 AAH91335 47.5 43 1 3.8 Creatine kinase, brain NP_036661 42.7 100 4 16 100 2 4.2 100 5 15. 7 43 3 13 43 2 9.7 43 6 21. 5 Alpha 2 actin P62738 42 40 3 9.8 40 3 9 Cytoplasmic beta-actin AAH63166 41.7 40 2 7.5 40 2 13. 2 Ubiquinol cytochrome c reductase Core protein 2 AAH83610 48.2 40 1 3.5 #Glutamate oxaloacetate Transaminase 1 AAH61877 46.3 38 3 11 38 4 4.4 Similar to Transcriptional activator Protein PUR-alpha XP_226016 34.9 34 1 2.8 Glyceraldehyde-3-phosphate Dehydrogenase AAH87743 33 34 5 23 34 3 13 # Glutamate oxaloacetate Transaminase 2 NP_037309 47.4 38 3 4.4 Malate dehydrogenase 1, NAD (soluble) NP_150238 36.4 34 6 21 34 1 3 34 3 10. 5 Malate dehydrogenase, mitochondrial NP_112413 35.5 32 8 20 32 7 25. 7 # Albumin precursor, PRO0883 Protein ABRTS 69.4 33 1 2.5 Lactate dehydrogenase A NP_058721 36.7 32 2 6.7 32 2 3 Lactate dehydrogenase B NP_036727 36.7 32 1 4.8
47 Table 2-2 (continued) L4-3-3, zeta NP_037143 28 28 4 19 28 2 10. 6 L4-3-3, eta NP_037184 28 28 3 17 28 1 5.7 L4-3-3, gamma NP_062249 28 28 2 9.7 L4-3-3, theta NP_037185 28 28 2 9 28 1 3.3 Phosphoglycerate mutase 1 JC1132 28.9 28 2 8.3 Calmodulin AAH58485 17 17 1 6 17 2 7.4 # Peptidylprolyl isomerase A; cyclophilin A NP_058797 18 13 2 20. 2 # Myelin proteolipid protein AAA41898 30.1 25 4 16. 2 # H+/K+-ATPase, alpha AAB93902 115 10 1 1.4 # undergoing proteolysis; potential bac kground proteins (Shevchenko et al., 2002).
48 Table 2-3. Functional grouping of putative novel CaMBPs Cellular communication of Signal transduction (21%) Adaptor Molecules Notes L4-3-3, zeta expressed in golgi and cytoplas m involved in complex scaffold activity L4-3-3, eta expressed in golgi and cytoplas m involved in complex scaffold activity L4-3-3, gamma expressed in golgi and cytoplas m involved in complex scaffold activity L4-3-3, theta expressed in golgi and cytoplas m involved in complex scaffold activity Transport/cargo protein Syntaxin binding protein 1 ATPase, H+/K+ transport., alpha Cytoskeletal associated protein Collapsin response mediator protein 2 GTPase Dynamin 1 primary function a GTPase activ ity involved in ve sicle endocytosis G protein coupled receptor G-protein coupled receptor P2Y4 Cell growth and/or maintenance (14%) Cytoskeletal associated protein Spectrin beta 2 isoform 1 Microtubule-associat ed protein 6 Structural protein Clathrin, heavy polypeptide Alpha 2 actin Cytoplasmic beta-actin Brain myelin proteolipid protein
49Table 2-3 (continued) Metabolism and Energy pathways (60%) Transport/cargo protein Na+/K+ -ATPase alpha 3 subunit Na,K-ATPase alpha-1 Na,K-ATPase alpha-2 ATPase, H+ transporting, V1 subunit B Isoform 2 ATP synthase alpha chain ATP synthase beta subunit Enzyme Aconitase 2, mitochondrial Hydratase Enolase Hydrolase 2',3'-cyclic nucleotide 3'phosphodiesterase Phosphodiesterase Ubiquinol cytochrome c reductase core Protein 2 Reductase Glutamate oxaloacetate transaminase 1 aminotransaminase Glutamate oxaloacetate transaminase 2 aminotransaminase Creatine kinase, mitochondrial 1 phosphotransferase Creatine kinase, brain phosphotransferase Phosphoglycerate mutase 1 Mutase Peptidylprolyl isomerase A; cyclophilin A Isomerase Glucose phosphate isomerase Isomerase Malate dehydrogenase 1, NAD (soluble) Dehydrogenase Malate dehydrogenase, mitochondrial Dehydrogenase Glutamate dehydrogenase 1 Dehydrogenase Glyceraldehyde-3-phosphate Dehydrogenase Dehydrogenase Lactate dehydrogenase A Dehydrogenase Lactate dehydrogenase B Dehydrogenase Glutamine synthase Dehydrogenase M2 pyruvate kinase phosphtransferase Pyruvate kinase-3 phosphtransferase
50Table 2-3 (continued) Unclassified (5%) Similar to Transcriptional activator protein PUR-alpha Albumin precursor, PRO0883 protein *proteins localized in mitochondria
51 (A) (B) Figure 2-1. Calmodulin binding proteome studies A) Schematic diagram of novel calmodulinaffinity purification RPLC-MSMS met hodology to study the rat brain calmodulin binding proteome. B) The intact and cal pain-2/caspase-3 digested CaM binding proteome visualized by Coomassie blue st aining. Each visible band was excised, ingel digestion and analyzed by RPLC-MSMS. C) Nave brain lysate was incubated with Sepharose-4B as a control. Sepharos e-4B eluent was compared to the CaMagarose eluent
52 Figure 2-2. Confirmation of CaM-affinity capture for 2 known CaM binding proteins ( IIspectrin and calcineurin). A) and their br eakdown products. (A) Control samples from nave and degraded brain lysates before CaM-agarose purification. (B) CaMBPs from CaM-agarose-elution in the presence of 15 mM EDTA were separated by SDS-PAGE and analyzed by immunoblot for IIspectrin and calcineurin A. Arrow head indicates II-spectrin breakdown product 150i.
53 Figure 2-3. Immuoblot of dynamin, be taII-spectrin as examples of putative CaMBP proteins and their breakdown products. (A, C) dynamin and II-spectrin were found degraded in vitro by calpain-2 and caspase-3 in rat cortex (crude brain lysate ). (B, D) intact dynamin and II-spectrin and their calpain-2/ caspase-3 breakdown products were found to have putative calmodulin bindi ng domains (CaM-agarose elution)
54 Figure 2-4. Quality control CaM-affinity purif ication immunoblot. Plain lysis buffer was incubated with CaM-agarose as a negative control, and compared with nave brain lysate that was CaM-affinity purified.
55 CHAPTER 3 CALPAIN-MEDIATED COLLAPSIN RESP ONSE MEDIATOR PROTEIN-1, 2 AND 4 PROTEOLYSIS FOLLOWING NEUROTOXIC AND TRAUMATIC BRAIN INJURY Introduction Collapsin response mediator prot eins (CRMPs) are a family of cytosolic proteins that are highly expressed in the brain (Wang and Strittm atter, 1996). They have been shown to be involved in different aspects of axonal outgrowth, neuronal morphogenesis and cell death (Charrier et al., 2003). CRMP-1, 2 and 5 play an essential role in growth cone collapse in response to repelling guidance cues, such as sema phorin 3A or lysophosphatid ic acid (Arimura et al., 2004; Bretin et al., 2005). CMRP-4 is highly expressed in post-mitotic neurons in the early embryonic brain and appears to be involved in brain development. Further, CRMP-4 is also found in brain regions that reta in the capability of neurogenesi s or axonal outgrowth and/or synaptic rearrangement during adulthood (Liu et al., 2003). CRMP-2, the first CRMP discovered, is found c oncentrated in growing axons, dendrites, and the cytoplasm of differen tiating neurons. Specifically, CR MP-2 is important in the determination of neuronal polarity and axonal el ongation (Inagaki et al., 2001; Fukata et al., 2002a). Additionally, when CRMP-2 is highly phosphorylated it may play a role in neurodegeneration, as observed in neurofibrilla ry tangles in Alzheimer's diseased brains (Yoshida et al., 1998; Gu et al., 2000; Butterfie ld et al., 2003; Uchida et al., 2005). A growing body of evidence suggests that CRMP-2 may also participate in the pathophysiology of other neurological disorders. Decreased expression of CRMP-2 has been re ported in fetal brains with Downs syndrome (Weitzdoerfer et al., 2001), pati ents with mesial temporal lobe epilepsy (Czech et al., 2004), focal rat brain ischemia (C hung et al., 2005) and in the frontal cortex of patients who suffer from psychiatric disorder s (schizophrenia, bipolar or major depression disorders) (Johnston-Wilson et al ., 2000). In contrast, an increa se in CRMP-2 has been observed
56 after chronic anti-depressant treatment in th e rat hippocampus (Khawaja et al., 2004). CRMP-2 has also been reported to mediate axonal da mage and neuronal death via a semaphorine-CRMP pathway (Barzilai et al., 2000; Gu and Ihara, 2000). Although the pat hophysiology of neuronal injury varies among neurological disorders (e.g. TB I, stroke, epilepsy, and Alzheimers disease), evidence points to CRMP-2s involvement in excitotoxic mechanisms of neuronal damage (Arundine and Tymianski, 2003). Recently, CRMP-3 was found truncated in response to excitotoxicity in vitro and cerebral ischemia in vivo by activated calpain. Calpain-cl eaved CRMP-3 was observed to translocate into the nucleus evoking neuronal cell death (Hou et al., 2006). We suspected that the other CRMPs might also have an association with excito toxic neuronal cell injury, since the sequence homology is high among CRMP family members ( 50-75%) (Schweitzer et al., 2005). Therefore, in this study, we examined the integrity of CRMP-1, 2, 4 and 5 following in vitro neurotoxin treatments and in vivo traumatic brain injury (TBI) in rats. Materials and Methods Primary Cortical Neuron Culture Primary cortical neurons from post-natal day one SpragueDawley rat brains were plated on poly-L-lysine coated culture pl ates (Erie Scientific, Portsmouth, NH, U.S.A.). In detail, cells were dissociated with trypsin and DNase I, re suspended in 10% plasma-derived horse serum (PDHS) in Dulbeccos modified Eagles medium (DMEM), and plated on poly-L-lysine treated 35 mm (density: 3.0 106 cells per well) plates. Cells were allowed to grow in an atmosphere of 10% CO2 at 37C for three days and then treated with 1 M 4-amino-6-hydrazino-7--Dribofuranosyl-7H-pyrrolo (2,3-d)-pyrimidine-5carboxamide (ARC) for two days. The ARC was removed and fresh 10% PDHS in DMEM was adde d after which the cells were grown for an additional 10 days. All animal studies confor m to guidelines outlined in the Guide for the
57 Care and Use of Laboratory Animals from the National Institutes of Health and are approved by the University of Florida. Neurotoxic Challenges and Pharmacologic Intervention In addition to untreated controls, the followi ng conditions were used: maitotoxin (MTX) (3 nM; WAKO Chemical USA Inc., Richmond, VA) as a calpain-dominated challenge for three hours; apoptotic inducer st aurosporine (STS) (0.5 M; Sigma, St. Louis, MO) for 24 hours; the Ca2+ chelator ethylene diam ine tetra-acetic acid (EDTA) (5 mM; Sigma) for 24 hours as a caspase-dominated challenge (Waterhouse et al., 1996; Chiesa et al., 19 98; Mizuno et al., 1998; McGinnis et al., 1999); and NMDA (200 M; Sigma) for 3 to 24 hours as an excitotoxic challenge. For pharmacologic intervention, cultures were pretreated one hour before the MTX, EDTA or NMDA challenge with either the calpa in inhibitor SJA6017 (Sen ju Pharmaceuticals, Kobe, Japan) (Fukiage et al., 1997; Kupina et al., 2001), or the pancaspase inhibitor Z-VAD (OMe)-FMK (R&D, Minneapolis, MN). Lactate Dehydrogenase Release Assay of Cell Death A lactate dehydrogenase (LDH) release assay (CytoTox One Reagent, Promega, Madison, WI) was performed to assess cell death. Primary co rtical neurons were seeded at 3.0 x 106 cells per well in 6-well plates, cultured for two weeks as described above, and then pretreated for one hour before introducing NMDA (200 M) with or without ca lpain inhibitor SJA6017 (30 M) or caspase inhibitor Z-VAD (20 M) in DMEM (Glucose 50 mM). Culture media were collected at three-hour intervals and assaye d for LDH release by following th e manufactures instructions. Three replicates for each time point were assayed.
58 Cell Lysate Collection and Preparation Primary neuronal cells were collected and lysed for 90 min at 4oC with a lysis buffer containing 50 mM Tris (pH 7.4) 5 mM EDTA, 1% (v/v) Triton X-100, 1 mM DTT, and a MiniComplete protease inhibitor cockta il tablet (Roche Biochemicals, Indianapolis, IN). The lysates were centrifuged at 10,000 g for 5 minutes at 4C to remove inso luble debris, and then were snap-frozen and stored at -80oC until use. Immunocytochemistry Primary cortical neurons were fixed with 4% paraformaldehyde (PFA) in PBS for 10 minutes, washed with PBS and permeabilized with 0.1% triton X-100 in PBS for 5 minutes. Immunocytochemistry staining was performed fo llowing a one-hour blocking step in 10% goat serum at room temperature. The neurons were incubated overnight at 4C with monoclonal mouse-anti-CRMP-2 (C4G, IBL, Aramachi, Takasa ki-shi, Gunma, Japan) at a dilution of 1:500. Alexa 488-conjugated goat-anti-mouse secondary antibody (Molecular Probes, Eugene, OR) was added at a dilution of 1:1000, followed by washi ng with PBS. The cells were mounted using medium containing 4, 6-diamidine-2-phenylindol e (DAPI) (Vector Laboratories, Burlingame, CA). Fluorescence images were captured w ith a 20x objective on the Zeiss Axiovert 200 Fluorescence Microscope with a CCD camera and merged using SPOT imaging software (Diagnostic Instruments, Sterling Heights, MI). Rat TBI Model A controlled cortical impact (CCI) device was used to mode l TBI (Dixon et al., 1991; Pike et al., 1998). Adult male (280-300 g) Sprague-Daw ley rats (Harlan: Indianapolis, IN) were anesthetized with 4% isoflura ne in a carrier gas of oxygen (4 min.) followed by maintenance anesthesia of 2.5% isoflurane in the same ca rrier gas. Core body temperature was monitored continuously by a rectal thermistor probe and was maintained at 37 1OC by placing an
59 adjustable temperature controll ed heating pad beneath the rats Animals were mounted in a stereotactic frame in a prone position and secured by ear and in cisor bars. A midline cranial incision was made and a unilateral (ipsilateral) craniotomy (7 mm diameter) was performed adjacent to the central suture, midway between bregma and lambda. The dura mater was kept intact over the cortex. Brain tr auma was produced by impacting th e right cortex (ipsilateral cortex) with a 5 mm diameter aluminum impact or tip (housed in a pneumatic cylinder) at a velocity of 3.5 m/s with a 1.6 mm compression a nd 150 ms dwell time (compression duration). Sham-injured control animals underwent identical surgical procedures, but did not receive an impact injury. Appropriate preand post-inj ury management was maintained to ensure compliance with guidelines set forth by the Univer sity of Florida Institutional Animal Care and Use Committee and the National In stitutes of Health guidelines detailed in the Guide for the Care and Use of Laboratory Animals. Brain Tissue Collection and Preparation At the 8 post-CCI time points (2, 6, 24 hour s and 2, 3, 5, 7, 14 days), animals were anesthetized and sacrificed by decapitation. Brains were immedi ately removed, rinsed with ice cold PBS and halved. Two different brain regi ons (cortex and hippocampus ) were removed from the right and left hemispheres, rinsed in ice cold PBS, snap-frozen in li quid nitrogen, and stored at C until use. For Western blot analysis, the brain samples were pulverized to a fine powder with a small mortar and pa stel set over dry ice. The pulverized brain tissue was then lysed for 90 minutes at 4C with lysis buffer containing 50 mM Tris-HCl (pH 7.4), 5 mM EDTA, 5 mM EGTA, 1 % Triton X-100, and 1 mM DTT (adde d fresh). Brain cortex lysates were then centrifuged at 10,000 g for 10 minutes at 4C. The supernatant was retained and a DC protein assay (Bio-Rad, Hercules, CA) was performed to determine protein concentration. Nave cortex lysate was prepared as above.
60 In Vitro Calpain-2 and Caspase-3 Digestion of Brain Lysate In vitro digestion of rat lysate (5 mg) wa s performed with the pur ified proteases human erythrocyte caplain-1, rat recombinant calpain-2 (Calbiochem, San Diego, CA) and recombinant human caspase-3 (Chemicon, Temecula, CA) in a buffer containing 100 mM Tris-HCl (pH 7.4) and 20 mM DTT. For the calpains 2 mM CaCl2 was also added to the lysate and then incubated at room temperature for 30 minut es. For caspase-3, samples were incubated at 37C for four hours. Protease reactions were stopped by the ad dition of either calpain inhibitor SJA6017 to a concentration of 30 M (Senju Pharmaceuticals, Kobe, Japan) or the pan-caspase inhibitor ZVAD to a concentration of 20 M and a protease inhibi tor cocktail solution. SDS-PAGE Electrotransfer a nd Immunoblot Analysis Protein concentrations of cell or tissue lysa tes were determined via Bio-Rad DC Protein Assay (Bio-Rad, Hercules, CA). Protein balanc ed samples were prepared for sodium dodecyl sulfate-polyacrylamide gel electro phoresis (SDS-PAGE) in a two-fold loading buffer containing 0.25 M Tris (pH 6.8), 0.2 M DTT, 8% SDS, 0.02% bromophenol blue, a nd 20% glycerol in distilled water. Samples were heated for 90 s econds at 90C, then centrifuged for 2 minutes. Samples were routinely resolved by SDS-PAGE on Tris-glycine gels for 2 hours at 130V. Following electrophoresis, separated proteins were laterally transferred to polyvinylidene fluoride (PVDF) membranes by the semi-dry method. Rat CRMP-5 antibody (targeting residues 369 Chemicon, Temecula, CA), dialyzed with PBS using Slide-A-Lyzer MINI Dialysis Units, (Pierce, 3.5MWCO, 69550, Rockford, IL), wa s then biotinylated using EZ-link, sulfoNHS-LC-LC-biotin following the manufactures inst ructions. Membranes were incubated with either anti-CRMP-1, or -4 (t argeting residues 499-511, Chemicon) biotinylated-anti-CRMP-5 antibodies or anti-CRMP-2 (C4G) (IBL) or a C-terminal anti-CRMP-2 antibody made at the University of Florida (raised against a synt hetic peptides of residues 551-559), and then
61 developed with biotin (except for CRMP-5) and avidin-conjugated alkaline phosphatase and nitroblue tetrazolium and 5-bromo-4-chloro-3-i ndolyl phosphate. Quantitative evaluation of protein levels was performed via computer-assis ted densitometric scanni ng (NIH ImageJ version 1.6 software) (Zhang et al., 2006b). Statistical Analysis All experiments described were performed at least in triplicate. Densitometric values represent the mean S.E.M. Statistical significance wa s determined using a one-way ANOVA test, with a significance level of p < 0.01. Results Proteolysis of CRMP-2 Following NMDA an d MTX Induction in Primary Cortical Neurons In our previous studies, the ionotrophic glut amate receptor agonists N-methyl-D-aspartate (NMDA) and Ca2+ channel opener maitotoxin (MTX ) were used to model acute neuronal injury in primary cortical neurons (Wang et al., 1996; Na th et al., 2000a; Dutta et al., 2002). With this method, the integrity of CRMP-2 following neurot oxin induction was examined. In this study, a marked reduction in the intact CRMP-2 (66 kDa and 62 kDa) was noticed along with the appearance of a 55 kDa band af ter excitotoxic injury (200 M NMDA) in rat primary cortical neuron cultures (Fig. 3-1A). Similar results we re observed after MTX treatment (Fig. 3-1A). Two anti-phospho-CRMP-2 specific antibodies (3 F4 and C-terminal Phospho-CRMP-2) were used to rule out the possibility that the 55 kDa band was due to de-phosphorylation. The altered profile of the 66 kDa and 62 kDa CRMP-2 wa s not observed after NMDA and MTX treatment (data not shown). It appears the 55 kDa fragment is a like ly breakdown product of CRMP-2. Suspecting that a similar proteolytic event may occur with other members of the CRMP family, we examined the integrity of CRMP-1, 4 and 5 under identical conditions (Fig. 3-1B). Decrease
62 in intact CRMP-1 and CRMP-4 were observed, as well as the increase of a 58 kDa CRMP-4 BDP. However, CRMP-5 levels remained unchange d following neurotoxic treatment. We further explored CRMP-2 dynamics following NMDA (200 M) induction using a time course analysis. The 55 kDa CRMP-2 BDP appeared by 3 hours and became prominent within 24 hours (Fig. 31C). Importantly, the densitometr ic analysis showed that the reduction of intact CRMP-2 was paralleled by the increased 55 kDa BDP over time (Fig. 3-1D). Calpain Inhibition Blocked the Proteolysis of CRMP-2 To rule out non-specific CRMP2 proteolysis during protei n extraction, we looked for matching CRMP-2 degradation in co rtical cultures pretreated with either the specific calpain inhibitor SJA6017 or the pan-casp ase inhibitor Z-VAD prior to neurotoxin treatment. The results show that the formation of the 55 kDa frag ment was blocked by SJ A6017, but not by Z-VAD (Fig. 3-2). Therefore, calpain appears to be re sponsible for proteolysis of the intact 62 kDa CRMP-2 protein into the 55 kDa breakdown product (BDP). Interestingly, despite preservation of the intact 62 kDa CRMP-2 by SJA6017, the 66 kD a CRMP-2 levels declin ed as explained in the Discussion section (Fig. 3-2). Next, the apoptosis inducer st aurosporine (STS, 0.5 M), a ca lpain and caspase activator challenge, and the calcium chelator EDTA (5 mM ), a caspase-dominant challenge, were used on primary cortical neurons (McG innis et al., 1999; Nath et al ., 2000b; Wang, 2000a). The results show that the intact 62 kDa CRMP-2 was rapi dly degraded into the 55 kDa BDP upon STS treatment, but not upon the caspase-activated ED TA treatment. STS-mediated generation of the 55 kDa CRMP-2 BDP was also effectivel y blocked by SJA6017 while Z-VAD offered no protection (Fig. 3-2, upper pa nel). The production of the 55 kDa CRMP-2 BDP strikingly paralleled the production of the 150 and 145 kDa II-spectrin breakdown products, which were
63 monitored as markers for calpain activity in NMDA and STS treatment (Nath et al., 2000b) (Fig. 3-2, lower panel). Calpain Inhibition Attenuates NMDA Induced Neuronal Cell Injury and Neurite Damage and Prevents CRMP-2 Redistribut ion Following NMDA Treatment LDH release assays were performed to inve stigate the role of calpain and caspase activation of NMDA induced neur onal cell injury and CRMP-2 degradation. The release of LDH, normally present in the cytoplasm of neurons into the cell culture media can be used as a measure of dying cells (Riss and Moravec, 2004 ). The results showed that NMDA treatment induced CRMP-2 proteolysis in a time-depende nt manner (Figs. 3-1C, D). NMDA treatment induced significant neuron death after 3 hours,, pe aking at 24 hours, which is consistent with the production of the 55 kDa CRMP-2 BDP. Moreover, the calpain inhib itor (SJA6017) provided significant protection for about 6 hours, while the caspase inhibito r (Z-VAD) offered no protection from NMDA treatment (Fig. 3-3A). We next examined the levels of CRMP-2 following 6 hours of NMDA treatment with and without the calpain and caspase inhibitors to further explore the association of CRMP-2 with NMDA induced neurite damage (F ig. 3-3B). Figure 3-3B show s neurons with healthy, long neurites (control, upper pane l). Under higher magnification, CRMP-2 is more prominently observed in neurites than in the cell body (arrow, control, lower panel) After six hours of NMDA challenge, neurites significantly retr acted (NMDA, upper panel). Under higher magnification, CRMP-2 redistribution to the ce ll body (arrow, lower panel) was observed. In contrast, neurites in most neurons appeared to be preserved with the calpain inhibitor pretreatment (N+SJA, upper panel). The re distribution of CRMP-2 to the cell body was prevented (N+SJA, lower panel). On the ot her hand, Z-VAD offered no protection and was unable to prevent CRMP-2 redistribution (N+VAD, Fi g 3-3B). Taken together, these data further
64 suggest that proteolysis of CRM P-2 may contribute to calpain-me diated neurite degeneration and cell death after NMDA challenge. CRMP-2 Integrity After TBI Next, we examined whether the CRMPs were proteolyzed in vivo following traumatic brain injury. We examined the contralateral and ipsilateral cortex and hippocampus tissues harvested 48 hours post-TBI since this is when ca lpain activation is most significant (Pike et al., 1998). A notable decrease in inta ct CRMP-1, 2 and 4 were observed in the ips ilateral cortex and hippocampus while at the same time the 55 kD a CRMP-2 BDP and 58 kDa CRMP-4 BDP were found to increase (Fig. 3-4A). However, there was no significant change in the intact CRMP-5 after TBI. Examination of the contralateral tissu e showed no noticeable d ecrease in the intact CRMPs and no production of BDPs. Tissue examin ation also verified that CRMP degradation only appeared following TBI, and not in sham (c raniotomy only) animals (Fig. 3-4B). Together, these results suggest that CRMP-2 proteolysis occu rred following TBI in our animal model and it is probable that the same occurred with CRMP-1 and CRMP-4. Post-CCI cortical and hippocampal rat brai n tissue was used to assess the temporal dynamics of CRMP-2 following TBI. The amount of intact 62 kDa CRMP-2 decreased from 6 hours to 3 days with a corresponding increase of the 55 kDa BDP (Fig. 3-5). The change was significant for both the intact a nd the 55 kDa CRMP-2 BDP at 24 a nd 48 hours for the ipsilateral cortex (Figs. 3-5A, B). Signifi cant changes in the hippocampus were, also, observed between 24 hours and 3 days (Figs. 3-5C, D). Interestingly, the level of intact and cleaved CRMP-2 seems to return to control levels by day 5 after TB I both in the cortex and hippocampus based upon Western blot analyses. Spectrin proteolysis wa s used to correlate CRMP-2 degradation with calpain and caspase activity (Fig. 3-5). We foun d that the formation of the SBDP150/145 calpain product paralleled the formation of the 55 kDa CRMP-2 BDP demonstrating that CRMP-2
65 proteolysis correlated well with the calpain act ivity over time after TBI. CRMP-2 fragmentation patterns are similar between TBI and brain lysate digestion. We previously identified calpain involvement in CRMP-2 proteolysis following TBI, just as we had in cell culture after ne urotoxin treatment (Figs. 3-1 and 3-3). To further test this as shown in Fig. 3-6, in vitro calpain-2 treatment of nave brain lysate resulted in the same fragmentation pattern observed after TBI in vivo. The 62 kDa and 66 kDa intact CRMP-2 bands disappeared. Pretreatment of the nave lysate with the calpain inhi bitor (SJA6017) blocked formation of the 55 kDa CRMP-2 BDP and preser ved the 62 kDa CRMP-2 bands. In contrast caspase-3 treatment did not produce the 55 kD a BDP, although the 66 kDa intact CRMP-2 band did fade even after applying the caspase inhi bitor Z-VAD (see Discussion). We confirmed the complete inhibition of calpain a nd caspase activity by parallel mon itoring of calpain and caspase associated II-spectrin degrada tion (Fig. 3-6). We then verifie d, using Phoretix 1D-gel imaging software, that the molecular weight of the calpain mediated CRMP2 proteolytic product matched that of the 55 kDa BDP observed post-TBI. Therefore, this suggests that calpain was involved in the cleavage of the 55 kDa CRMP-2 BD P, while caspase-3 was not. Taken together, these data suggest that CRMP-2 proteolysis is due to calp ain activation following TBI. Discussion In this work we demonstrated for the first time the degradation of CRMP-1, 2, and 4 after acute neuronal injuries by in vivo TBI and in vitr o glutamate excitotoxicity (Figs. 3-1 and 3-4) and that the proteolytic enzyme involved was calpain-2. As the current manuscript was under review, Bretin and colleagues also reporte d that CRMP-1, 2, 4 ar e proteolyzed by NMDA treatment in cortical neurons (Bretin et al., 2006). Chung et al. (Chung et al., 2005) also recently reported proteolysis of CRMP-2 in focal ischemic rat brain, however, the responsible protease was not identified. In agreement with those findings, our study demonstrates that CRMP-1, 2 and
66 4 are proteolyzed after neurotoxic injury and TBI. Moreover, we have confirmed that the decrease of intact CRMP-2 occurs with a conc urrent increase of a 55 kDa CRMP-2 fragment due to calpain-2 proteolysis, in vi tro and in vivo, and that the calp ain-mediated CRMP-2 proteolysis appears to be associated with neuron al cell injury and neurite damage. The data demonstrate CRMP-2 was proteo lyzed into a 55 kDa BDP under two calpaindominant challenges (MTX and NMDA) and with the apoptotic inducer STS in cortical neurons (Figs. 3-1A and 3-2). Interesti ngly, the decrease of the intact CRMP-2 and increase of CRMP-2 55 kDa BDP following NMDA treatment (Figs. 31C, D) noticeably paralleled NMDA induced neuronal cell death over time (F ig. 3-3A). In addition to at tenuating cell death induced by NMDA treatment for about 6 hours (Fig. 3-3A), pretr eatment of primary cortical culture with the cell-permeable calpain-inhibitor SJA6017, and not the caspase inhibitor Z-VAD, prevented the formation of the 55 kDa CRMP-2 BDP, preserving the intact 62 kDa CRMP-2 protein (Fig. 3-2). Furthermore, calpain inhibition prevented redi stribution of CRMP-2 from neurites to the cell body, and preserved the architecture of neurites (Fig. 3-3B). These data suggested CRMP-2 proteolysis may be linked to calpain mediated neurite degeneration. Any further relationship between calpain-mediated truncatio n of CRMPs and neurite damage that may exist has yet to be determined. In addition to LDH assay, other met hods such as 3-(4, 5-di methylthiazol-2-yl)2,5diphenyl-tetrazolium bromide (MTT) assay, TUNEL assay Hoechst and propidium iodide (PI) staining may be used to conc ordantly assess and confirm ne uronal cell injury or death. Consistent with other therapeutic approaches using other calpain inhibitors in vitro (Adamec et al., 1998; Araujo et al., 2004; Lopez-Picon et al., 2006) and in vivo (M arkgraf et al., 1998; Kupina et al., 2001), preincubati ng with the calpain inhibitor (SJA6017) did not effectively protect neurons from excitoxic in jury after 12 hours. This suggest s, in addition to the calpains,
67 other pathways, such as calmodulin (Pohorecki et al., 1990), phospholipase protein kinase A and so forth (Adamec et al., 1998; Llansola et al ., 2000), may also contribu te to the excitotoxic neuron death. It was also noticed that proteoly sis of CRMP-2 was only exhibited in the injured brain regions where there was the formation of the CRMP-2 55 kDa BDP which appears to correlate with calpain activation over time follo wing TBI (Figs. 3-4 and 3-5). The apparent rebound of CRMP-2 by day 5 after TBI may be due to necrotic tissue loss thereby leaving the remaining more intact and/or recovering tissue for sampling (Sahuquillo et al., 2001; Liu et al., 2006). Finally, treatment of nave brain lysate with purified and activated calpain-2, but not caspase-3, resulted in the 55 kDa cleavage product identical to th at observed after TBI (Fig. 3-6). Taken together, these data demonstrate that ca lpain mediated CRMP-2 pr oteolysis occurred and that it appears to be linked to neuronal cell inju ry and neurite damage after excitotoxic injury. Interestingly, both calpain and caspase treatm ent resulted in the disappearance of the 66 kDa CRMP-2 band, whereas only calpain treatment induced formation of the 55 kDa CRMP-2 BDP (Fig. 3-6). Previous studies showed that the 66 kDa CRMP-2 is a phosphorylated form of CRMP-2 while the 62 kDa form was unphosphorylated. Work by Gu et al. demonstrated that following incubation of brain ly sate with EGTA, the 66 kDa CRMP-2 was dephosphorylated to the 62 kDa CRMP-2 form (Gu et al., 2000). Our caspase digestion buffer contained EGTA and EDTA the disappearance of the 66 kDa CRMP-2 band following caspase-3 treatment was probably due to dephosphorylation, but this was not preventable by the pan-caspase inhibitor (Fig. 3-6). CRMP-2 has two domains, a dihydropyrimidinas e (DHPase) domain (residues 64-413), and a C2 domain (residues 486-533), illustrate d in Fig. 3-7A following SBASE analysis (http://hydra.icgeb.trieste.it/sbase). Even though CRMP-2 has high sequence similarity to
68 DHPase, it has no known enzyme activity (hydro toinase or dihydrophrimidinase) (Wang and Strittmatter, 1997). Important phosph orylation sites are located with in the C2 domain (Gu et al., 2000; Cole et al., 2004; Yoshimura et al., 2005). Sequence analyses also suggest there are multiple putative calpain cleavage sites within th e C-terminus region of CRMP-2 (Fig. 7) with the preferred residues Leu, Thr and Val in posit ion P2, and Lys, Tyr and Arg in position P1 (Tompa et al., 2004). Along with the C4G anti body, whose epitope is found in the large boxed sequence in Fig. 3-7 (residue 486 to 528) (Gu et al., 2000), another C terminal antibody was used (epitope residue 551-559) to narrow down the possi ble cleavage sites involved in forming the 55 kDa CRMP-2 BDP (Arimura et al ., 2005) (Fig. 7). With the latt er antibody, the intact CRMP-2 was shown to decrease following excitotoxic treatment and TBI, but the 55kDa BDP was not observed (data not shown). This suggests that the cleavage site is lo cated between residue 486 and 551. There are also two PEST regions (res idues 496-511 and 535-552) located on the Cterminal end of CRMP-2 (residues 486-559) as identified by PESTfind analysis (http://www.at.embnet.org/embnet/ t ools/bio/PESTfind). PEST regions tend to depict regions of rapid degradation in proteins. A previous correlation between the presence of a PEST site and calpain susceptibility provided evidence that calp ain was the enzyme that degraded PEST regioncontaining proteins (Wang et al., 1989a). The first PEST region RGLYDGPVCEVSVTPK (residues 496-511) contains preferred calpain cleavage site Leu498-Tyr499-Asp. Furthermore, cleavage at residue 499 will result in a trunc ated CRMP-2 with a theoretical mass 54.7 kDa matching the experimental mass of 55 kDa for the CRMP-2 BDP. The other putative calpaincleavage sites highlighted in Fig. 3-7 may also meet the criteria of the calpain digestion, but their theoretical fragment mass would be larger than the observed 55 kDa BDP. Thus, after analyzing the sequence and based upon our experimental data we propose that the most likely calpain
69 cleavage is between residues LY499 D500 in CRMP-2 (Fig. 3-7). Nterminal microsequencing is underway to confirm the exact cleavage site. The biological significance of calpain-mediated CRMP-2 proteoly sis is not yet clear. An important function of CRMP-2 is its involveme nt in axonal regenera tion or elongation. Overexpression of CRMP-2 induces the formation of multiple axons and primary axon elongation. Truncation of 24 amino acids or more from the Cterminus abolishes the ability of CRMP-2 to promote multiple axons (Inagaki et al., 2001). St udies have shown that phosphorylation at Thr514 or Thr509, Ser518 and Ser522 of CRMP-2 by GSK3 and Cdk5 regulates axon outgrowth (Fig. 3-7) (Gu et al., 2000; Cole et al., 2004; Yoshimura et al., 2005). Additionally, recent studies showed that Rho-kinase (ROCK) phosphorylates CRMP-2 at Thr-555 (Arimura et al., 2000) inhibiting the ability of CRMP-2 to bi nd tubulin and Numb, th ereby inducing growth cone collapse (Nishimura et al., 2003; Arimura et al., 2005). Furthermore, CRMP-2 binds to the light chain of kinesin-1 (KLC ) at region 440-572, regulating sol uble tubulin transport to the distal part of the growing axon (Kimura et al ., 2005). Based on these previous studies, it is apparent that calpain tr uncated CRMP-2, between LY499 D500, would remove all the phosphorylation sites and disrupt CRMP-2 in teractions with KLC or ROCK, thereby deregulating axonogenesis and neur onal polarity (Fig. 3-7). In addition, calpain-cleaved CRMP2 may also be involved in neuronal cell damage. A recent study showed that over-expression of the CRMP-2 C413 mutant, which lacks the C-terminal residues 413, led to CRMP-2 leakage into the nuclear compartment, i nducing apoptosis (Tah imic et al., 2006). In addition to CRMP-2, this report moreove r showed that CRMP-1 and CRMP-4 were also degraded after excitotoxic neuronal injury and TBI (Figs. 1, 4). These results are consistent with the recent study on calpa in-mediated truncation of CRM P-4 in response to NMDA and
70 H2O2 toxicity (Kowara et al., 2005). Until now, there has been no report on the degradation of CRMP-1 following neurotoxic injury or TBI. Ho wever, it was reported that truncation of 183 amino acids from the C-terminus caused a si gnificant decrease of neurite formation and extension following NT3 treatment of DRG neur ons (Quach et al., 2004). CRMP-3 was, also, recently shown to be vulnerable to calpain proteolysis, and that calpain-cleaved CRMP-3 translocated into the nucleus, inducing neur onal cell death (Hou et al., 2006). Although all CRMPs high sequence homology, CRMP1 and 4 exhibit the closest identity with each other (68-75%), while CRMP-5 has the relatively least identity with the rest of family members (4950%). As a result, CRMP-5 may best be classified as a member of a different subfamily (Fukada et al., 2000). It is therefore not surprising that CRMP-5 levels we re unaffected after TBI. Further studies are being conducted to el ucidate the pathophysio logical significance of the cleavage of CRMP-2 in the nervous system. This may give us new insights into the mechanism of axon or neuron damage in the central nerv e system, such as may occur afte r spinal cord injury, TBI and stroke.
71 Figure 3-1. Effect of neurotoxins on the integrity of CRMP-1, 2, 4 and 5 in primary cortical neurons. NMDA (200 M) was used as an excitotoxic challenge and maitotoxin (MTX) (3 nM) was used as calpain-domin ant neurotoxin challenge in rat primary cortical neuron culture. (A) Primary cort ical neurons were exposed to NMDA (200 M) for 6 hours or maitotoxin (3 nM) for 3 hours. Total protein extracts were separated by SDS-PAGE and immunobl otted with anti-CRMP-2 (C4G) and -actin. A marked reduction of intact CRMP-2 was noticed along with the appearance of a 55 kDa band following NMDA and MTX treatment (B) The same samples were probed with CRMP-1, CRMP-4 and CRMP-5 antibod ies. Decrease of CRMP-1 and CRMP-4 with a 58 kDa and 62 kDa doublet was observed following NMDA and MTX treatment; however, no remarkable change in CRMP-5 was detected. Results shown are representatives of three experiments. (C) NMDA induced CRMP-2 proteolysis in a time-dependent manner. Primary cortical neurons were subjected NMDA treatment (200 M) at different time points. A repres entative Western blot probed with CRMP2 is shown. (D) Densitometric analysis of intact CRMP-2 and BDP 55 kDa following NMDA treatment at the various time points. Values were means S.E.M. n = 3.
72 Figure 3-2. Effects of calpain and caspase-3 inhibition on neur otoxin induced proteolysis of CRMP-2 in primary cortical neurons. Primary cortical neurons were pretreated with calpain inhibitor (30 M SJA6017) and cas pase-3 inhibitor (20 M Z-VAD) for one hour, and then exposed to NMDA (200 M) for 6 hours, apoptosis inducer STS (0.5 M) and EDTA (5 mM) for 24 hours. Total prot eins in the cell lysates were resolved by SDS-PAGE and immunoblotted with anti-CRMP-2 (C4G) (top). Paralleled IIspectrin (middle) proteolysis was monitore d as a marker of calpain and caspase activity. -actin (bottom) was used as a loading control. Calpain inhibition preserved the intact CRMP-2, while caspase in hibition did not. Results shown are representatives of three experiments.
73 Figure 3-3. Calpain inhibition at tenuated NMDA induced neurona l death and prevented CRMP-2 redistribution and neurite damage fo llowing NMDA induction. Primary cortical neurons were pretreated with calpain (30 M SJA6017) and caspase-3 (20 M ZVAD) inhibitors for one hour or with serum free medium alone, and then exposed to NMDA (200 M) for various hours. (A) Ti me course of LDH release following NMDA treatment. Values represent means S.E.M. n = 4. A difference was considered to be statistically significant when the P value was less than 0.01 (* P < 0.01). (B) Immunocytochmistry of CRMP-2 (C4G) following NMDA treatment (6 hours) with or without calpai n and caspase inhibition. Arrows indicate the accumulation of CRMP-2. Scale bar=50m
74 Figure 3-4. Immunoblot analysis of integrity of CRMPs in rat brai n after TBI. (A) The integrity of CRMPs (1, 2, 4 and 5) in Nave (N) and 48 hours post-TBI (T) contralateral, ipsilateral cortex and hippocampus (Hippo). Those tissue lysates were subjected to immunoblotting and probed with CRMPs and -actin antibodies. (B) Nave, shams and 48 hours post-TBI ipsilateral cortex ly sates were subjecte d to immunoblotting and probed with CRMP-2 (C4G) (top), and -actin (bottom) antibodies. A marked decrease in the 62 kDa CRMP-2 was noti ced with the appearance of a 55 kDa CRMP-2 band. -actin was used as a loadi ng control. Results shown are representative of three experiments.
75 Figure 3-5. Time-course of CRMP-2 proteolysis in rat cortex and hippocampus following TBI. The brain samples were collected from nave, ipsilateral cortex (A) and ipsilateral hippocampus (B) with TBI (1.6mm impact) 2 hours to 14 days after injury (2h, 6h, 24h, 48h, 3d, 5d, 7d, 14d). Brain lysates were subjected to immunoblotting and probed with CRMP-2 (C4G) (top), II-spectrin (middle) or -actin (bottom) antibodies. Quantitative analysis of the intact CRMP-2 (62 kDa) and 55 kDa breakdown product of CRMP-2 in ipsilateral cortex (C) and ipsilateral hippocampus (D) were done by densitometry an alysis. Values represent means S.E.M. n = 3. *p<0.01 compared with nave.
76 Figure 3-6. Fragmentation pattern s of CRMP-2 after TBI matche s calpain-2 digested CRMP-2 in vitro Nave cortex lysate incubated with ca lpain-2 or caspase-3 in the presence or absence of calpain inhibitor (30 M SJA 6017) or caspase inhibitor (20 M Z-VAD). The same amount of TBI ipsilateral cortex 48 hours after injury and in vitro calpain-2 or caspase-3 treated nave cortex lysate were analyz ed by immunoblots and probed with CRMP-2 (C4G) antibody. Calpain treatment resulted in CRMP-2 proteolysis to 55 kDa BDP that matched fragmentation pattern after TBI in vivo Fragmentation was blocked by calpain inhibitor, but not by caspase-3 inhibitors. II-spectrin proteolysis was monitored in parallel as a mark er of calpain and caspase activity.
77 Figure 3-7. Sequence analysis of CRMP-2 and potential calpai n cleavage sites assignment. Part A shows the domain architecture of CRMP-2 predicted by SBASE SVM (http://hydra.icgeb.trieste.it/s base). Part B depicts a sc hematic representation of residues 486-559 toward the C-terminus of CRMP-2. PEST regions are underlined, while portions used for generating CRM P-2 antibody (C4G) (residue 486-528) and CRMP-2 (c-terminal) antibody (residue 551-559) are boxed. Arrows beyond the box indicate putative calpain cleavage sites.
78 CHAPTER 4 CRMP-2 IS A NEW CALMODULIN-BINDING PROTEIN Introduction CRMP-2, also known as CRMP62, TOAD-64 (turne d on after division 64 kDa), Ulip-2 (Unc-33-like phosphoprotein), a nd DRP2 (dihydropyrimidinase-re lated phosphoprotein), is one of at least five isoforms (CRMP1-5) in th e CRMP family (Wang and Strittmatter, 1996, 1997). CRMP-2 is a developmentally regulated protein and highly expressed in the nervous system. Two splice isoforms of CRMP-2, namely CRMP-2 A and 2B, have been identified. CRMP-2A, the long N-terminal isoform is specifically associ ated with axons of th e corpus callosum, the bundles of the striatum, and the mossy fibe rs of the hippocampus, whereas CRMP-2B is ubiquitously distributed throughout the cell body, in the axons or dendr ites (Yuasa-Kawada et al., 2003; Bretin et al., 2005). A growing body of evidence has demonstrated the importance of CRMP-2 and its phosphorylated isoforms in various neuronal functions. CRMP-2 was first identified as an intracellular protein involved in the transducti on cascade initiated by semaphorin 3A (Sema 3A). Sema 3A induced neuronal grow th-cone collapse which is mediated via the sequential cdk5 and GSK3 phosphorylation of CRMP-2 on Ser 522 and Thr509 (Goshima et al., 1995; Ito et al., 2000). In contra st, when lysophosphatidic acid (LPA) induces growth-cone collapse, CRMP-2 is phosphorylated on Thr-555 by Rho-kinase, downstream of RhoA(Arimura et al., 2000). CRMP-2 has also b een reported to regulate neuron al polarity and axon elongation. Overexpression of CRMP-2 induces the forma tion of multiple axons and elongation of the primary axon in hippocampal neurons(Fukata et al., 2002b; Fukata et al., 2002a). A number of studies have demonstrated that CRMP-2 interacts with a number of proteins, including tubulin, kinesin, Numb, Sra/WAVE an d phospholipase D2, to exhibit its functions (Fukata et al., 2002b; Lee et al., 2002; Nishimura et al., 2003; Kawano et al., 2005; Kimura et al.,
79 2005). CRMP-2 binds to tubulin heterdimers to promote microtubule assembly and regulates polarized Numb-mediated endocytosis of neuronal adhesion molecule L1 in the growth cone. CRMP-2 interacts with Sra-1/WAVE1, a regulator of the actin cytoskeleton, which in turn transports the protein complex to distal part of the growi ng axon in a kinesin-1 dependent manner and thereby regulating axon outgrowth and formation (Kawano et al., 2005). In addition, our recent CaM affinity capture coupled with reversed-phase liquid chromatography tandem mass spectrometry (RPLC-MSMS) profiling CaM-bi nding proteome suggests that CRMP-2 may also interact with CaM (Zhang et al., 2006a). Upon elevation of intracellular Ca2+, CaM undergoes conformational change and binds to different CaM binding proteins, resulting in multiple cellular functions. Many of the downstream targets that CaM binds to and activates are unab le to bind calcium themselves; therefore they need CaM as a calcium sensor and signal transdu cer. With the extraordinarily high concentration of CaM (10 to 100 M) found in neuron of the central ner vous system, CaM binding to enzymes, cytoskeletal proteins, receptors and ion cha nnels, regulates neurona l/axonal response upon stimulus(Wang et al., 1989b; O'Day, 2003; Xia a nd Storm, 2005). For example, CaM binding to calcinuerin and neurofilament (N F) regulates the proteolysis and phosphorylation of NF, which in turn modifies axonal injury after calcium influx (Johnson et al., 1991). Similarly, CRMP-2 has been reported undergoes proteolysis or phosphorylation modification under different neurobiological disorders, such as epilepsy, st roke, traumatic brain injury and Alzheimers disease (Gu et al., 2000; Cole et al., 2004; Czech et al., 2004; Chung et al ., 2005; Jiang et al., 2007; Zhang et al., 2007). One common feature of th ese diseases is the al teration of intracellular Ca2+ homeostasis which appears to pl ay a central role in the mechanisms of the neuronal/axonal
80 injury that underlies th ese diseases (Sola et al., 1999; Smith et al., 2003; Czogalla and Sikorski, 2005; Johnston, 2005). Given the relevance of the Ca2+/CaM signaling system in the processing of Ca2+ signals and taking into account the impor tance of CRMP-2 in the axon formation and elongation, it is possible to hypothesize that CaM binding may in fluence the proteolysis or phosphorylation of CRMP-2, thereby modulating CRMP-2 involve d cytoskeleton reorganization upon Ca2+ influx. To address these questions, we validated the interaction between CaM and CRMP-2 and then further investigated the possible roles that CaM may play in post translational modification of CRMP-2 and CRMP-2 related actin reorganization in response to neuronal or axonal injury. Materials and Methods Animal Surgery, Brain Tissue Coll ection and Protein Extraction Animal surgery procedures were conducted in compliance with the Animal Welfare Act and the University of Florida Institutional An imal Care and Use Comm ittee and the National Institutes of Health guidelines detailed in the Guide for the Care and us e of Laboratory Animals. Adult male (280-300 g) Sprague-Dawley rats (Harla n: Indianapolis, IN) we re anesthetized with 4% isoflurane in a carrier gas of oxygen (4 mi n.) followed by maintenance anesthesia of 2.5% isoflurane in the same carrier gas. Cortex, hi ppocampus brain tissues were removed, rinsed with ice-cold PBS, and halved. Brain tissues were ra pidly dissected, rinsed in ice-cold PBS, snapfrozen in liquid nitrogen, and stored at -80C until us ed. The brain samples were pulverized with a small mortar and pestle set over dry ice to a fine powder. The pulve rized brain tissue powder was then lysed for 90 minutes at 4C with 50 mM Tris-HCl (pH 7.4), 5 mM EDTA, 5 mM EGTA, 1% Triton X-100, and 1 mM DTT (added fresh). Brain cortex lysates were then centrifuged at 100,000 g for 10 minutes at 4C. The supernatant was retained and a DC protein assay (Bio-Rad; Hercules, CA) was perfor med to determine protein concentration.
81 Cell Culture Primary cortical neurons from first post-nata l day SpragueDawley rat brains were plated on poly-L-lysine coated culture pl ates (Erie Scientific, Portsmouth, NH, U.S.A.). In detail, cells were dissociated with trypsin and DNase I, re -suspended in 10% plasma-derived horse serum (PDHS) in Dulbeccos modified Eagles medium (DMEM), and plated on poly-L-lysine treated 35 mm (density: 3.0 106 cells per well) plates. Cells were a llowed to grow in an atmosphere of 10% CO2 at 37C for three days and then treated with 1 M 4-amino-6-hydrazino-7--Dribofuranosyl-7H-pyrrolo (2,3d)-pyrimidine-5-carboxamide (ARC) for two days. The ARC was removed and fresh 10% PDHS was added in DMEM after which the cells were grown for an additional 10 days. All animal studies confor m to guidelines outlined in the Guide for the Care and Use of Laboratory Animals from the National Institutes of Health, and are approved by the University of Florida. PC-12 cells were main tained at 37C in Dulbecos modified Eagles medium (DMEM) supplied with 10% fetal bovine serum (FBS), 5% heat-inactivated horse serum, 100 g/ml of streptomycin, 100 U/ml of penici llin and 1% Fungizone (Gibco, Rockville, MD) in a humidified 5% CO2 incubator. HEK293 cells were maintained in DMEM supplied with 10% FBS, 100 g/ml of streptomycin, 100 U/ml of penicillin and 1% Fungizone. Plamid Constructs and Transfection The cDNA of human CRMP-2 was amplif ied by PCR from IMAGE clone#3870039 using the primers 5-ACCAGAATTCAGATGT CTTATCAGGGGAAGAAAA-3 and 5ATCAGTCGACCTAGCCCAGGCTGGTGATGTT-3. Pr imers were purchased from IDT (Coralvile,IA). The PCR products were subcloned into pAcGFPV1 (Clontech). HEK293 cells were transfected with GFP-hCRMP-2 or G FP vector using Lipofectamine 2000 (Invitrogen, Carlsbad, CA, USA). Stable GFP-hCRMP-2 expr ession HEK293 cells were selected by using media containing 400 ug/ml G418.
82 Cell Lysate Collection and Preparation Primary neuronal cells or PC-12 cells were collected and lysed for 90 minutes at 4oC with a lysis buffer containing 50 mM Tris (pH 7.4) 5 mM EDTA, 1% (v /v) Triton X-100, 1 mM DTT, with or without a Mini-Com plete protease inhibitor cocktail tablet (Roche Biochemicals, Indianapolis, IN). The lysates were centrif uged at 100,000 g for 5 minutes at 4C to remove insoluble debris, and then were snap-frozen and stored at -80oC until use. Preparation of Biotinylated Calmodulin Purifed bovine or human brain CaM (Calbiochem, San Diego, CA) was dialyzed with PBS using Slide-A-Lyzer MINI Dialysis Units (Pierce, 3.5MWCO, 69550, Rockford, IL). The concentration of dialyzed CaM was determined by a DC protein assay (Bio-Rad; Hercules, CA). Then CaM was biotinylated using EZ-linkTM, su lfo-NHS-LC-LC-biotin (Pierce, Rockford, IL cat.21338) by following the manual. Calmodulin Overlay Experiments The purified CRMP-2 and calcineur in protein was subjected to SDS-PAGE. Proteins were separated by SDS-PAGE electrophoresis and el ctrtrotransferred to a PVDF membrane. The PVDF membranes were incubated with 5% (W /V) bovine serum albumin in TBST medium containing either 1mM CaCl2 or 1mM EDTA to prevent non specific binding. Biotinlyated bovine or human brain CaM was next added (20 ng/ml) and incubated for another hour. The membrane were then extensively washed in TBST in the presence of 1mM CaCl2 or EDTA and further incubated for 1 h with avidin-conjugated alkaline phosphatase in the presence or absence of Ca2+. Again, Blots were extensively washed, a nd the positive CaMBP bands were detected using nitroblue tetrazolium and 5-br omo-4chloro-3-indolyl phosphate.
83 In Vitro Calpain-2 and Caspase-3 Digestion of Cell Lysate or Purified hCRMP-2 PC-12 cell lysate was prepared as above. In vitro digestion of rat lysate (5 mg) or purified CRMP2 protein with the purified proteases, hum an erythrocyte caplain-1, rat recombinant calpain-2 (Calbiochem, San Diego, CA) a nd recombinant human caspase-3 (Chemicon, Temecula, CA) were performed in a buffer c ontaining 100 mM Tris-HCl (pH 7.4) and 20 mM DTT. For calpain-1 or -2, 1 mM CaCl2 or 5 mM CaCl2 was added respectively with or without 50 M W7, and then incubated at room temperatur e for 30 minutes. For caspase-3, samples were incubated at 37C for 4 hours. The protease reaction was stopped by the addition of 30 M calpain inhibitor (SJA 1670) or 100 M pan-caspase inhibitor (Z -D-DCB) (Calbiochem, San Diego, CA) and a protease inhib itor cocktail solu tion (Roche Biochemicals, Indianapolis, IN). Neurotoxin Challenges and Ph armacologic Intervention In addition to untreated controls, the follo wing conditions were used: maitotoxin (MTX, 3 nM; WAKO Chemical, USA Inc) as a calpain -dominated challenge for 3 hour, NMDA (200 M; Sigma) for 8-24 hours as an excitotoxin challe nge. For pharmacologic in tervention, cultures will be pretreated 1 hour before MTX and NMDA cha llenge with calpa in, caspase-3 inhibitors. Immunocytochemistry Primary cortical neurons were fixed with 4% paraformaldehyde (PFA) in PBS for 10 minutes, washed with PBS and permeabilized with 0.1% triton X-100 in PBS for 5 minutes. CRMP-2 staining was performed fo llowing a one-hour blocking step in 10% goat serum at room temperature. Then the neurons were incubated overnight at 4C with monoclonal mouse-antiCRMP-2 (C4G, IBL, Aramachi, Takasaki-shi, Gu nma, Japan) at a d ilution of 1:500. Alexa 488conjugated goat-anti-mouse second ary antibody (Molecular Probes, Eugene, OR) was added at a dilution of 1:1000, followed by washing with PBS. F-actin was fluorescently labeled with 5 units/ml Rhodamine-conjugated phallo idin for 30 minutes at room temperature. The cells were
84 mounted using medium containing 4, 6-diamidin e-2-phenylindole (DAPI) (Vector Laboratories, Burlingame, CA). Fluorescence images were captu red with a 20x objective on the Zeiss Axiovert 200 Fluorescence Microscope with a CCD camera, and combined using SPOT imaging software (Diagnostic Instruments, Sterling Heights, MI). Results Identification of CRMP-2 as a Calmodulin-binding Protein In our previous studies, we developed a novel CaM-affinity capture coupled with RPLC-MSMS method to enrich and identify pr oteins or peptides which have CaM binding domains. Data suggested that CRMP-2 might be a putative nove l CaM binding protein. Ca M binding proteome of rat brain was visualized by Coomassie blue staining (Figure 4-1). CRMP-2 was marked by red rectangle red rectangle. Our ma ss spectrometry results showed a total of 4 peptides of CRMP-2 that were sequenced by RPLC-MSMS. By compar ing those peptides to full length of rat CRMP2, they corresponded to 238-255, 258-270,440-45 2, and 470-481 residues, respectively (underline in Figure 4-1 ). The overall sequence coverage of CRMP-2 identified by RPLCMSMS is 9.6%. Moreover, the peptides 258-270 and 470-481 were found twice. Two putative calmodulin binding domains (CaMBDs) were found by searching CaMBD search program (http://calcium.uhnres.utoronto.c a/ctdb/ctdb/sequence.html )(Yap et al., 2000). One is located between residues 258-275, and the other between residues 475-493. Helical wheel projections of these two putative CaMBDs were shown in Figure 4-1B CaMBD-2 shows an appropriate amphiphilic helix distribution with predominan tly hydrophobic face on one side and a positively charged on the other side. Studies have shown that CRMP-2 can be phos phorylated at Thr514 or Thr509, Ser518 and Ser522 by GSK3 and Cdk5 or Thr-555 by Rho-kinase (G u et al., 2000; Cole et al., 2004; Yoshimura et al., 2005). Additionally, recent stud ies have shown that the most likely calpain
85 cleavage site is between residues 486 and 551. The schematic diagram of CRMP-2 ( Figure 41C ) indicates that CaMBD-2 overlaps the c-te rminal C2 domain (residues 486-533) which is very close to the calpain cleavage sites and mu ltiple phosphorylation sites. Thus, it suggests that CaM binding might affect proteolysis and/or phosphorylation of CRMP-2. To confirm that CRMP-2 can be purified by CaM-agarose, we probe d the CaM-affinity EDTA elution and Ca2+ flow through with CRMP-2 antibody. Figure 4-2A shows that CMRP-2 can be pulled down with Ca2+ chelator EDTA. However, the detection of CRMP-2 in the CaMagarose pull down elution with EDTA does not exclude the po ssibility that CRMP-2 could be isolated because of indirect association with other CaM-binding proteins Hence, we performed biotinylated CaM overlay experiment with purif ied GST-CRMP-2 protein in either the presence or absence of Ca2+. In the presence of Ca2+, a strong band was observed and was accompanied by positive control calcineurin after development with streptavidian-peroxidase. However, in the absent of Ca2+, no band observed ( Figure 4-2B ). Therefore, it suggests that CaM directly and specifically binds to CRMP-2 in a Ca2+-dependent manner. CaM Regulates CRMP-2 Mediated Filopodia Formation The dynamics of filopodia is necessary for unde rstanding the growth cone formation, thus we investigated the signaling pathways that might involved in cytoskeleton change. Here, rhodamine-phalloidin F-actin immunostaining and monoclonal CRMP-2 immunostaining (green) were performed with DAPI counter staining (blue) of the nuclear DNA. Consistent with results from other groups, CRMP-2 was found to colo calize with F-actin in PC-12 cells ( Figure 4-3 ). Interestingly, before treatment with the CaM antagonist W7, CMRP-2 overlapped with F-actin, and found to uniform lie in parallel with each ot her within the cell body. Af ter thirty minutes of treatment, organized F-actin structure was lost as was the cytoskeleton structure ( Figure 4-3B ). We were intrigued by the fact that CRMP-2 might interact w ith F-actin and that CaM was
86 involved in F-actin dynamics. Therefore, we suspecte that CaM may modulate CRMP-2 mediated filopodia formation. It is well know n that over-expressing CRMP-2 in hippocampal neurons induces axon formation (Inagaki et al., 2 001). Our data suggests that the over-expression of GFP-CRMP-2 in HEK293 cell induces longe r filopodia formation when compared to overexpression of GFP alone. After thirty minut es CaM antagonist W7 challenge, the long filopodia retracted. In contrast, there is no ch ange in the GFP-over-exp ressing HEK293 cells filopodia. The change of filopodia is even more dramatic in cells tr eated with LiCl, a GSK3 inhibitor. GSK3 is implicated in axonal/neurite formation and by inhibiting GSK3 leads to dephosphorylation of CRMP-2, thus resul ting in axon elongation. When CRMP-2 overexpressing HEK293 cells are incubated with LiCl for one hour, more cells were found bearing longer filopodia. Application of W7 dramatically reduces the numbers of cells with longer filopodia ( Figure 4-4 ). Taken together, our data suggests that CaM may be involved in CRMP-2 mediated filopodia formation. CaM Binding Retards Calpain-mediat ed CRMP-2 Proteolysis in Vitro Our previous studies suggested that calpai n mediates CRMP-2 proteolysis (Zhang et al., 2007). In order to assess the effect of CaM binding on CRMP-2 proteolysis by calpain in vitro we compared the cleavage efficiency of the CR MP-2 with or without CaM by calpain in vitro. Purified GST-CRMP-2 protein was tr eated with calpain-1 or calpain -2 in either the presence or absence of CaM. The addition of CaM to GST-CRMP-2 signifi cantly retards CRMP-2 pr oteolysis as noted by the decrease in the level of the CRMP-2 breakdown product (CRMP-2 BDP) in both calpain1 and calpain-2 treated panels ( Figure 4-5A ). To further examine the effect of endogenous CaM binding on proteolysis of CRMP-2, we treated PC12 cell lysates with calpain-1 or calpain-2 after applying CaM antagonist W7. The W7 trea ted panel exhibits higher level of CRMP-2 BDP
87 compared to the untreated cells (Figure 4-5B) Thus, this indicates that CaM inhibition enhances CRMP-2 proteolysis by calpain-1 or 2 in vitro Altogether, these results demonstrate that CaM binding prevent calpain-mediated CRMP-2 proteolysis in vitro Post-translation Modification of CRMP-2 Following Ca2+/CaM Binding ex vivo As we have shown that CaM binding reta rds calpain-mediated CRMP-2 proteolysis in vitro. N ext we examined whether Ca2+/CaM modulates CRMP-2 proteolysis ex vivo CRMP-2 BDP was used as a marker for calpain activity after MTX treatment. We treated GFP and GFPCRMP-2 over-expressing HEK293 cells with di fferent concentrati ons of W7, from 5 M to 100 M, for thirty minutes. Only one band was visi ble on the immunoblot representing the intact CRMP-2 meaning that the protein was not cleav ed into the 55 kDa CRMP-2 BDP (Data not shown). However, when we added W7 to the Ca2+ channel opener MTX treat ed cells, a robust 55 kDa was clearly visible more than MTX treated only cells ( Figure 4-5C ). This indicates that the application of W7 application acts synergistically with MTX to cleave CRMP-2. Therefore, CaM may be involved in the regulation of CRMP-2 proteolysis after intracellular Ca2+ elevation ex vivo It has been reported that two isoforms of CRMP-2 are expressed in primary cortical neurons. Consistent with Bretin s mice data (Bretin et al., 2005), our previous and ongoing studies suggest that CRMP-2A a nd 2B isoforms do exist in rat primary cortical neurons and human SY5Y cell line as well, but not in cerebe llar granule neurons a nd PC-12 cells (Data not shown). Since the antibody C4G can recogni ze both phosphorylated and non-phosphorylated CRMP-2 isoforms in primary cortical neurons and since phosphorylation dynamics of CRMP-2 is more important in determinating axon elonga tion, we examined the effect of CaM binding on CRMP-2 phosphorylation dynamics. We found that elevated intracellular Ca2+ concentration due to MTX or NMDA treatment causes CRMP-2 to undergo dephosphorylation and proteolysis
88 even in the presence of hi gh level of endogenous CaM. Alt hough the calpain inhibitor SJA6017 can protect CRMP-2 from proteolysis, it can not prevent CRMP-2 dephosphorylation in the pathology state ( Figure 4-6). Next, we studied the effect of CaM on phosphorylation dynamics of CRMP-2. A GSK3 inhibitor, LiCl, was used to induce CRMP-2 de phosphorylation in primary cortical neurons. As shown in Figure 4-7A there was a significant decrease of the phosphor-CRMP-2B after LiCl treatment, while no change was observed following co-treatment with W7. Quite remarkably, the protein phosphotase PP1 and PP2A inhibitor okadaic acid significantly en hanced the level of phosphor-form of both CRMP-2A and B, while PP2B inhibitor cybermethrin did not (Figure 47C) CaM antagonist W7 produces no dramatic effect on the phosphorylation dynamics of CRMP-2 induced by okadaic acid and cybermethri n. Thus, this indicates that phosphotases PP1 and PP2A were responsible for the dephosphorylation of CRMP-2 in vivo and CaM does not affect the phosphorylation dynamics. Discussion The RPLC-MSMS based CaM-pull down proteomi c data and the CaM overlay experiment with recombinant CRMP-2 indicated that CRMP-2 binds to CaM in a Ca2+-dependent manner (Figure 4-1) There are two CaMBDs for CRMP-2 as predicated by the CaMBD predication program. CaMBD2 (residues 475-493) has been s hown to form a helix in a recent crystal structure study of CRMP-2 (Stenmark et al., 2007). Furthermore, this region is rich in basic and hydrophobic residues. The helical wheel projection of CaMBD2 shows this region could form an appropriate amphiphilic helix with characte ristic hydrophobic amino acid residues, namely Phe475, Ala 484, Leu488 and Leu491, on one face and on the other face with a cluster of charged amino acid residues. This is in agreemen t with the results of pr evious studies showing that CaM binding sequence potentially forms a basic amphiphilic a-helix (Wang, 2000b). Thus,
89 CaM more likely binds to CRMP-2 through the CaMBD2. A further deletion mutation study or synthesized peptides of CRMP-2 is needed to further identify the exact CaMBD. Next, the schematic diagram of CRMP-2 indi cates that CaMBD-2 overlaps its c-terminal C2 domain which is very close to the calpain cleavage sites and the multiple phosphorylation sites. Thus, this suggests that CaM binding mi ght affect proteolysi s or phosphorylation of CRMP-2. In agreement with previous studies, different phosphor-forms of CRMP-2 exist in cortical neurons in its ph ysiological state (Figure 4-6 to 4-7 ). Intact CRMP-2 may bind to F-actin constitutively to stabilize axon and growth cone structure. However, under pathological condition with the elevated cytosolic Ca2+, CRMP-2 may be undergoing different posttranslational modification in response to the stimulus with or w ithout CaM regulation ( Figure 45 to 4-8 ). Initially, activ ation of calpains and protein phos phatases PP1 and PP2A followed by Ca2+ influx results in CaM indepe ndent CRMP-2 dephosphorylation ( Figure 4-5 to 4-7 ). With the sustained Ca2+ influx and calpains activation, CRMP-2 undergoes CaM independent proteolysis, which may contribute to gr owth cone collapse and axonal injury ( Figure 4-6 ). Concurrently, neurons need to activate profound in trinsic repair signaling pathways to prevent axonal/neuronal injury and boost axon al recovery in response to Ca2+ influx. In our proposed model, with the Ca2+ influx, CaM undergoes conformational change, binds to CRMP-2 and protects CRMP-2 from calpain-mediated cleavage ( Figure 4-5 ). In addition, CaM binding to CRMP-2 appears to play a role in F-actin bundling ( Figure 4-3 ), thereby maintaining a stable structure or enhanci ng axonal regeneration ( Figure 4-4 ). It is well established that actual extension of a xons occurs at its distal tip, the growth cone. The peripheral and central domains of the growth cone are composed of by two F-actin based structures, filopodia and lamellipodia. These actin -rich structures contribute to the force
90 necessary for the forward extension of the growth cone. It is ther efore important to understand of how actin dynamics are regulated and this could provide key insi ghts into how axonal regeneration may be promoted (Dent and Gertle r, 2003). There have not been any systematic studies of the interaction CRMP-2 with actin, however, there is some evidence suggesting that CRMP-2 may interact with F-actin (Gungabi ssoon and Bamburg, 2003; Arimura et al., 2005; Kawano et al., 2005). CRMP-2 merged with actin at growth cone. Our data suggests that CRMP2 co-localizes with F-actin and that the CaM antagonist W7 disput es this F-actin based structure in PC-12 cells. Furthermore, W7 reduced the number and length of filopodia induced by CRMP2 over-expression. Therefore, our study of Ca M affects on filopodia formation may have application on the growth cone dynamics of ax on regeneration. Further studies of how CaM modulates CRMP-2 involvement in the actin dynam ics in the growth cone of neuron are needed. In our studies, this is th e first time that CaM was found to protect CRMP-2 from the cleavage by calpain. We in terpret this effect as a conseque nce of direct CaM binding to the corresponding domain of the CRMP-2 CaMBD2 (residues 475-493), which is adjacent to the cleavage site. As a Ca2+ sensor, CaM binds to numerous di fferent CaMBPs resulting in the specific intracellular response to the Ca2+ signal. For example, CaM binding to the plasma membrane Ca2+ ATPase, GAP-43 leads to inhibition of calpain-mediated cleavage at several sites located within the CaMBD. In other cases, CaM either accelerated the proteolysis (brain spectrin, calponin and calcineurin) by calpain or changed the pattern of cleavage (myosin, myosin light chain kinase) (Kosaki et al., 1983 ; Harris et al., 1989; Ts unekawa et al., 1989; Croall et al., 1996). Furthermore, studies have demonstrated that CaM binding to N-MethylAspartate Receptor (NMDA) subunit 1 facilitat es calcium-dependent inactivation of NMDA Receptor, which in turn serves as a negative feedb ack to fine tune Ca2+ influx after injury(Zhang
91 et al., 1998; van Dalen et al., 2003). Previous studie s have also showen th at the proteolysis of CRMP-2 also occurs in ischemia and TBI (Zhang et al., 2007). Similarly, it is possible that CaM binding to CRMP-2 also in turn prevents its degrad ation and stabilizes grow th cone structure in a negative feedback loop to fine tune the response to injury in vivo Phosphorylation of CRMP-2 is important in determining the function of CRMP-2 in axon elongation and growth cone collapse (Arimura et al., 2005; Yoshimura et al., 2005). Gu et al suggested that PP1 and PP2A may be invol ved in the dephosphorylation of CRMP-2 in vitro (Gu et al., 2000). A recent proteomic study also found that CRMP-2 is phosphorlated upon okadaic acid treatment (Hill et al., 2006). Given the f act that NMDA or MTX induction induces CRMP-2 dephosphorylation and PP1 and PP2 inhibitor enha nced CRMP-2 when co-treated with calpain inhibitor, we proposed PP1 and PP2A were re sponsible for CRMP-2 dephosphorylation both in vitro and in vivo. Different from other cytosk eletonal proteins, such as NF, tau and MAP2 (Yamamoto et al., 1985; Hashimoto et al., 2000 ),our data shows that CaM binding does not influence the phosphorylation of CRMP-2. This ma y be due to high concentration of endogenous CaM or highly dynamics of temporal and spatial phosphorylation modifications that occur in vivo In summary, our study indicat es that CRMP-2 is a new cal modulin binding protein and CaM binding may retard proteo lysis of CRMP-2, but not phos phorylation, thereby modulating CRMP-2 involvement with F-actin dynamics in response to certain pathological states. Understanding the signaling mechanisms of Ca2+/CaM and CRMP-2 in cytoskeleton dynamics may open up the possibility of developing novel stra tegies to alleviate cell death and promote axon regeneration in vivo
92 Figure 4-1. CRMP-2 is a putative CaM binding pr otein. (A) Identification of CRMP-2 as a putative calmodulin binding protein by mass spectrometry. The rat brain CaM binding proteome was visualized by Coomassi e blue staining. Each visible band was excised, in-gel digested and analyzed by RPLC-MSMS. CRMP-2 band was boxed as indicated. (B) Putative CaM-binding do mains (CaMBD) of CRMP-2 analysis. Putative CaMBDs were searched by using CaMBD search program (http://calcium.uhnres.utoronto.ca/ctdb/ct db/sequence.html). Helical wheel projections of amino acids from putativ e CaMBDs were present as B1 and B2 (http://cti.itc.virginia.edu/~cmg/Demo/w heel/wheelApp.html). Positively charged residues are marked with a plus sign, wh ereas hydrophobic residues are marked as yellow. (C) Schematic diagram of locali zation of potential CaMBDs in CRMP-2.
93 Figure 4-2. CRMP-2 binds to CaM in a direct and specific Ca2+-dependent manner. Immuoblot of CRMP-2 as a putative CaMBP. Lysate before CaM-agaros e purification, CaMagarose-elution and flow through were separated by SDS-PAGE and analyzed by immunoblot for CRMP-2. (B) The purif ied GST-CRMP-2 and bovine brain calcineurin were subjected to overlay with 20 ng/ml biotinylated bovine CaM in the presence of 2 mM CaCl2 or 2 mM EDTA. The PVDF membranes were further incubated for 1 h with avidin-conjugated alkaline phosphatase in the presence or absence of Ca2+ and detected using nitroblue te trazolium and 5-bromo-4chloro-3indolyl phosphate.
94 Figure 4-3. CRMP-2 co-localizes with F-actin and CaM antagonist W7 altered F-actin organization. PC-12 cells were culture in the presence of DMSO (upper panel) or 10 M of W7 (lower panel). PC-12 cells were doubly staining with anti-CRMP-2 (green) and rhodamine F-actin (red). Yellow colo r in merge photographs corresponds to colocalization of CRMP-2 with F-actin. Bl ue is DAPI stained nuclei. Scale bar=10 M
95 Figure 4-4. CaM antagonist W7 induces filopod ia retraction in CRMP-2 overexpressing HEK293 cells. HEK293 cells with over-expressed G FP (A) or GFP-CRMP-2 (B) were treated with W7, LiCl or both. F-actin was staine d by rhodamine-phalloidin. Yellow color in merged photographs corresponds to colocaliz ation of CRMP-2 with F-actin. Typical Changes of filopodia are indi cated by arrows. Blue is DAPI stained nuclei. Scale bar=10 M. (C) Quantification of cells with at least one fil opodia greater than half the diameter of the cell body following various treatments. For each experiment, at least 30 cells were randomly measured.
96 Figure 4-5. CaM modulates calpai n-mediated CRMP-2 proteolysis in vitro (A) Calpain mediated purified CRMP-2 prot eolysis is retarded by CaM in vitro Purified GSTCRMP-2 protein (1 g) was incubated with calpain-1/ 2 in the presence of calmodulin (10 M) or Ca2+ (2 M). (B) Calpain mediated CRM P-2 proteolysis in cell lysate is enhanced by CaM antagonist in vitro Whole PC-12 cell lysates (20 g) were incubated with calpain-1/2 in the presence of CaM antagonist W7 (50 M) or SJA (30 M) (C) W7 application acts synergis tically with MTX to induce CRMP-2 proteolysis. Therefore, CaM might regul ate CRMP-2 proteolysis after intracellular Ca2+ elevation. HEK293 cells with ove r-expressed GFP or GFP-CRMP-2 were treated with MTX with or without W7 or cal pain inhibitor SJA. Total proteins in the cell lysates were resolved by SDS-PAGE and immunoblotted with anti-CRMP-2 (C4G), aII-spectrin and actin.
97 Figure 4-6. Elevation of intracellular Ca2+ results in CRMP-2 proteo lysis and dephosphorylation. Primary cortical neurons were pretreated with or without calpain inhibitor (30 M SJA6017) or caspase inhibitor Z-VAD fo r one hour, and then exposed to Ca2+ channel opener maitotoxin (MTX 3 nM) for 3 hours, and NMDA (200 M) for 12 hours. Total proteins in the cell lysa tes were resolved by SDS-PAGE and immunoblotted with anti-CRMP-2 (C4G).
98 Figure 4-7. Okadaic acid enhances CMRP-2 phosphorylation but CaM does not affect the dynamics of CRMP-2 phosphorylation. Primary co rtical neurons were pretreated with calmodulin antagonist (W7, 10 M) for thirty minutes, and then exposed to LiCl (20mM), okadaic acid (50 nM) and cybermet hrin (4 M) for one hour. Total proteins in the cell lysates were resolved by SD S-PAGE and immunoblotted with anti-CRMP2 (C4G) (A). -actin (B) was used as a loading cont rol. (C) Densitometric analyses of CRMP-2 and phospho-CRMP-2 immunoblots were performed. Values were means S.E.M. n = 4. Statistical significance of differences between the control and each treated group was determined by one-way ANOVA with Dunnett's multiple comparison tests. A difference was considered to be statistically significant when the P value was less than 0.05 (* P < 0.05).
99 Figure 4-8. Proposed model of the role of calmo dulin in CRMP-2 post-translation modification under pathological conditions. CRMP-2 is phosp horylated when in physiological state (Figures 4-6 and 4-7). Under pathological states with the elevated cytosolic Ca2+, activation of calpain and protein phosphatases PP1 and PP2A induces CaM independent CRMP-2 dephosphorylation and proteolysis (Figure 4-7). With the sustained Ca2+ influx and calpain activation, CRMP-2 undergoes CaM independent proteolysis, which may contribute to grow th cone collapse and axonal injury (Figure 4-6). Meanwhile, in our model, CaM undergoes conformational changes upon Ca2+ influx, binds to CRMP-2 and prevents CRM P-2 from calpain-mediated cleavage, thereby serving as a negative feedback as part of the cells re sponse after injury (Figure 4-5). In addition, Ca M binding to CRMP-2 might play a role in F-actin bundling (Figure 4-3) and, in turn, maintain s stable structure or enhancing axonal regeneration (Figure 4-4).
100 CHAPTER 5 DIRECT RHO-ASSOCIATED KINASE INHIBITION INDUCES COFILIN DEPHOSPHORYLATION AND NEURI TE OUTGROWTH IN PC-12 CELLS Introduction Unlike in the peripheral nervous system, axons in the adult central nervous system (CNS) undergo little spontaneous regeneration. The lack of regeneration is due to a diverse class of neuritogenic inhibitors that prev ail in the CNS. Some of these inhibitors have already been identified, such as Nogo-A, myelin associated glycoprotein (MAG) (McKerracher et al., 1994), chondroitin sulfate proteoglycans (CSPGs) and oligodendrocyte myelin glycoprotein (OMgp) (Wang et al., 2002; McKerracher and David, 2004). Approaches targeting these molecules could promote axonal regeneration. A number of studies have demonstrated that inhibition of small Rho-GTPases and Rho kinase can promote axona l regeneration by overcom ing the inhibitory effects of myelin, as well as Nogo-A, CSPG, and MAG (Lehmann et al., 1999; Yamashita et al., 2002; Fournier et al., 2003; Yamash ita et al., 2005). In contrast, ne urotrophic factors promote the growth and survival of neurons in the mature ner vous system. They also pl ay a significant role in influencing nerve fiber growth. In addition to nerve growth fact or (NGF) (Bonini et al., 2003), several other molecules exhibit neurotrophic properties, includi ng fibroblast growth factor (FGF), brain-derived neurotroph ic factor (BDNF) and neurotrophi n-3 (NT-3) (Ebadi et al., 1997; Petruska and Mendell, 2004). Also reported, NGF and NT-3 promote axonal outgrowth via the suppression of Rho-A activity (Ozdinler and Erzurumlu, 2001; Yamaguchi et al., 2001). Accumulating evidence has linked the Rho-A family with permissive as well as inhibitory pathways of neurite outgrowth. ROCKs (also known as Rho kinases), a class of serine/threonine kinases, were the first downstream effectors of Rho to be discovered (Kwon et al., 2002) Numerous studies showed the ROCKs mediate a large proportion of the signals from Rho in regulating dynamic reorganization
101 of the cytoskeleton (Amano et al., 2000; Hall, 2005). Initially, activ ation of ROCKs was characterized by direct phosphor ylation of myosin light ch ain (MLC) (Somlyo and Somlyo, 2000) and by indirect inhibi tion of MLC phosphatase (MLCP) (Kawano et al., 1999) in mediating Rho-A induced stress fi bers and focal adhesions. ROCKs consist of an amino-terminal kinase domain and an autoinhi bitory carboxy-terminal region, which includes the Rho-binding (RB) domain and the pleckstrin homology (P H) domain. Both the RB and PH domains can interact independently from the amino-terminal ki nase domain and in turn inactivate the enzyme. Activated Rho interacts with the RB domain, a nd disrupts the negative regulated interaction between the kinase domain and the autoinhibito ry region, thereby freei ng kinase activity. Two ROCK isoforms have been identi fied: ROCK I (also known as ROK and P160ROCK) and ROCK II (ROK ). The kinase domains of ROCK I and ROCK II are 92% iden tical, and so far there is no evidence that they have diffe rent functions (Rient o and Ridley, 2003). Recently, the Rho-ROCK pathway has been dem onstrated to play a prominent role in mediating neurite retraction, grow th cone collapse and axonal growth through ROCK inhibition on dorsal root ganglion (DRG) on an inhibitory substrate (suc h as MAG) (Amano et al., 2000; Fournier et al., 2003). Strategies for prom oting axonal regeneration in the CNS are therapeutically attractive for treatment of various diseases such as traumatic brain injury, stroke and Alzheimers disease. The rat pheochromocyto ma cell line, PC-12, has been widely used as an important model for neuronal differentiation. PC-12 cells differentiate into a neuronal phenotype in response to various ne urotrophins. For instance, nerv e growth factor (NGF) treated PC-12 cells exhibit proliferati on arrest, neurite outgrowth (NOG) and electrical excitability (Greene and Tischler, 1976). In previous studi es we also demonstrated that repeated amphetamine treatment induces NOG in PC-12 cells, similar to that found with known
102 neurotrophic factors (Park et al., 2003). Moreover, PC-12 cells elicit NOG via Rho inhibition by Clostridium b. C-3 exoenzyme treatment of PC -12 cells (Lehmann et al., 1999; Sebok et al., 1999) or DRG on inhibitory subs trate (Fournier et al., 2003), making it an invaluable model system for studying potential Rho-ROCK downst ream signal transduction pathways in NOG. Cytoskeletal reorganization plays a striking role in NOG, such as through actin and microtubule remodeling. Thus, it is essential to understand the signaling pa thways that control cytoskeleton dynamics (Tojima and Ito, 2004). Rho GTPase has been shown to influence actin cytoskeleton dynamics (Sebok et al., 1999; De nt and Gertler, 2003). ROCKs, also mediate signals to the actin cytoskelet on through various substrates, such as adducin and LIM kinase (LIMK) (Dent and Gertler, 2003). In turn, LI MK phosphorylates cofilin, an actin associated protein, which binds to actin and serves to enhance depolymerizat ion of actin filaments. Once phosphorylated, cofilin is inactivated and lose s its filament severing and monomer binding abilities (Maekawa et al., 1999; Ohashi et al., 2000). In addition, ROCKs phosphorylate other neurite intermediate filament or microtubule-associated proteins su ch as NF-L (Hashimoto et al., 1998), Tau and MAP2 (Amano et al., 2003). A number of ROCK inhibitor compounds have been developed, including H-89, HA-1077, Y-27632 (Davies et al., 2000), H-1152 (Ikenoya et al., 2002) and Wf-536 (Nakajima et al., 2003). Among commercially available inhibitors (R)-(+)-trans-N-(4-Pyr idyl)-4-(1-aminoethyl)cyclohexanecarboxamide (Y-27632), have shown high potency and selectivity for ROCK inhibition (Amano et al., 2000; Ishizaki et al., 2 000). This selective inhi bition of ROCKs makes Y-27632 very useful for evaluating ROCK functions in CNS since ROCKs are highly expressed in the brain.
103 In this study, the role of ROCK inhibitor (Y-27632 or H-1152) was systematically evaluated for promoting NOG in the well-define d PC-12 model. Dynamics in cell morphology and cytoskeleton components (actin, cofilin and III-tubulin) were characterized following treatment with ROCK inhibitor Y-27632. Results suggest ROCK inhibition might be a potential therapeutic avenue for promoting NOG after CNS injury. Materials and Methods Chemicals and Antibodies Culture media and sera were obtained from Gibco Inc. (Rockville, MD). Y-27632, H-1152, Ro-32-0432, PD 98059, and H-89 were purchased fr om Calbiochem (San Diego, CA). The primary antibodies used include : polyclonal antibodies against phospho-cofilin and cofilin (Cell Signaling, Beverly, MA), monoclonal antibody against III-tubulin (Covance, Denver, PA) and FITC-conjugated phalloidin (Mol ecular Probes, Eugene, OR). Cell Culture PC-12 cells were maintained at 37C in Dulbecos modified Eagles medium (DMEM) supplied with 10% fetal bovine serum (FBS), 5% heat-inactivat ed horse serum, 100 g/ml of streptomycin, 100 U/ml of penicil lin and 1% Fungizone (Gibco, Ro ckville, MD) in a humidified 5% CO2 incubator. To induce neurite outgrowth, PC-12 cells were plated in the same medium with 25 M Y-27632. Y-27632 was prepared as a 25 mM stock solution in dimethlsulfoxide (DMSO) and added directly into the medium. An equal amount of DMSO was added to control plates. Micrographs of cells were taken at 32x with an Axiocam digital camera using a Zeiss Axiovert 135 microscope. For the kinase inhi bition study, PC-12 cells were treated with 25 M Y-27632 for 2 hours, and then continued to be co-treated with 300 nM RO-32-0432 or 500 nM H-89 or 30 M PD 98059 for 14 hours.
104 Quantification of Neurite Outgrowth Cell processes were defined as neurites when longer than the diam eter of the cell body. The percentage of neurite-bearing cells was calcul ated as the number of cells with one or more neurites divided by the total cell number (Park et al., 2003). Neurite length was evaluated by manually tracing the longest ne urite per cell us ing the software ImageJ (version 1.29x, NIH, USA) and referenced to a known length. Each ex periment was conducted in triplicate. Images were taken with 15 or more cells per field. For each experiment, at least 50 cells were randomly measured. Immunoblotting PC-12 cells were treated for various time pe riods, washed twice with phosphate-buffered saline (PBS), and solubilized with lysis bu ffer containing 20 mM Tr is-HCl, pH 7.4, 150 mM NaCl, 5 mM EDTA, 5 mM EGTA, 1% Triton X-100, 1 mM NaF, 1 mM Na3VO4, and a protease inhibitor cocktail tablet (Roche, Indianapolis, IN). The cell lysates were briefly sonicated before clarificati on by centrifugation at 15,000 g for 10 minutes at 4C. Protein concentration of the supernatant was determ ined by a modified Lowry method (DC Protein Assay Kit, Bio-Rad, Herc ules, CA). Samples (20 g protein) were reso lved by 10-20% sodium dodecyl sulfate-polyac rylamide gel electrophoresis (S DS-PAGE) and transferred onto a polyvinylidene difluoride (PVDF) membrane by th e semi-dry method. Membranes were blocked with 5% non-fat milk in tris-buffered sa line containing 0.1% tween-20 (TBST) and then incubated with the primary antibody in 5% non-fat milk in TBST at 4C overnight. Following washes with TBST, membranes were incubate d for one hour at room temperature with a biotinylated secondary antibody. Following another series of washes, the membrane was incubated with avidin-conjugated alkaline phosphata se for 30 minutes. Prot eins were visualized using nitroblue tetrazolium and 5-bromo-4chloro-3-indolyl ph osphate. The membranes were
105 scanned, and the optical density of the bands wa s quantified with the software ImageJ (version 1.29x, NIH, USA). Immunocytochemistry PC-12 cells were seeded onto LabTek II cham ber slides (Nunc, Naperville, IL) followed by overnight incubation. On the next day, the medium was replaced with or without 25 M Y27632. Twenty-four hours followi ng treatment, PC-12 cells were fixed with 4% paraformaldehyde (PFA) in PBS for 10 minutes, washed with PBS and permeabilized with 0.1% triton X-100 in PBS for 5 minutes. F-actin wa s fluorescently labeled with 5 units/ml FITCconjugated phalloidin for 20 minutes at room temperature. III-tubulin staining was performed, following a one-hour blocking step in 10% goat seru m at room temperature. Then the cells were incubated overnight at 4C with monoclonal rabbit antiIII-tubulin at a dilution of 1:2000. Alexa 488-conjugated goat-anti-rabbit secondary antibody (Molecular Probes, Eugene, OR) was added at a dilution of 1:1000, followed by washi ng with PBS. The cells were mounted using medium with 4, 6-diamidine-2phenylindole (DAPI) (Vector Labor atories, Burlingame, CA). Fluorescence images were captured with a 20x ob jective on the Zeiss Axioplan 2 Fluorescence Microscope with a CCD camera and combined using SPOT imaging software (Diagnostic Instruments, Sterling Heights, MI). Results Dose-Dependent Neurite Outgrowth Induced by ROCK Inhibition in PC-12 Cells The effect of ROCK inhibitor Y-27632 on a low-density culture of PC-12 cells over multiple concentrations is shown in Figure 1. At the low dose of 0.01 M PC-12 cells appear similar with control cells that are slightly larger and rounde r, but possess few visible neurites (Fig. 1 A, top panel). At higher Y-27632 doses, from 1 M to 100 M, an increasing number of PC-12 cells present multiple long br anched neurites, compared to sparse growth in controls (Fig.
106 1 B-F, top panel). The qualified data shows that the percentage of neurite-bearing cells significantly increased in a dose-dependent manner, peaking between 25 and 100 M with greater than 90% (Fig.1, bottom pa nel). However, at 100 M, cell de tachment from the substrate and neurite loss was noticed after 48 hours of exposure (data not shown). Dynamics of Neurite Outgrowth in ROCK Inhibitor Treated PC-12 Cells To further examine NOG dynamics after RO CK inhibition, we executed a time course analysis of NOG in PC-12 cells using 25 M Y-27632. NOG was immediate initiated within 5 minutes, with small protrusion veils (lamellipodia) developing (Fig. 2B), and 10 minutes later with a few spikes (filopodia) forming (Fig. 2C ). Prominent neurite elongation was visualized between 30 minutes and 10 hours. Robust and fully extended neurites we re observed following 6 to 10 hours of treatment, which was sustained for 24 hours after Y-27632 stimulation (Fig. 2D-I). There was a rapid increase in neurite-bearing ce lls after Y-27632 exposure, with greater than 85% of the cells showing neurites within 6 hours (Fig. 2J), The lengtheni ng of neurites appears biphasic, with a rapid increase within 6 hours, an d a more gradual extension to maximal length by 24 hours (Fig. 2K).Since protein kinase i nhibitory agents, including Y-27632, can crossinhibit other protein kinase, we further tested another more specific ROCK inhibitor, H-1152 (Fig. 3). We observed that even at a very low concentration (1 M), H-1152 produced very rapid cell shape changes, lamellipodia and filopodia developing within 5-30 minutes (Fig. 3B, C), followed by fully extended neurites by 24 hours (Fig. 3F). We also found that 10 M of H-1152 produced the same NOG effect (Fig. 3H), wh ile 0.1 M also produced partial NOG by 24 hours (Fig. 3G).
107 Remodeling of Cytoskeletal Architecture in ROCK Inhibition Mediated Neurite Outgrowth With the cytoskeleton considered a prin cipal determinant of NOG, we examined cytoskeleton reorganization dur ing Y-27632 treatment (Tojima and Ito, 2004). FITC-phalloidin F-actin immunostaining and neuronal specific III-tubulin immunostain ing (green) were performed against DAPI staining (blue) of the nucl ear DNA, respectively. Before treatment, PC12 cells were round and F-actin was uniformly locali zed in the periphery of the soma (triangle in Fig. 4A), whereas III-tubulin was evenly di stributed within the cel l body (Fig. 4C). After 12 hours treatment, F-actin accumulated in growth c one-like structures (short arrow in Fig. 4B), newly formed neurites (long arrow in Fig. 4B) and lamellipodia (triangle in Fig. 4B). In comparison with F-actin, III-tubulin was preferentia lly concentrated along the nascent neurites (arrow in the Fig. 4D). ROCK Inhibition Induces Transient Cofilin Dephosphorylation The polymerization/depolymerization of actin was shown to be necessary for NOG (Dent and Gertler, 2003), thus we investigated the signaling pathwa ys that involved actin dynamics. Cofilin is the most well characterized stimulus -responsive mediator of actin dynamics. Cofilin dissociates from F-actin when phosphorylated by LIMK. Since LIMK is a direct downstream effector of ROCK, ROCK inhibi tion decreases LIMK1 activity a nd dephosphorylates cofilin. In a series of experiments, we sought to identify the cytoskeleton signa l transducti on pathways through which Y-27632 mediated NOG in PC-12 cel ls. The morphology change of PC-12 cells in response to Y-27632 indicated that the init iation of NOG occurred within 5 minutes. Correlated with morphology change, more than 60% of cofilin underwent dephosphorylation within 5 minutes. Subsequent pa rtial recovery of phospho-cofili n was noticed during the neurite elongation and maintenance periods (6 to 24 hr) (Fig. 5A and B). There was no change in the
108 total expression level of cofilin during ROCK inhibitor treatment (Fig. 5A). Thus, dephosphorylation-phosphorylation of cofilin appears to be involve d in the initiation of NOG. We were intrigued by the observation that cofilin underwent partia l re-phosphorylation following the initial phase of dephosphorylation, de spite the presence of a constant level of ROCK inhibitor (Y-27632, 25 M) (Fig. 5). We hypothesized the involvement of additional protein kinases or protein kinase cross talk. Protein kinase-A, pr otein kinase C and MAPK have all been implicated in neurite outgrowth in PC -12 cells (Hundle et al., 1995; Obara et al., 2002; Christensen et al., 2003). PC-12 cells were first subjected to 25 M of Y-27632 for the maxima cofilin dephosphorylation. Then various protein kina se inhibitors were intr oduced to see if they would suppress cofilin re-phosphorylation. Howe ver, attenuated rephosphorylation was not bserved with any of the kinase inhibitors (Fig. 6). To further confirm that cofilin dephosphorylation is ROCK mediated, we again employed a second specific ROCK inhibitor H1152 (Fig. 7). Again a rapid dephosphorylation of cofilin was observed with a low concentration of H-1152 (1 M) within 30 minutes. Interestin gly, H-1152 differed from Y-27632 in that the Pcofilin reduction was sustained for up to 24 hours (Fig.7). Discussion Although ROCK inhibitor (Y-27632) induced NOG in PC-12 cells had been briefly mentioned in previous studies (Birkenfeld et al., 2001; Fujita et al., 2001; Kishida et al., 2004), there has been no comprehensive biochemical anal ysis of this important phenomenon to define the role of ROCK inhibition to date. For example, Birkenfeld et al. (Birkenfeld et al., 2001) only described briefly the effects of Y-27632 on NOG without detailed biochemical studies. In addition, they only used the less selective RO CK inhibitor Y-27632, but we now for the first time use H-1152 to complement the use of the le ss selective Y-27632 (Fig. 1-3). In our studies, both Y-27632 and H-1152 rapidly initiated NOG within 5 to 30 minutes with the formation of
109 small protrusions (neurite ini tiation) followed by neurite extens ion in 6 to 10 hours (elongation) (Fig. 2). Importantly, both ROCK inhibitors produce rapid cofilin-dephosphorylation. Since NGF-induced NOG in PC-12 cells by activating protein kinase C (PKC) (Christensen et al., 2003), it is possible that Y-27632 might exert its effects on PKC. However, this possibility was ruled out as PKC was not inhibited at all at 10 M of Y-27632, yet 1-5 MY-27632 already produced NOG effects (Fig. 1). Y-27632 is also known to inhibit MAP kinase activated protein kina se-1b (MAPAP-K1b) (IC50 19 M) but at a higher concentration than ROCK (IC50 800 nM) in vitro (Ishizaki et al., 2000). Under our cell treatment conditions, Y-27632 even as low as 1-5 M already produced NOG effects, so it is unlikely that it was attributable to MAPAP-K1b i nhibition. Furthermore, a more selective ROCK inhibitor H-1152 (Ikenoya et al., 2002; Sasaki et al., 2002) produced the same NOG effects at a concentration as low as 1 M (Fig. 3). Neuronal differentiation processes were medi ated by cytoskeletal reorganization, as observed with F-actin and micr otubules. Two major F-actin networks were observed in the filopodia and lamellipodia of the neuronal grow th cones (Tojima and Ito, 2004). Neuron-specific III-tubulin usually is expressed in cell bodies and neurites and highl y concentrated in neurites of mature neurons (Braun et al., 2002). In our st udy, F-actin accumulated in growth cone-like structures, lamellipodia and ne urites (Fig. 4B), whereas III-tubulin was prefer entially enriched along the nascent neurites of PC-12 cells follow ing ROCK inhibition (Fig. 4D). Similar to NGFinduced NOG, ROCK inhibition induced PC-12 cell differentiation into a neuronal phenotype as indicated by neuronal specific III-tubulin antibody staining a nd morphological changes (Fig. 4D). Thus, differentiated PC-12 cells are represen tative of neurons and can be used to further explore the downstream ROCK pathways in NOG.
110 Interestingly, ROCK inhibitor induced a ra pid decrease and then subsequent gradual increase in phosphorylation of cofilin, which co rrelated with the ini tiation and elongation of neurites (Fig. 5). It is important to point out that Maekawa et al had previously established that ROCK inhibition causes cofilin dephosphoryla tion through LIMK-1 (Maekawa et al., 1999). However, they did not study the relationshi p between ROCK-cofilin dephosphorylation and neurite outgrowth in model cell system such as PC-12 cells. Tojima and Ito recently proposed a signal transduction cascade th at involves ROCK inhibition decreasing LIMK 1 activity and dephosphorylating cofilin, thus inhibiting neurit ogenesis (Tojima and Ito, 2004). Yet, based on our studies, instead of inhibiting neuritoge nesis, our results showed that cofilin dephosphorylation coincided with NOG initiation (5 to 30 minutes), while cofilin rephosphorylation occurred during elo ngation and maintenance phase (6 to 24 hours) of NOG (Fig. 5A). Our results were consistent with the prev ious work of Aizawa et al, where a sequential cofilin phosphorylation-dephophorylation cycling occured during semaphoring 3A (Sema-3A) treatment on DRG cells on inhibitory substrate (Aizawa et al., 2001). Sema-3A, a chemorepulsive axonal guidance molecule, induces growth cone collapse via the LIMK-cofilin pathway regulating actin filament dynamics. The dynamic cofilin dephosphorylationphosphorylation found in our work indicates that in addition to LIMK, other signaling pathways may also be involved in the mechanism of regulat ing the cofilin cycling in NOG of PC-12 cells. Upon ROCK inhibition LIM kinase s are inactivated leading to dephosphorylation of cofilin and resulting in massive new barbed ends for initiation neurites. We attempted to elucidate the mechanism of cofilin re-phosphor ylation shown during longer Y27632 treatments (6 h to 24 h) (Fig. 5). Since protein kinase A, C and MAPK have all been implicated at NOG previously in PC-12 cells (Hundle et al., 1995; Ob ara et al., 2002; Christensen et al., 2003); we tested the
111 potential effects of pharmacological inhibi tors of these kinases on post-Y-27632 cofilin rephosphorylation. However, protein kinase A, protein kinase C a nd MAPK inhibition all failed to prevent cofilin re-phosphorylation (Fig. 6). Ot her signaling pathways, such as inhibition of the Slingshot phosphatase (Niwa et al., 2002) or type 1 and type 2A serine/threonine phosphatases (Ambach et al., 2000), may be invol ved in the re-phosphorylation of cofilin, which in turn leads to actin polymerization, and subsequently cont ributes to the elongati on and maintenance of neurites. This mechanism may also apply to the formation and/or stabi lity of essential actinbased structures in growth cones and postsynaptic densities (Revenu et al., 2004). In addition to inhibiting ROCKs, Y-27632 also inhi bits protein kinase C-related kinase (PRK)2 in vitro, which may contribute to the dephosphorylation-phosphoryl ation dynamics of cofilin (Davies et al., 2000). We thus tested a more specific ROCK inhib itor (H-1152) (Ikenoya et al., 2002; Sasaki et al., 2002). It is of interest to note that, when PC-12 cells treated with H-1152, cofilin dephosphorylation was sustained for 24 hours w ithout notable rephospho rylation (Fig. 7). Further work is needed to understand crosstalk between signaling cascades in ROCK inhibition mediated NOG. Understanding the signaling mech anisms of ROCK inhibition mediated NOG opens up the possibility for developing novel st rategies to promote a xon regeneration in vivo. Clinically, the use of a ROCK i nhibitor may be useful for deve loping therapies in CNS following damage by Alzheimers disease (Zhou et al., 20 03), spinal cord injury (Ellezam et al., 2002), traumatic brain injury (Brabeck et al., 2004) and stroke (Brabeck et al., 2003).
112 Figure 5-1. Neurite outgrowth of PC-12 cells in response to ROCK inhibitor Y-27632 in a dosedependent manner. Top panel : A through F are phase-contra st images of PC-12 cells following treatment by different concen tration of Y-27632 for 24 hours. (A) 0.01 M; (B) 1 M; (C) 5 M; (D) 10 M; (E) 25 M; and (F) 100 M of Y-27632. Scale bar represents 50 m. Bottom Panel : Quantification of neurite outgrowth following Y27632 treatment for 24 hours. Cells with at l east one neurite greater than the diameter of the cell body were counted and expressed as a percentage of the total number of cells in a field. For each experiment, at le ast 50 cells were randomly measured. Data shown are mean values S.E.M, n = 3. Statistical significance of differences ( P <0.01) between the control and each tr eated group was determined by one-way ANOVA with Dunnett's multiple comparison tests.
113 Figure 5-2. ROCK inhibitor Y-27632 induced neur ite outgrowth in PC-12 cells in a timedependent manner. Panels A through I are phase-contrast imag es of PC-12 cells following 25 M Y-27632 treatment for different time points: A ) Control B ) 5 min; C ) 10 min; D ) 30 min; E ) 1 hr; F ) 3 hr; G ) 6 hr; H ) 10 hr; and I ) 24 hr. Arrows indicate examples of typi cal neurite outgrowth in PC12 cells over time. Cells were plated onto 6-well plates at a cell density of 4 x 103/cm2. Scale bar represents 50 m.
114 Figure 5-3. Quantification of neurite outgrow th post Y-27632 treatment in PC-12 cells. Quantification of cells with neurites A ) and neurite lengths B ) were performed after PC-12 cells were treated with 25 M Y-27632 at various time points. The length of the longest neurite was counted for cells with at least one identif ied neurite. Neurite length was determined by manually tracing the length of the longest neurite per cell. For each experiment, at least 50 cells were randomly measured. Values represent means S.E.M. n = 4.
115 Figure 5-4. Reorganization of cytoskeletal arch itecture in ROCK inhibition mediated neurite outgrowth. PC-12 cell nuclei were visuali zed with DAPI-DNA staining (blue). Panel A and B were FITC-phalloidin F-actin-stain ing (green) following treatment with Y27632. Before treatment, F-actin is uniformly localized in the periphery of the soma (triangle in panel A). After 12-hour treatme nt, F-actin became highly accumulated in growth cone-like structures (short arrow in panel A ), the neurites (long arrow in panel C) and lamellipodia (triangle in panel B ). Panels C through D were Alexa 488conjugated neuronal specific III-tubulin immunostaining (green). Most III-tubulin is concentrated in nasc ent neurites (arrow in panel D ). Cells were treated by DMSO as a control (A, C) or 25 M Y-27632 (B, D) for 12 hours prior to fixation. Scale bar represents 50 m.
116 Figure 5-5. Immunoblot analysis of cofilin phosphorylation following Y-27632 treatment. A ) PC-12 cells were treated with ROCK inhibitor (Y-27632; 25 M) for various time points. Total protein lysate were extracted for immunoblotting analysis with phosphocofilin (P-cofilin) and total cofilin antibodies Representative blots were shown here. B ) Densitometric analyses of cofilin immunoblots were performed. The optical densities of phospho-cofilin we re normalized to the corresponding values for total cofilin. Values represent means S.E.M. n = 4. Statistical significance of differences between the control and each treated group was determined by one-way ANOVA with Dunnett's multiple comparison tests. A difference was considered to be statistically significant when the P value was less than 0.01 (* P < 0.01).
117 Figure 5-6. Effect of various pr otein kinase inhibitiors on co filin rephosphorylation in the presence of Y-27632. (A) PC-12 cells were treated with ROCK inhibitor (Y-27632; 25 M) for 2 h. For the same conditions, prot ein protein kinase-A (H-89; 500 nM), protein kinase-C (Ro-32-0432, 300 nM) or MAPK (PD98059, 30 M) attenuated cofilin rephosphoryaltion. The cells were further incubated to 16 hour before cell lysate was collected for immunoblotting anal ysis with phospho-cofilin (P-cofilin) and total cofilin antibodies. (B) Densitometric analyses of cofilin immunoblots were performed. The optical densities of phospho-cofilin were normalized to the corresponding values for total cofilin. Values represent average from two experiments (*P < 0.01).
118 Figure 5-7. Persistent cofilin phosphorylation fo llowing ROCK inhibitor H-1152 treatment. (A) PC-12 cells were treated with ROCK i nhibitor (H-1152; 1 M ) for various time points (30 min, 2 h, 12 h and 24 h). Total protein lysate were extracted for immnuoblotting analysis with phosphor-cofilin (P-cofilin) and total cofilin antibodies. Representative blots were shown here. (B) Densitometric analyses of cofilin immunoblots were performed. The optical densities of P-cofilin were normalized to the corresponding values for total cofilin. Va lues represent means from two separate experiments.
119 CHAPTER 6 SYSTEMS BIOLOGY APPROACH TO DE CIPHER NEURITOGENESIS: ROCK PATHWAYS IN MEDIATING NEURI TE OUTGROWTH IN PC-12 CELLS Introduction Traumatic brain injury (TBI) is a major cause of death and disability in young people less than 24 years old (www.cdc.gov). The pathogenesis of TBI involves two components: the initial mechanical injury and the subsequent secondary events, which further increase the brain damage. The secondary injury results in a massive en largement of the initial small primary lesion, typically characterized by disrupt ed axons, a cystic cavity encas ed within a g lial scar, and a variable amount of intact tissu e (Hayes et al., 1998; Denslo w et al., 2003; Wang et al., 2004). The clinical challenge is to pr omote compensatory sprouting from neurons with intact axons or the re-growth of axons across the injury site. Unli ke the peripheral nervous system, axons in the adult central nervous system (CNS) rarely under go spontaneous regeneration after injury. This lack of regeneration appears to be due to inhib itory elements within the CNS environment, such as myelin and the glial scar. Over the past tw o decades, Nogo-A, myelin-associated glycoprotein (MAG), and oligodendrocyte myelin glycoprotein (OMgp) have been well-characterized as CNS/myelin-associated inhibitors (Yamashita et al., 2005). Other neurite growth inhibitors have also been found directly at the le sion site or glial scar tissue. Th ese inhibitors include chondroitin sulfate proteoglycans (CSPG), semaphorins, ephr ins, and other repulsive guidance molecules (McKerracher et al., 1994; Kwon et al., 2002; McKerracher and David, 2004). Interestingly, recent studies indicate that Rho-Rho kinase (ROC K) is at the convergent point of those multiple neurite growth inhibitory pathwa ys. A number of studies have de monstrated that inhibition of small Rho-GTPases by bacterial toxin Clostridiu m botulinum exoenzyme C3 transferase (C3) and Rho kinase (a downstream effecter of RhoA) by Y-27632 can promote axonal regeneration by overcoming the inhibitory effects of CSPG, as well as those of semaphorin 4A, ephrins A5
120 and other repulsive guidance mol ecules (Wahl et al., 2000; Monnier et al., 2003a; Yukawa et al., 2005). More specifically, Nogo, MAG, OMgp through downstream receptor NgR, p75 and LINGO-1 complex, activate the RhoA-ROCK pathway to execute the neurite growth inhibition (Selzer, 2003; Tang, 2003; Yamashita et al., 2005). Most importantly, a variety of evidences indi cate that TBI or spinal cord injury (SPI) induces activation of RhoA and RhoB at the lesi on site in human brains (Brabeck et al., 2003; Brabeck et al., 2004) and of ROCK I and ROCKII in rats (Aimone et al., 2004). Intriguingly, the observed upregulation of RhoA and RhoB is still detectable months after TBI in humans (Brabeck et al., 2004) and at least for a period of 4 weeks after spinal cord injury (SCI) in rats (Conrad et al., 2005). Due to the persistent ac tivation of Rho-ROCK pathway around the lesion site, makes Rho-ROCK inhibition an attractive site for therapy intervention not only for acute and sub-acute treatments, but also for delayed interventions after CNS injury. Fortunately, a number of specific ROCK i nhibiting compounds have been developed, including H-89, HA-1077, Y-27632 (Davies et al., 2000), wf-536 (Nakajima et al., 2003), and H1152 (Ikenoya et al., 2002) The selective inhi bition of ROCK makes Y-27632 very useful for evaluating ROCK functions in CNS since ROCK is highly expressed in the brain. Taken together, targeting the ROCK pathway offers the advantage of antagonizing multiple neurite growth inhibitors. However, the Rho-ROCK path way is also involved in a number of ubiquitous cell functions unrelated to neurite genesis in cluding ruffling, motility, cytokinesis, and cell spreading (Schwartz, 2004; Hall, 2005). Therefor e, it is reasonable to assume a subgroup of downstream targets from ROCKar e directly involved in neurit e growth while the rest are involved in other ROCK-m ediated cell functions.
121 In this study, a system biological approach for studying the PC-12 cells neurite outgrowth model in vitro was applied to identify the molecular players in the ROCK pathways linked to neurite genesis. Here, we conducted gene re gulatory networks, protein networks, and relationships between genomic and proteomic le vels with neurite mor phological changes that might be linked with the axon regeneration. This new research generated a massive data set, from which emerged the junction of two different fields, biomedicine and data analysis. One of the pioneering interdisciplinary fields that ha s enormous potential for making the most of the information such as in the above data sets is systems biology. Systems biol ogy is a latest addtion to biology that aims at system-l evel understanding of complex bi ological processes. The scope of systems biology combines multiple advanced resear ch areas, such as biosciences, data mining, control theory and a number of different engineering fields (Kitano, 2002). The novelty of the present study is in combining advanced data mining methods with the novel application of systems biology to study neurite outgrowth and axonal regeneration. Specifically, we applied correspondence analysis and DTAselect to the da ta sets produced by sophisticated biological experiments that were carried out to study the underlying mechanisms at the gene and protein levels (Oda et al., 2005). This study will help si gnificantly in increasing our understanding of the mechanisms involved in axonal regeneration and w ill help in developing strategies to manipulate axonal regeneration in various CN S pathologies such as TBI, st roke and Alzheimers disease. Experimental and Computational Methods Cell Culture PC-12 cells were maintained at 37C in Dulbecos modified Eagles medium (DMEM) supplied with 10% fetal bovine serum (FBS), 5% heat-inactivat ed horse serum, 100 g/ml of streptomycin, 100 U/ml of penicil lin and 1% Fungizone (Gibco, Ro ckville, MD) in a humidified 5% CO2 incubator. To induce neurite outgrowth, PC12 cells were plated in the same medium
122 with 25 M Y-27632 (Calbiochem, San Diego, CA). An equal amount of DMSO was added to control plates. Micrographs of cells were taken at 32 x with an Axiocam digital camera using a Zeiss Axiovert 135 microscope. RNA Isolation Total RNA from control and Y-27632 treated PC12 cells were isolated using the RNeasy Mini Kit (Qiagen, Maryland USA) according to th e manufacturer's instructions. The amount and quality of RNA was determined using the Agile nt 2100 system before performing the microchip arrays. Two independent pairs of samples were performed. Affymetrix GeneChip Differential gene expression analysis wa s performed on Affymetrix GeneChip Rat genome 230 2.0 (Santa Clara, CA). Briefly, a T7 promoter-dT primer was used to generate cDNA with SuperScript II revers e transcriptase (Invitrogen, Ca rlsbad, CA) and amplified by in vitro transcription using T7 RNA polymerase. A second round of cDNA synthesis was performed using random hexamers as primers, and a final in vitro transcription using biotinlabeled nucleotides (Bioarray Hi gh Yield RNA; Enzo Life Scie nces, Farmingdale, NY) yielded labeled cRNA for microarray hybri dization. Fragmentation of th e cRNA, hybridization, staining, and scanning of the microarray were performed according to the GeneChip Expression Analysis Manual provided by Affymetrix. Sample Preparation for Protein Differential Analysis Control and ROCK inhibitor treated PC-12 ce lls were solubilized in 0.1% SDS, 150mM NaCl, 3 mM EDTA, 2 mM EG TA, 1% IGEPAL, 1 mM Na3VO4 and a protease inhibitor cocktail tablet (Roche, Indianapolis, IN). 1mg samples of each were prepared for protein sepression. A Bio-Rad (Hercules, CA) Biologic DuoFlow syst em with QuadTec UV detector and BioFrac fraction collector was used with Uno series SCX (S1) and SAX (Q1) serially placed ion
123 exchange columns. Ice cold 20 mM Tris-HCl buffe rs (pH 7.5) were used (mobile phase A) with 1 M NaCl (Fisher Scientific, crystalline 99.8% cert ified) as the elution buffer (mobile phase B). Thirty-one 1ml fractions were automatically coll ected. Differential analys is between control and treatment groups were performed by pairing fr actions for loading side by side on Bio-Rad Criterion 10-20% gradient sodi um dodecyl sulfate-polyacrylam ide gel electrophoresis (SDSPAGE). In-Gel Digestion Bands (4 x 1 mm) with differential protein expression were excised and washed with HPLC water, then 50% 100 mM ammonium bi carbonate/ 50% acetonitrile. Pieces were dehydrated with 100% acetonitrile followed by speed vac. Cubes were rehydrated with 10 mM dithiothreitol for 30 min at 56 C. Dithiothreitol was re placed by 50ul of 55mM ammonium bicarbonate and incubated for 30 minut es in the dark at room temp erature. This was followed by washing step followed with 100% acetonitrile, 15 ul of a 12.5 ng/ul trypsin solution for 30 minutes at 4 C, then 20 ul of 50 mM ammonium bicarbonate was added overnight at 37C. The peptide extract was dried and re suspended in mobile phase. Capillary RPLC-MSMS Based Protein Identification Capillary reversed phase liquid chromat ography tandem mass spectrometry protein identification was performed as described previously. Briefly, sample digests (2 L) were loaded via an autosampler onto a 100 m x 5 cm c-18 reversed phase capillary column at 1.5 l/min. Peptide elution was performed by linear gr adient: 5% to 60% meth anol in 0.4% acetic acid over 30 minutes at 200 nL/m in. Tandem mass spectra were co llected in data-dependant mode (3-most intense peaks) on a Thermo Electron LCQ Deca XP plus ion trap mass spectrometer. Tandem mass spectra were searched against a NCBI rat indexed RefSeq protein database using Sequest. Filtering and sorti ng was performed with DTAselect software by
124 peptide number and Sequest cross correlation values (Xcorr values of 1.8, 2.5, 3.5 for +1, +2, +3 charge states(Tabb et al., 2002). Peptides filtered and sorted by DTAselect were assigned to specific protein accession numbers (National Center for Biotechnology Information [NCBI]). Feature Selection Two related statistical techniques are used in this study. The first tec hnique is implemented using the two-sample t-test given that the normality assumption holds for the data. The other technique is based on the Wilcoxon rank-sum test, wh ich is a non-parametric alternative to the ttest, allowing us to avoid the above assumption. Results Correpsondence Analysis for the Microarray Da ta of ROCKi Induced Neurite Outgrowth We performed a numerical analysis of the data set describing the gene expression levels that were obtained in the micr oarray experiments. By using co rrespondence analysis, all samples and data points of a given data set can be ma pped onto one low-dimensional space visualized as either as a biplot (2-D) or as a 3-D diagram (Fe llenberg et al., 2001). Each ax is of such a plot or diagram has the remarkable tendency for providing an insightful characterization of the data set. Furthermore, the samples/data points with a hi gh degree of similarity with respect to such characterization also have very similar coordinates on the corres pondent axis. A data set can be conveniently described by a rectangular matrix A = (aij)mn of n samples (columns of the matrix) and m data points (rows of A). When working with the microarray data, genes are represented by rows, so that the value aij specifies the expression level of gene i in sample j. Given that we handle both columns and rows of th e data matrix in a unified manner. This is in addition to the fact that we ar e always able to use the transp ose of the data matrix without making any changes in the algorithm, it follows th at, without loss of gene rality, we can assume
125 that m > n. In order to perform corresponding an alysis, we first construct the correspondence matrix P = (pij)mn by computing 11 ij ij mn ij ija p a To some extent, the correspondence matrix is similar to a two-dimensional probability distribution table, where the sum of the elements equals 1. Next we calculate masses of rows and columns as follows: 1 11 n ij j i mn ij ija r a 1 11 m ij i j mn ij ija c a Notice that the masses of columns or rows are also comparable to the marginal probability densities. Subsequently, we build the matrix S = (sij)mn, to which SVD is later applied ijij ij ij p rc s rc By means of the SVD the matrix S is represen ted as a product of thr ee matrices as follows: S = U^V T where matrices U and V are unitar y, and D is a diagonal matrix. In particular, columns of the matrix U = (uij)mn are orthono rmal vectors spanning the columns of S, while columns of the matrix V = (vij)nn are orthono rmal vectors spanning the rows of S. One can easily show that in order to de scribe the optimal low-dimensi onal subspace (where the columns of S are to be projected) wher e the information loss would be mi nimum, it is enough to consider a suitable number of first colu mns of U. Likewise, the optimal low-dimensional subspace for projecting the rows of S can be constructed using an equal number of first columns of V. In
126 addition, as a consequence of the aforementioned properties of the matrix S, the columns and rows of the original data set, matrix A, may be represented in one low-dimensional space of dimensionality K < n in the following manner: ,1,2,...,kik ik iu f kK r gives the k-th coordinate of the row I, and ,1,2,...,kjk jk jv gkK c gives the k-th coordinate of the j-th colu mn in the new space. Clearly, by choosing K = 2, we obtain a bi-plot, while taking K = 3 gives us a 3-D diagram of the analyzed data set. All genes shaving at least one absent call in the data se t were discarded. The resu lting data array had 4 samples and 11,419 features (genes). Next, a statis tical analysis was performed to select the informative genes by means of the Wilcoxon rank -sum test. The feature selection algorithm based on the Wilcoxon test selected a total of 4,002 relevant features out of the analyzed 11,419 features. With a high probability we assumed that the selected relevant features represent the genes that are involved in regulation of th e processes underlying neurogenesis and axonal regeneration. To observe the up-regulated gene s from the down-regulated genes we further obtained a two-dimensional projec tion of the data using corres pondence analysis. The resulting projection onto the plain is shown in Figure 6-1. From the biplot of the data in Figure 6-1, we can see that the first (horizontal) axis discriminates all the active (up-regulated) genes ag ainst all the genes that need to be suppressed for a neuritogenesis in a PC-12 cell to take place (i.e. down-regulated). Part icularly, the points of the biplot with the most positive coordinates on this axis correspond to the up-regulated genes, whereas the points having the most negative c oordinates represent th e genes that are downregulated.
127 Identification of ROCKi-induced Neurite Outgrowth Transcriptome The list of genes whose biplot projection points have the most positive first coordinates gives the up-regulated genes upon ROCK inhibiti on induced neurite outgr owth, shown in Table 6-1 below. Next, we also selected points with the most negative first c oordinate. The resulting down-regulated genes are summarized in Table 62. All genes with pertur bation were searched against the Gene Ontology Consortium (http://www.geneontology.org (Harris et al., 2004)) and grouped into categories that defi ne their biological and molecula r functions, as indicated in Tables 1 and 2. Many of those genes were found to be functionally unclassified. The rest of the groups of up-regulated genes are involved in tr anscription, cell signaling, cell cycle regulation, cytoskeleton/motility protein, cell adhesion recepto r, trafficking proteins, cell receptors and metabolism et al. Some of the up-regulated gene s include rhoB, drebrin, neuritin, S-100 calcium binding protein, olfactomedin 1, Per1 interactin g protein, protein kina se C, BCL-2 and bone morphogenetic protein 2 (BMP2). neurexophi lin 3, prostaglandin E synthase, bone morphogenetic protein 7, histone dea cetylase 10, and Nogo-66 receptor (NgR). Identification of Differentially Ex pressed Proteins by RPLC-MSMS Differential protein expression analysis was accomplished by multi-dimensional separations involving biphasic ion-exchange chromatography in tandem followed by LC-MS/MS protein identification. Thirty-two fractions colle cted from each CAX experiment were paired (i.e., fraction 1of control with fr action 1 of ROCK inhibitor treatment) and loaded side-by-side onto 1D-PAGE for the second dimension protein se paration. The gels we re visualized with Coomassie blue stain for differential band an alysis. Thirty-five ba nds with an observed difference in densitometry were selected and exci sed for proteomic analysis as boxed and labeled in Figure 6-2. In all, more than 20 proteins were confirmed to be differentially expressed between control and Y-27632 treated PC-12 cells as shown in Table 6-3. Those proteins are
128 presented here with their resp ective GI numbers along with th e calculated mass spectrometry sequence coverage along with the SDS-PAGE appa rent molecular mass. The identified proteins were grouped into upregulated or downregulated abundance in Table 6-4. The proteins that upregulated upon ROCK inhibitor tr eatment include cytoskeleton associated proteins (talin, 143-3 protein and filamin), S-100 calcium binding protein, heat shock protein, ubiquitin carboxyl extension protein 80 and programmed cell death 6 interacting protein and others. Among the downreuglated proteins are annexin II, cerebella r postnatal development protein 1, proliferation related acidic leucine rich protein PAL31 and heat shock protein 70K. Discussion In this paper we employed an innovative methodology for applying the systems biology approach to the complex biological syst em underlying neurite outgrowth and axonal regeneration. We presented an integrated appr oach to data mining of microarray data that incorporates feature selection ba sed on statistical analysis with the dimensionality reduction by means of projection methods (Felle nberg et al., 2001). In particular the relevant features were selected from the obtained gene expression data using the Wilcoxon rank-sum test, and then the selected data was projected onto a two-dimensional space us ing the correspondence analysis algorithm. Utilizing statistical pr ocedures to perform feature selection significantly aids in reducing the number of features in a data set. In addition, the Wilcoxon test is a non-parametric procedure, and hence does not require any assumptions about th e distribution of data to hold. Also application of correspondence analysis allo ws us to obtain an informative projection of high-dimensional data onto a low-dimensional sp ace, and then visually explore the data. In addition, correspondence analysis offers some add itional advantages. First, this technique does not require any prior information about classification of samples or features of the data set. Second, unlike many other data mining procedures, correspondence analysis is very
129 computationally efficient. To summarize, applic ation of data mining techniques to the study of systems biology of PC-12 cells proves to be especial ly useful as it allows us to answer the key questions about the processes unde rlying neurogenesis at the gene level. Furthermore, such approach can be applied to many others biomedi cal studies based on analysis of the microarray data that can be naturally s ubdivided into some classes. Of those differentially expressed genes, some of them are particularly interesting and need to be further studied. It has been reported that applic ation of the selectiv e ROCK inhibitor, Y27632 preferentially lowered brain levels of Abeta42 in a transgenic mouse model of Alzheimer's disease (AD) (Z hou et al., 2003). However, the mechanism underlying the ROCK inhibition contribution to AD is still unknown. Our data suggests that Y-27632 induces upregulation of drebrin and down-re gulation of prostaglandin E synt hase. Drebrin, an acitn-binding protein, is localized in postsynapt ic terminals of adult brains a nd regulates synaptic plasticity. A remarkable reduction of drebrin was found in Alzh eimer's disease brains (Harigaya et al., 1996; Counts et al., 2006). In another case, Prostaglandin E (PGE) synt hase highly influences the synthesis of the inflammatory factor PGE. Thus, Y-27632 may alleviat e the pathogenesis and memory loss of AD by inhibiting prostaglandin E synthase and stimulating the expression of drebrin. Interestingly, several apoptosis related genes have also been implicat ed in relation to the ROCK inhibitor induced neurite outgrowth model, such as Bcl-2 and Annexin II. The role of Bcl-2 family members on axonal regeneration ha s thus far been controversial. Delivering recombinant, or over-expressed Bcl-2, into retin al tissue stimulates a xonal initiation but not axonal elongation after crus h injury to retinal expl ants, which is independent of its anti-apoptotic role. The subsequent activation of ERK and CREB may underlie the ability of Bcl-2 to promote
130 axon regeneration (Chen et al., 1997; Jiao et al ., 2005; Dietz et al., 2006). ROCK inhibition might lead to increased levels of Bcl-2, there by contributing to the axon regeneration after brain injury. It has been shown that MAG, NogoA a nd OMgp can bind to NgR, Lingo and p75 or TROY to form trimeric receptor complexes, and thus activate Rho-ROCK pathway leading to its inhibitory effects on axons (Yamashita et al., 2005). Consistent with previous studies, ROCK inhibition reduces NgR expre ssion, thereby enhancing neurite outgrowth. Neuexophilin 3 is highly expressed in the brain (Beglopoulos et al., 2005; Duda nova et al., 2006; Craig and Kang, 2007). A group of bone morphogenetic proteins also have been ch anged with Y-27632 treatment. The exact influence of ROCK on those proteins is still unknown. Furthe r genetic studies are needed to explore their functions. There were 35 pairs of bands were cut in our proteomic study. Eight een proteins were identified as being up-regulated and four proteins were found to be down-regulated after ROCKi treatment. Of these identified proteins, onl y one protein, S-100 calcium binding protein, was shown to be differentially regulated at the mRNA level using microarray analysis. The lack of correlation between mRNA and protein levels may be due to several reas ons. In terms of the transcriptome, microarrays provide a abundance meas urement for a finite set of targets, while a typical proteome analysis does not. The current technical challenges, such as the presence of splice isoforms in the transcriptome, the inco mplete protein index database and sensitivity limitations of mass spectrometry in proteome analys es affect the data sets generated. In addition, instead of being static, mRNAs and proteins are highly dynamic, they change in numerical, spatial and temporal patterns. T ypical transcriptomic and proteo mic profiles are snapshots of a specific time point. There is no t enough resolving power to di stinguish newly-synthesized
131 transcripts or proteins fro m those accumulated over time. Dynamic and high-throughput quantitative assays might be needed to dete rmine the proteome. Finally, the proteomes we examined at a specific point in time isnot a replica of the underlying transcriptomes. After transcription from DNA to RNA, the gene transcript can be spliced in different ways prior to its translation into a specific protein. Following tran slation, most proteins are chemically altered through post-translational modifications. As a cons equence, the information from a single gene may encode many different proteins. There ar e already compelling data suggesting that proteomics is a complement or alternative to mRNA based measuremen ts. Those quantitative data sets will give us insight into how cells res ponsed to a given stimuli and allow us to construct a systems biology.
132 Figure 6-1. Correspondence analysis bi pot for PC-12 microarray dataset
133 Figure 6-2. One representative proteome of ROCK inhibiti on induced neurite outgrowth visualized on 1D-PAGE following CAX Fractionation. One milligram of PC-12 lysate was divided into 32 CAX fracti ons resolved on 1D-polyacrylamide gel electrophoresis and gels were stained with Coomassie Blue staining. Control and treatment samples were run side by side.
134 Table 6-1. The up-regulated genes upon Y-27632 treatment Functional group % Transcription 4.2 Cell cycle 1.4 Oncogenes and tumor suppressors 6.9 Membrane channels and transporters 1.4 Trafficking/targeting proteins 1.4 Metabolism 5.6 Post-translational modification/protein folding 2.8 Apoptosis associated proteins 1.4 Cell receptors 4.2 Cell signaling, extr acellular communication proteins2.8 Intracellular transducer s/effectors/modulators 5.6 Cytoskeleton/motility proteins 2.8 DNA synthesis, recombination, and repair 1.4 Unclassified 56.9 Table 6-2. The down-regulated genes upon Y-27632 treatment Functional group % Transcription 5.6 Cell cycle 1.1 Cell adhesion receptors/protein 1.1 Oncogenes and tumor suppressors 1.1 Stress response proteins 2.2 Membrane channels and transporters 1.1 Trafficking/targeting proteins 3.3 Metabolism 3.3 Translation 1.1 Cell receptors 5.6 Cell signaling, extr acellular communication proteins2.2 Unclassified 72.2
135 Table 6-3. Proteins with Increa sed Abundance post Y-27632 treatment Protein Name GI Number MW of Protein(kDa) MW from gel(kDa) Matched peptides in Control Sequence Coverage in Control (%) Matched peptides in ROCKi Sequence Coverage in ROCKi (%) Aconitase 2, mitochondrial gi|18079339 85.5 82 1 1.2 13 18.2 Fibroblast growth factor receptor 4 gi|6679789 89.8 90 0 0 1 1.4 Aconitase 2, mitochondrial [Mus musc ulus gi|18079339 85.5 82 1 1.7 8 12.6 S100 calcium-binding protein A gi|4506761 11.2 11 0 0 1 7.2 Ubiquitin carboxyl extension prot ein 80 gi|4506713 18 10 2 16 4 25.6 Programmed cell death 6 interacting protein gi|6755002 96 95 1 1.2 12 13.3 Transketolase gi|6678359 67 65 0 0 4 2.7 Eukaryotic translation elongation factor 2 gi|33859482 95 90 0 0 8 10.8 Gamma filamin gi|4557597 288 280 3 1.6 12 5.1 Heat shock 90kDa protein 1 gi|2014594 84 85 1 1 3 3.2 Laminin receptor 1 gi|31560560 33 38 3 11.9 6 20.3 L4-3-3 protein, zeta polypeptide gi|6756041 27.7 28 1 4.9 2 10.6 L4-3-3 protein tau gi|5803227 27.7 28 1 6.5 2 12.2 Talin 1 gi|16753233 270 270 0 0 9 4.1 Alpha 2 macroglobulin precursor gi|4557225 163 180 0 0 2 1.4 SET translocation gi|13591862 33 35 0 0 1 3.5 Table 6-4. Proteins with decrea sed Abundance post Y-27632 treatment Protein Name GI Number MW of Protein(kDa) MW from gel(kDa) Matched peptides in Control Sequence Coverage in Control (%) Matched peptides in ROCKi Sequence Coverage in ROCKi ( % ) Heat shock protein 70k gi|3198169070.8 71 1 12.2 7 14.1 Annexin II gi|4757756 38.6 35 6 20.4 0 0 Proliferation related acidic leucine ri ch protein PAL31gi|1877777031 31 1 5.1 0 0 Cerebellar postnatal development protein 1 gi|3154213129.6 30 1 5.4 0 0
136 CHAPTER 7 CONCLUSIONS AND FUTURE DIRECTIONS Many significant advances have been made recently to sharpen our understanding of important molecular mechanisms involving in axon/neuronal injury. Given one of the key mediator of axonal injury is Ca2+ dyshomeostas is, calmodulin binding proteins (CaMBPs) are particularly vulnerable to two abnormal pr ocesses: (i) over-activation by calcium bound calmodulin (CaM), and (ii) proteolytic processing by calpains and caspases. Thus, it is imperious to study the calmodulin signal pathway in a systema tic manner after traumatic injury. In chapter 2, we developed a novel CaM-affinity capture method coupled with reversed-phase liquid chromatography tandem mass spectrometry (RPLCMSMS) to identify the calcium-dependent CaM-binding proteome in rat brain and its vulne rability to calpain and caspase. Our results suggested that this is a simple and efficient way to explore th e CaM-binding proteome and its vulnerability to proteolysis. A comprehensive pr otein-protein interaction map was constructed to facilitate understanding how brain cells respond in terms of in itiating proteases to Ca2+ stimuli. In this map, brain enriched CRMP-2 is of most interest to further study. CRMP-2 plays a crucial role in neurite outgrowth and axon formation. As described in chapter 3, it is the first time we demonstrated the degradation of CRMP-2 after ac ute neuronal injuries by in vivo TBI and in vitro glutamate excitotoxicity. Furthermore, calpa in-2 was identified as the possible proteolytic mediator of CRMP-2 following ex citotoxic injury and TBI which appears to correlate well with neuronal cell injury and neurite damage. CaM bi nds to CRMP-2 in a calcium dependent manner, thereby preventing proteolysis of CRMP-2 both in vivo and in vitro. Meanwhile, CaM binding to CRMP-2 might play a role in F-actin bundling, and in turn maintain a stable structure or enhance axonal regeneration.
137 Although it is excited to construct a Ca2+/C aM interactions map, the mechanisms of axonal injury in CNS is enormously complex. A xons fail to regenerate in the adult central nervous system (CNS) after injury. Fortunatel y, both inhibitory and permissive pathways converge at the Rho/ROCK (Rhoassociated kinase) pathway. T hus, Rho-ROCK is an emerging target to promote axonal regeneration and func tionary recovery. Howeve r, inhibition of ROCK pathway in vivo is a complex proc ess, with numerous specific dow nstream effectors that may or may not involve in axonal regeneration. This is further complicated by the temporal and spatial nature of many of these effects. As a result, it is critical to evaluate the entire cellular system from the neuritogenesis perspective to include di fferential expression of tr anscriptome, proteome and even phosphor-proteome. In chapter 5, we found that ROCK inhibitor can induce robust neurite outgrowth in PC-12 cells and it initiate d neuritis through dephosphorylation of cofilin. Next, differential gene transcriptome and prot ein expression influenced by ROCK inhibition were identified by affymetrix Mciroarray and pr oteomic studies in a systems biology approach. More than 200 genes and 20 proteins, many of which are known to be associated with neuritogenesis or cell growth, we re potentially involved in RO CK inhibition induced neurite outgrowth in PC-12 cells. Post translation m odification might contribute to the lack of correlation between mRNA and protein level. Thus a comprehensive picture emerged afte r applying ROCK inhibi tion from not only one single downstream effecter but whol e cell response. Studies such as these can ultimately be used to understand the molecular events that accomp any neuritogenesis, and therefore to develop more specific therapeutic targets and assess the ce ll response after applying the medicine. Further investigation is underway be explor ing the role of these genes or pr oteins in injury animal model by a combination of biochemistr y, cell biology and molecular biology.
138 Taken together, data from our studies provide clues regarding the perturbation factors that are important in the axonal injury/regeneration. We will further investigate the role of these proteins in pathophysilog ical significance. Clinically, this systemic biological approach to decipher signal pathways may provide promis ing potential in findi ng novel biomarkers and therapeutic targets after axonal injury.
139 LIST OF REFERENCES Adamec E, Beermann ML, Nixon RA (1998) Calpain I activation in rat hippocampal neurons in culture is NMDA receptor selective and not e ssential for excitotoxi c cell death. Brain Res Mol Brain Res 54:35-48. Aguayo AJ, David S, Bray GM (1981) Influences of the glial environment on the elongation of axons after injury: transpla ntation studies in adult r odents. J Exp Biol 95:231-240. Aimone JB, Leasure JL, Perreau VM, Thallmair M (2004) Spatial and temporal gene expression profiling of the contused rat sp inal cord. Exp Neurol 189:204-221. Aizawa H, Wakatsuki S, Ishii A, Moriyama K, Sa saki Y, Ohashi K, Sekine-Aizawa Y, SeharaFujisawa A, Mizuno K, Goshima Y, Yahara I (2001) Phosphorylation of cofilin by LIMkinase is necessary for semaphorin 3A-induced growth cone collapse. Nat Neurosci 4:367373. Amano M, Fukata Y, Kaibuchi K (2000) Regulati on and functions of Rho-associated kinase. Exp Cell Res 261:44-51. Amano M, Kaneko T, Maeda A, Nakayama M, Ito M, Yamauchi T, Goto H, Fukata Y, Oshiro N, Shinohara A, Iwamatsu A, Kaibuchi K (2003) Identification of Ta u and MAP2 as novel substrates of Rho-kinase and m yosin phosphatase. J Neurochem 87:780-790. Ambach A, Saunus J, Konstandin M, Wesselbor g S, Meuer SC, Samstag Y (2000) The serine phosphatases PP1 and PP2A associate with and ac tivate the actin-binding protein cofilin in human T lymphocytes. Eur J Immunol 30:3422-3431. Araujo IM, Verdasca MJ, Leal EC, Bahr BA, Ambrosio AF, Carvalho AP, Carvalho CM (2004) Early calpain-mediated proteolysis follo wing AMPA receptor activation compromises neuronal survival in cultured hippoc ampal neurons. J Neurochem 91:1322-1331. Arimura N, Inagaki N, Chihara K, Menager C, Nakamura N, Amano M, Iwamatsu A, Goshima Y, Kaibuchi K (2000) Phosphorylation of colla psin response mediator protein-2 by Rhokinase. Evidence for two separate signaling pa thways for growth cone collapse. J Biol Chem 275:23973-23980. Arimura N, Menager C, Fukata Y, Kaibuchi K (2004) Role of CRMP-2 in neuronal polarity. J Neurobiol 58:34-47. Arimura N, Menager C, Kawano Y, Yoshimura T, Kawabata S, Hattori A, Fukata Y, Amano M, Goshima Y, Inagaki M, Morone N, Usukura J, Kaibuchi K (2005) Phosphorylation by Rho Kinase Regulates CRMP-2 Activity in Gr owth Cones. Mol Cell Biol 25:9973-9984. Arundine M, Tymianski M (2003) Molecula r mechanisms of calcium-dependent neurodegeneration in excitoto xicity. Cell Calcium 34:325-337.
140 Babcock DF, Herrington J, Goodwin PC, Park YB Hille B (1997) Mitochondr ial participation in the intracellular Ca2+ ne twork. J Cell Biol 136:833-844. Barnes JA, Gomes AV (1995) PEST sequences in calmodulin-binding proteins. Mol Cell Biochem 149-150:17-27. Barzilai A, Zilkha-Falb R, Daily D, Stern N, Offen D, Ziv I, Melamed E, Shirvan A (2000) The molecular mechanism of dopamine-induced apopt osis: identification and characterization of genes that mediate dopamine t oxicity. J Neural Transm Suppl:59-76. Beglopoulos V, Montag-Sallaz M, Rohlmann A, Pi echotta K, Ahmad M, Montag D, Missler M (2005) Neurexophilin 3 is highl y localized in cortical a nd cerebellar regions and is functionally important for sensorimotor ga ting and motor coordination. Mol Cell Biol 25:7278-7288. Benaim G, Villalobo A (2002) Phosphorylation of calmodulin. Functional implications. Eur J Biochem 269:3619-3631. Berglund A, Backman L, Shanbhag VP (1986) The 240-kDa subunit of human erythrocyte spectrin binds calmodulin at micromolar ca lcium concentrations. FEBS Lett 201:306-310. Bignone PA, Baines AJ (2003) Spectr in alpha II and beta II isoforms interact with high affinity at the tetramerization site. Biochem J 374:613-624. Birkenfeld J, Betz H, Roth D (2001) Inhibi tion of neurite extension by overexpression of individual domains of LIM ki nase 1. J Neurochem 78:924-927. Blesch A, Grill RJ, Tuszynski MH (1998) Neurot rophin gene therapy in CNS models of trauma and degeneration. Prog Brain Res 117:473-484. Blesch A, Tuszynski MH (2007) Transient growth factor delivery sustains regenerated axons after spinal cord injur y. J Neurosci 27:10535-10545. Boivin P, Galand C (1984) Is spectrin a ca lmodulin-binding protei n? Biochem Int 8:231-236. Bonini S, Rasi G, Bracci-La udiero ML, Procoli A, Aloe L (2003) Nerve growth factor: neurotrophin or cytokine? In t Arch Allergy Immunol 131:80-84. Bouche N, Yellin A, Snedden WA, Fromm H (2004) Plant-Specific Calmodulin-Binding Proteins. Annu Rev Plant Biol. Brabeck C, Beschorner R, Conrad S, Mittelbron n M, Bekure K, Meyermann R, Schluesener HJ, Schwab JM (2004) Lesional expression of R hoA and RhoB following traumatic brain injury in humans. J Neurotrauma 21:697-706. Brabeck C, Mittelbronn M, Bekure K, Meyermann R, Schluesener HJ, Schwab JM (2003) Effect of focal cerebral infarctions on lesional R hoA and RhoB expression. Arch Neurol 60:12451249.
141 Braun H, Schafer K, Hollt V (2002) BetaIII tubulin-expressing neurons reveal enhanced neurogenesis in hippocampal and cortical struct ures after a contusion trauma in rats. J Neurotrauma 19:975-983. Bretin S, Reibel S, Charrier E, Maus-Moatti M, Auvergnon N, Thevenoux A, Glowinski J, Rogemond V, Premont J, Honnorat J, Gauchy C (2005) Differential expression of CRMP1, CRMP2A, CRMP2B, and CRMP5 in axons or dendr ites of distinct neurons in the mouse brain. J Comp Neurol 486:1-17. Bretin S, Rogemond V, Marin P, Maus M, Torre ns Y, Honnorat J, Glowinski J, Premont J, Gauchy C (2006) Calpain product of WT-CRM P2 reduces the amount of surface NR2B NMDA receptor subunit. J Neurochem. Buki A, Okonkwo DO, Wang KK, Povlishock JT (2000) Cytochrome c release and caspase activation in traumatic axonal injury. J Neurosci 20:2825-2834. Buki A, Povlishock JT (2006) All roads lead to disconnection?--Traumatic axonal injury revisited. Acta Neurochir (W ien) 148:181-193; discussion 193-184. Butterfield DA, Boyd-Kimball D, Castegna A (2003) Proteomics in Alzheime r's disease: insights into potential mechanisms of ne urodegeneration. J Neurochem 86:1313-1327. Cai D, Deng K, Mellado W, Lee J, Ratan RR, Filb in MT (2002) Arginase I and polyamines act downstream from cyclic AMP in overcomi ng inhibition of axonal growth MAG and myelin in vitro. Neuron 35:711-719. Carafoli E (2003) Historical review: mitochondr ia and calcium: ups and downs of an unusual relationship. Trends Biochem Sci 28:175-181. Chan CC, Khodarahmi K, Liu J, Sutherland D, Oschipok LW, Steeves JD, Tetzlaff W (2005) Dose-dependent beneficial and detrimenta l effects of ROCK i nhibitor Y27632 on axonal sprouting and functional r ecovery after rat spinal cord injury. Exp Neurol. Charrier E, Reibel S, Rogemond V, Aguera M, Thomasset N, Honnor at J (2003) Collapsin response mediator proteins ( CRMPs): involvement in nervous system development and adult neurodegenerative diso rders. Mol Neurobiol 28:51-64. Chen DF, Schneider GE, Martinou JC, Tonegawa S (1997) Bcl-2 promotes regeneration of severed axons in mammalian CNS. Nature 385:434-439. Chiesa R, Angeretti N, Del Bo R, Lucca E, Mu nna E, Forloni G (1998) Extracellular calcium deprivation in astrocytes: regulation of mRNA expression and a poptosis. J Neurochem 70:1474-1483. Christensen AE, Selheim F, de Rooij J, Dremier S, Schwede F, Dao KK, Martinez A, Maenhaut C, Bos JL, Genieser HG, Doskeland SO (2003) cAMP analog mapping of Epac1 and cAMP kinase. Discriminating analogs demons trate that Epac and cAMP kinase act synergistically to promote PC-12 cell neurite extension. J Biol Chem 278:35394-35402.
142 Chung MA, Lee JE, Lee JY, Ko MJ, Lee ST, Kim HJ (2005) Alteration of collapsin response mediator protein-2 expression in focal is chemic rat brain. Neuroreport 16:1647-1653. Cole AR, Knebel A, Morrice NA, Robertson LA Irving AJ, Connolly CN, Sutherland C (2004) GSK-3 phosphorylation of the Alzheimer epitope within collapsin response mediator proteins regulates axon elongation in primary neurons. J Biol Chem 279:50176-50180. Conrad S, Schluesener HJ, Trautmann K, Joannin N, Meyermann R, Schwab JM (2005) Prolonged lesional expression of RhoA and R hoB following spinal cord injury. J Comp Neurol 487:166-175. Counts SE, Nadeem M, Lad SP, Wuu J, Mufson EJ (2006) Differential expression of synaptic proteins in the frontal and temporal cort ex of elderly subjects with mild cognitive impairment. J Neuropathol Exp Neurol 65:592-601. Cousin MA (2000) Synaptic vesi cle endocytosis: calcium works ove rtime in the nerve terminal. Mol Neurobiol 22:115-128. Craig AM, Kang Y (2007) Neurexin-neuroligin si gnaling in synapse development. Curr Opin Neurobiol 17:43-52. Croall DE, Chacko S, Wang Z (1996) Cleavage of caldesmon and calponin by calpain: substrate recognition is not dependent on calmodulin binding domains. Biochim Biophys Acta 1298:276-284. Czech T, Yang JW, Csaszar E, Kappler J, Baumgartner C, Lubec G (2004) Reduction of hippocampal collapsin response mediated protein2 in patients with mesial temporal lobe epilepsy. Neurochem Res 29:2189-2196. Czogalla A, Sikorski AF (2005) Spec trin and calpain: a 'target' and a 'sniper' in the pathology of neuronal cells. Cell Mol Life Sci 62:1913-1924. Dasgupta M, Honeycutt T, Blumenthal DK ( 1989) The gamma-subunit of skeletal muscle phosphorylase kinase contains two noncontiguous domains that act in concert to bind calmodulin. J Biol Chem 264:17156-17163. Davies SJ, Fitch MT, Memberg SP, Hall AK, Raisma n G, Silver J (1997) Regeneration of adult axons in white matter tracts of the ce ntral nervous system. Nature 390:680-683. Davies SJ, Silver J (1998) Adult axon regenerati on in adult CNS white matter. Trends Neurosci 21:515. Davies SP, Reddy H, Caivano M, Cohen P (2000) Specificity and mechanism of action of some commonly used protein kinase inhibitors. Biochem J 351:95-105. Del Rio JA, Soriano E (2007) Overcoming chondroitin sulphate proteoglycan inhibition of axon growth in the injured brain: lessons from chondroitinase ABC. Curr Pharm Des 13:24852492.
143 Denslow N, Michel ME, Temple MD, Hsu CY, Saatman K, Hayes RL (2003) Application of proteomics technology to the field of neurotrauma. J Neurotrauma 20:401-407. Dent EW, Gertler FB (2003) Cyto skeletal dynamics and transport in growth cone motility and axon guidance. Neuron 40:209-227. Dergham P, Ellezam B, Essagian C, Avedissi an H, Lubell WD, McKerracher L (2002) Rho signaling pathway targeted to promote spin al cord repair. J Neurosci 22:6570-6577. Dietz GP, Dietz B, Bahr M (2006) Bcl-x(L) in creases axonal numbers but not axonal elongation from rat retinal explants. Brain Res Bull 70:117-123. Dixon CE, Clifton GL, Lighthall JW, Yaghmai AA, Hayes RL ( 1991) A controlled cortical impact model of traumatic brain injury in the rat. J Neuros ci Methods 39:253-262. Dudanova I, Sedej S, Ahmad M, Masius H, Sarg syan V, Zhang W, Riedel D, Angenstein F, Schild D, Rupnik M, Missler M (2006) Importa nt contribution of alpha -neurexins to Ca2+triggered exocytosis of secretor y granules. J Neurosci 26:10599-10613. Dutta S, Chiu YC, Probert AW, Wang KK (2002) Selective release of calpain produced alphalIspectrin (alpha-fodrin) breakdown products by acute neuronal cell death. Biol Chem 383:785-791. Ebadi M, Bashir RM, Heidrick ML, Hamada FM Refaey HE, Hamed A, Helal G, Baxi MD, Cerutis DR, Lassi NK (1997) Neurotrophins and th eir receptors in nerve injury and repair. Neurochem Int 30:347-374. Ellezam B, Dubreuil C, Winton M, Loy L, Der gham P, Selles-Navarro I, McKerracher L (2002) Inactivation of intrace llular Rho to stimulate axon grow th and regeneration. Prog Brain Res 137:371-380. Farkas O, Povlishock JT (2007) Cellular and su bcellular change evoked by diffuse traumatic brain injury: a complex web of change exte nding far beyond focal damage. Prog Brain Res 161:43-59. Fellenberg K, Hauser NC, Brors B, Neut zner A, Hoheisel JD, Vingron M (2001) Correspondence analysis applied to microa rray data. Proc Natl Acad Sci U S A 98:1078110786. Fournier AE, Takizawa BT, Strittmatter SM ( 2003) Rho kinase inhibition enhances axonal regeneration in the injured CNS. J Neurosci 23:1416-1423. Fujita A, Hattori Y, Takeuchi T, Kamata Y, Ha ta F (2001) NGF induces neurite outgrowth via a decrease in phosphorylation of myosin light chain in PC12 cells. Neuroreport 12:35993602.
144 Fukada M, Watakabe I, Yuasa-Kawada J, Kaw achi H, Kuroiwa A, Matsuda Y, Noda M (2000) Molecular characterization of CRMP5, a novel me mber of the collapsin response mediator protein family. J Bi ol Chem 275:37957-37965. Fukata Y, Kimura T, Kaibuchi K (2002a) A xon specification in hippocampal neurons. Neurosci Res 43:305-315. Fukata Y, Itoh TJ, Kimura T, Menager C, Nishimura T, Shiromizu T, Watanabe H, Inagaki N, Iwamatsu A, Hotani H, Kaibuchi K (2002b) CRMP-2 binds to tubulin heterodimers to promote microtubule assemb ly. Nat Cell Biol 4:583-591. Fukiage C, Azuma M, Nakamura Y, Tamada Y, Nakamura M, Shearer TR (1997) SJA6017, a newly synthesized peptide alde hyde inhibitor of calpain: am elioration of cataract in cultured rat lenses. Bioc him Biophys Acta 1361:304-312. Gallagher CN, Hutchinson PJ, Pickard JD (2007) Neuroimaging in trauma. Curr Opin Neurol 20:403-409. Geddes JF, Whitwell HL, Graham DI (2000) Trau matic axonal injury: pr actical issues for diagnosis in medicolegal cases. Ne uropathol Appl Ne urobiol 26:105-116. Gerendasy D (1999) Homeostatic tuning of Ca 2+ signal transducti on by members of the calpacitin protein famil y. J Neurosci Res 58:107-119. Glenney JR, Jr., Weber K (1985) Separation of fodrin subunits by affinity chromatography on calmodulin-Sepharose. Anal Biochem 150:364-368. Gold L (2001) mRNA display: diversity matters du ring in vitro selection. Proc Natl Acad Sci U S A 98:4825-4826. Goshima Y, Nakamura F, Strittmatter P, Stritt matter SM (1995) Collapsin-induced growth cone collapse mediated by an intracellular prot ein related to UNC-33. Nature 376:509-514. Goto H, Kosako H, Tanabe K, Yanagida M, Sakurai M, Amano M, Kaibuchi K, Inagaki M (1998) Phosphorylation of vimen tin by Rho-associated kinase at a unique amino-terminal site that is specifically p hosphorylated during cytokinesi s. J Biol Chem 273:11728-11736. Greene LA, Tischler AS (1976) Establishment of a noradrenergic clonal line of rat adrenal pheochromocytoma cells which respond to ne rve growth factor. Proc Natl Acad Sci 73:2424-2428. Gu Y, Hamajima N, Ihara Y (2000) Neurofibri llary tangle-associated collapsin response mediator protein-2 (CRMP2) is highly phosphorylated on Thr-509, Ser-518, and Ser-522. Biochemistry 39:4267-4275. Gu Y, Ihara Y (2000) Evidence that collapsin re sponse mediator protein2 is involved in the dynamics of microtubules. J Biol Chem 275:17917-17920.
145 Gungabissoon RA, Bamburg JR (2003) Regula tion of growth cone actin dynamics by ADF/cofilin. J Histochem Cytochem 51:411-420. Haeseleer F, Palczewski K (2002) Calmodulin and Ca2+-binding proteins (C aBPs): variations on a theme. Adv Exp Med Biol 514:303-317. Hall A (2005) Rho GTPases a nd the control of cell behavi our. Biochem Soc Trans 33:891-895. Hannila SS, Filbin MT (2007) The role of cyclic AMP signaling in promoting axonal regeneration after spinal cord injury. Exp Neurol. Harigaya Y, Shoji M, Shirao T, Hirai S (1996) Disappearance of actin-b inding protein, drebrin, from hippocampal synapses in Alzheimer's disease. J Neurosci Res 43:87-92. Harris MA, Clark J, Ireland A, Lo max J, Ashburner M, Foulger R, Eilbeck K, Lewis S, Marshall B, Mungall C, Richter J, Rubin GM, Blake JA Bult C, Dolan M, Drabkin H, Eppig JT, Hill DP, Ni L, Ringwald M, Balakrishnan R, Cherry JM, Christie KR, Costanzo MC, Dwight SS, Engel S, Fisk DG, Hirschman JE, Hong EL, Nash RS, Sethuraman A, Theesfeld CL, Botstein D, Dolinski K, Feierbach B, Berardini T, Mundodi S, Rhee SY, Apweiler R, Barrell D, Camon E, Dimmer E, Lee V, Chisholm R, Gaudet P, Kibbe W, Kishore R, Schwarz EM, Sternberg P, Gwinn M, Hannick L, Wortman J, Berriman M, Wood V, de la Cruz N, Tonellato P, Jaiswa l P, Seigfried T, White R (2004) The Gene Ontology (GO) database and informatic s resource. Nucleic Acids Res 32:D258-261. Harris AS, Croall DE, Morrow JS (1989) Calmodulin re gulates fodrin susceptibility to cleavage by calcium-dependent protease I. J Biol Chem 264:17401-17408. Hashimoto R, Nakamura Y, Goto H, Wada Y, Sakoda S, Kaibuchi K, Inagaki M, Takeda M (1998) Domainand site-specific phosphoryl ation of bovine NF-L by Rho-associated kinase. Biochem Biophys Res Commun 245:407-411. Hashimoto R, Nakamura Y, Komai S, Kashiwagi Y, Tamura K, Goto T, Aimoto S, Kaibuchi K, Shiosaka S, Takeda M (2000) Site-speci fic phosphorylation of neurofilament-L is mediated by calcium/calmodulin-dependent protei n kinase II in the ap ical dendrites during long-term potentiation. J Neurochem 75:373-382. Hata K, Kubo T, Yamaguchi A, Yamashita T (2006) Signaling mechanisms of axon growth inhibition. Drug News Perspect 19:541-547. Hayes RL, Wang KK, Kampfl A, Posmantur RM Newcomb JK, Clifton GL (1998) Potential contribution of proteases to neuronal damage. Drug News Perspect 11:215-222. Hergan K, Schaefer PW, Sorensen AG, Gonzal ez RG, Huisman TA (2002) Diffusion-weighted MRI in diffuse axonal injury of the brain. Eur Radiol 12:2536-2541. Hill JJ, Callaghan DA, Ding W, Kelly JF, Chak ravarthy BR (2006) Identification of okadaic acid-induced phosphorylation events by a mass spectrometry approach. Biochem Biophys Res Commun 342:791-799.
146 Hou ST, Jiang SX, Desbois A, Huang D, Kelly J, Tessier L, Karchewski L, Kappler J (2006) Calpain-cleaved collapsin response mediator protein-3 induces ne uronal death after glutamate toxicity and cerebral ischemia. J Neurosci 26:2241-2249. Hundle B, McMahon T, Dadgar J, Messing RO (1995) Overexpression of ep silon-protein kinase C enhances nerve growth factor-induced phos phorylation of mitogen-activated protein kinases and neurite outgrow th. J Biol Chem 270:30134-30140. Ikenoya M, Hidaka H, Hosoya T, Suzuki M, Yamamoto N, Sasaki Y (2002) Inhibition of rhokinase-induced myristoylated alanine-rich C kinase substrate (M ARCKS) phosphorylation in human neuronal cells by H-1152, a novel a nd specific Rho-kinase inhibitor. J Neurochem 81:9-16. Inagaki N, Chihara K, Arimura N, Menager C, Kawano Y, Matsuo N, Nishimura T, Amano M, Kaibuchi K (2001) CRMP-2 induces axons in cultured hippocampal neurons. Nat Neurosci 4:781-782. Ishizaki T, Uehata M, Tamechika I, Keel J, Nonomura K, Maekawa M, Narumiya S (2000) Pharmacological properties of Y-27632, a specific inhibitor of rho-associated kinases. Mol Pharmacol 57:976-983. Itano T, Matsui H, Doi A, Ohmura Y, Hata se O (1986) Identificati on of calmodulin-binding proteins in pure mitochondria by photo affinity labeling. Biochem Int 13:787-792. Ito T, Kagoshima M, Sasaki Y, Li C, Udaka N, Kitsukawa T, Fujisawa H, Taniguchi M, Yagi T, Kitamura H, Goshima Y (2000) Repulsive axon guidance molecule Sema3A inhibits branching morphogenesis of fetal mouse lung. Mech Dev 97:35-45. Jiang SX, Kappler J, Zurakowski B, Desbois A, Aylsworth A, Hou ST (2007) Calpain cleavage of collapsin response mediator proteins in ischemic mouse brain. Eur J Neurosci 26:801809. Jiao J, Huang X, Feit-Leithman RA, Neve RL, Snider W, Dartt DA, Chen DF (2005) Bcl-2 enhances Ca(2+) signaling to support the in trinsic regenerative cap acity of CNS axons. Embo J 24:1068-1078. Jin J, Smith FD, Stark C, Wells CD, Fawcett JP, Kulkarni S, Metalnikov P, O'Donnell P, Taylor P, Taylor L, Zougman A, Woodgett JR, Langeberg LK, Scott JD, Pawson T (2004) Proteomic, functional, and domain-based an alysis of in vivo 14-3-3 binding proteins involved in cytoskeletal re gulation and cellular organi zation. Curr Biol 14:1436-1450. Johnson GV, Greenwood JA, Costello AC, Troncos o JC (1991) The regulatory role of calmodulin in the proteolysis of individual neurofilament proteins by calpain. Neurochem Res 16:869-873. Johnston MV (2005) Excitotoxicity in peri natal brain injury. Br ain Pathol 15:234-240.
147 Johnston-Wilson NL, Sims CD, Hofmann JP, A nderson L, Shore AD, Torrey EF, Yolken RH (2000) Disease-specific alterations in frontal cortex brain prot eins in schizophrenia, bipolar disorder, and major depressive disorder. The Stanley Neuropathology Consortium. Mol Psychiatry 5:142-149. Kawano Y, Fukata Y, Oshiro N, Amano M, Nakamura T, Ito M, Matsumura F, Inagaki M, Kaibuchi K (1999) Phosphoryl ation of myosin-binding s ubunit (MBS) of myosin phosphatase by Rho-kinase in vivo. J Cell Biol 147:1023-1038. Kawano Y, Yoshimura T, Tsuboi D, Kawabata S, Kaneko-Kawano T, Shirataki H, Takenawa T, Kaibuchi K (2005) CRMP-2 is involved in kinesin-1-dependent transport of the Sra1/WAVE1 complex and axon formation. Mol Cell Biol 25:9920-9935. Khawaja X, Xu J, Liang JJ, Barrett JE (2004) Pr oteomic analysis of protein changes developing in rat hippocampus after chroni c antidepressant treatment: Im plications for depressive disorders and future therapies. J Neurosci Res 75:451-460. Kimura T, Watanabe H, Iwamatsu A, Kai buchi K (2005) Tubulin and CRMP-2 complex is transported via Kinesi n-1. J Neurochem 93:1371-1382. Kishida S, Yamamoto H, Kikuchi A (2004) Wnt-3a and Dvl induce neurite retraction by activating Rho-associated ki nase. Mol Cell Biol 24:4487-4501. Kitano H (2002) Systems biology: a brief overview. Science 295:1662-1664. Kobeissy FH, Ottens AK, Zhang Z, Liu MC, De nslow ND, Dave JR, Tortella FC, Hayes RL, Wang KK (2006) Novel differential neuroproteomics analysis of traumatic brain injury in rats. Mol Cell Proteomics 5:1887-1898. Kosaki G, Tsujinaka T, Kambayashi J, Morimoto K, Yamamoto K, Yamagami K, Sobue K, Kakiuchi S (1983) Specific cleavage of calmodulin-binding proteins by low Ca2+requiring form of Ca2+-activated neutral pr otease in human platelets. Biochem Int 6:767775. Kosako H, Amano M, Yanagida M, Tanabe K, Nishi Y, Kaibuchi K, Inagaki M (1997) Phosphorylation of glial fibril lary acidic protein at the same sites by cleavage furrow kinase and Rho-associated ki nase. J Biol Chem 272:10333-10336. Kowara R, Chen Q, Milliken M, Chakravart hy B (2005) Calpain-mediated truncation of dihydropyrimidinase-like 3 protein (DPYSL3) in response to NMDA and H2O2 toxicity. J Neurochem 95:466-474. Krucker T, Siggins GR, McNamara RK, Linds ley KA, Dao A, Allison DW, De Lecea L, Lovenberg TW, Sutcliffe JG, Gerendasy DD ( 2002) Targeted disruption of RC3 reveals a calmodulin-based mechanism for regulating me taplasticity in the hippocampus. J Neurosci 22:5525-5535.
148 Kupina NC, Nath R, Bernath EE, Inoue J, Mits uyoshi A, Yuen PW, Wang KK, Hall ED (2001) The novel calpain inhibitor SJA6017 impr oves functional outcome after delayed administration in a mouse model of diffu se brain injury. J Neurotrauma 18:1229-1240. Kurz JE, Parsons JT, Rana A, Gibson CJ, Hamm RJ, Churn SB (2005) A significant increase in both basal and maximal calcineurin activity following fluid perc ussion injury in the rat. J Neurotrauma 22:476-490. Kwon BK, Borisoff JF, Tetzlaff W (2002) Molecula r targets for therapeu tic intervention after spinal cord injury. Mol Intervent 2:244-258. Lakshmikuttyamma A, Selvakumar P, Sharma AR, Anderson DH, Sharma RK (2004) In vitro proteolytic degradation of bovine brain calcineurin by m-calpa in. Neurochem Res 29:1913-1921. Lee S, Kim JH, Lee CS, Kim Y, Heo K, Ihara Y, Goshima Y, Suh PG, Ryu SH (2002) Collapsin response mediator protein-2 inhibits neur onal phospholipase D(2) activity by direct interaction. J Biol Chem 277:6542-6549. Lehmann M, Fournier A, Selles-Navarro I, Dergham P, Sebok A, Leclerc N, Tigyi G, McKerracher L (1999) Inactivation of R ho signaling pathway promotes CNS axon regeneration. J Neurosci 19:7537-7547. Lingor P, Teusch N, Schwarz K, Mueller R, Mack H, Bahr M, Mueller BK (2007) Inhibition of Rho kinase (ROCK) increases neurite outgrow th on chondroitin sulpha te proteoglycan in vitro and axonal regeneration in the adult optic nerve in vivo. J Neurochem 103:181-189. Liu F, Grundke-Iqbal I, Iqbal K, Oda Y, Tomizawa K, Gong CX (2005) Truncation and activation of calcineurin A by calpain I in Alzheimer disease brain. J Biol Chem 280:37755-37762. Liu MC, Akle V, Zheng W, Kitlen J, O'steen B, Larner SF, Dave JR, Tortella FC, Hayes RL, Wang KK (2006) Extensive degradation of my elin basic protein isoforms by calpain following traumatic brain injury. J Neurochem 98:700-712. Liu PC, Yang ZJ, Qiu MH, Zhang LM, Sun FY ( 2003) Induction of CRMP-4 in striatum of adult rat after transient br ain ischemia. Acta Pharmacol Sin 24:1205-1211. Llansola M, Monfort P, Felipo V (2000) I nhibitors of phospholipase C prevent glutamate neurotoxicity in primary cultures of cereb ellar neurons. J Pharmacol Exp Ther 292:870876. Lopez-Otin C, Overall CM (2002) Protease degr adomics: a new challenge for proteomics. Nat Rev Mol Cell Biol 3:509-519.
149 Lopez-Picon FR, Kukko-Lukjanov TK, Holopainen IE (2006) The calpain inhibitor MDL-28170 and the AMPA/KA receptor antagonist CNQX inhibit neurofilament degradation and enhance neuronal survival in ka inic acid-treated hippocampal sl ice cultures. Eur J Neurosci 23:2686-2694. Luk SC, Ngai SM, Tsui SK, Fung KP, Lee CY Waye MM (1999) In vivo and in vitro association of 14-3-3 epsilon isoform with calmodulin: implication for signal transduction and cell proliferation. J Cell Biochem 73:31-35. Maekawa M, Ishizaki T, Boku S, Watanabe N, Fu jita A, Iwamatsu A, Obinata T, Ohashi K, Mizuno K, Narumiya S (1999) Signaling fr om Rho to the actin cytoskeleton through protein kinases ROCK and LI M-kinase. Science 285:895-898. Markgraf CG, Velayo NL, Johnson MP, McCarty DR, Medhi S, Koehl JR, Chmielewski PA, Linnik MD (1998) Six-hour window of opportunity for calpain inhibiti on in focal cerebral ischemia in rats. Stroke 29:152-158. Matsuzawa K, Kosako H, Azuma I, Inagaki N, Inagaki M (1998) Possible regulation of intermediate filament proteins by Rhobinding kinases. Subcell Biochem 31:423-435. McGinnis KM, Wang KK, Gnegy ME (1999) Alterations of extracellular cal cium elicit selective modes of cell death and prot ease activation in SH-SY5Y hu man neuroblastoma cells. J Neurochem 72:1853-1863. McGinnis KM, Whitton MM, Gnegy ME, Wang KK W (1998) Calcium/calmodulin-dependent protein kinase IV is cleaved by caspase-3 and calpain in SH-SY5 Y human neuroblastoma cells undergoing apoptosis. J Biol Chem 273:19993-20000. McKerracher L, David S (2004) Easing the brak es on spinal cord repair. Nat Med 10:1052-1053. McKerracher L, David S, Jackson DL, Kottis V, Dunn RJ, Braun PE (1994) Identification of myelin-associated glycoprotein as a major my elin-derived inhibito r of neurite growth. Neuron 13:805-811. Medana IM, Esiri MM (2003) Axonal damage: a key predictor of outcome in human CNS diseases. Brain 126:515-530. Mimura F, Yamagishi S, Arimura N, Fujitani M, Kubo T, Kaibuchi K, Yamashita T (2006) Myelin-associated glycoprotein inhibits micr otubule assembly by a Rho-kinase-dependent mechanism. J Biol Chem 281:15970-15979. Mizuno N, Yoshitomi H, Ishida H, Kuromi H, Ka waki J, Seino Y, Sein o S (1998) Altered bcl-2 and bax expression and intracellu lar Ca2+ signaling in apoptosis of pancreatic cells and the impairment of glucose-induced in sulin secretion. Endocrinology 139:1429-1439. Monnier PP, Sierra A, Schwab JM, Henke-F ahle S, Mueller BK (2003a) The Rho/ROCK pathway mediates neurite growth-inhibitory acti vity associated with the chondroitin sulfate proteoglycans of the CNS glial s car. Mol Cell Neurosci 22:319-330.
150 Monnier PP, Sierra A, Schwab JM, Henke-F ahle S, Mueller BK (2003b) The Rho/ROCK pathway mediates neurite growth-inhibitory acti vity associated with the chondroitin sulfate proteoglycans of the CNS glial s car. Mol Cell Neurosci 22:319-330. Mueller BK, Mack H, Teusch N (2005) Rho kina se, a promising drug target for neurological disorders. Nat Rev Drug Discov 4:387-398. Mukerjee N, McGinnis KM, Gnegy ME, Wang KKW (2001) Caspase-mediated calcineurin activation contributes to IL-2 release during T cell activation. Biochem Biophys Res Commun 285:1192-1199. Mukerjee N, McGinnis KM, Park YH, Gne gy ME, Wang KKW (2000) Caspase-mediated proteolytic activation of calci neurin in thapsigargin-mediated apoptosis in SH-SY5Y neuroblastoma cells. Arch Biochem Biophys 379:337-343. Nakajima M, Hayashi K, Egi Y, Katayama K, Ama no Y, Uehata M, Ohtsuki M, Fujii A, Oshita K, Kataoka H, Chiba K, Goto N, Kondo T (2003) Effect of Wf-536, a novel ROCK inhibitor, against metastasis of B16 me lanoma. Cancer Chemother Pharmacol 52:319-324. Nath R, Davis M, Probert AW, Kupina NC, Ren X, Schielke GP, Wang KK (2000a) Processing of cdk5 activator p35 to its tr uncated form (p25) by calpain in acutely injured neuronal cells. Biochem Biophys Res Commun 274:16-21. Nath R, Huggins M, Glantz SB, Morrow JS McGinnis K, Nadimpalli R, Wanga KK (2000b) Development and characterizati on of antibodies specific to caspase-3-produced alpha IIspectrin 120 kDa breakdown product: marker for neuronal apoptosis. Neurochem Int 37:351-361. Newcombe VF, Williams GB, Nortje J, Bradle y PG, Harding SG, Smielewski P, Coles JP, Maiya B, Gillard JH, Hutchinson PJ, Pi ckard JD, Carpenter TA, Menon DK (2007) Analysis of acute traumatic axonal injury us ing diffusion tensor im aging. Br J Neurosurg 21:340-348. Ng J, Luo L (2004) Rho GTPases regulate a xon growth through convergent and divergent signaling pathways. Neuron 44:779-793. Nishimura T, Fukata Y, Kato K, Yamaguchi T, Matsuura Y, Kamiguchi H, Kaibuchi K (2003) CRMP-2 regulates polarized Numb-mediated e ndocytosis for axon growth. Nat Cell Biol 5:819-826. Niwa R, Nagata-Ohashi K, Takeichi M, Mi zuno K, Uemura T (2002) Control of actin reorganization by Slingshot, a family of phosphatases that dephosphorylate ADF/cofilin. Cell 108:233-246. Obara Y, Aoki T, Kusano M, Ohizumi Y (2002) Be ta-eudesmol induces neurite outgrowth in rat pheochromocytoma cells accompanied by an activation of mitogen-activated protein kinase. J Pharmacol Exp Ther 301:803-811.
151 Oda K, Matsuoka Y, Funahashi A, Kitano H (2005) A comprehensive pathway map of epidermal growth factor receptor si gnaling. Mol Syst Biol 1:2005 0010. O'Day DH (2003) CaMBOT: profiling and char acterizing calmodulin-binding proteins. Cell Signal 15:347-354. O'Day DH, Myre MA (2004) Calmodulin-binding domains in Alzheimer's disease proteins: extending the calcium hypothesis. Bi ochem Biophys Res Commun 320:1051-1054. Ohashi K, Nagata K, Maekawa M, Ishizaki T, Narumiya S, Mizuno K (2000) Rho-associated kinase ROCK activates LIMkinase 1 by phosphorylation at threonine 508 within the activation loop. J Biol Chem 275:3577-3582. Ottens AK, Kobeissy FH, Golden EC, Zhang Z, Haskins WE, Chen SS, Hayes RL, Wang KK, Denslow ND (2006) Neuroproteomics in ne urotrauma. Mass Spectrom Rev 25:380-408. Ottens AK, Kobeissy FH, Wolper RA, Hask ins WE, Hayes RL, De nslow ND, Wang KKW (2005) A multidimensional differential proteomic platform using dual-phase ion-exchange chromatography polyacrylamide gel el ectrophoresis/reversed-phase liquid chromatography tandem mass spect rometry. Anal Chem 77:4836-4845. Ozdinler PH, Erzurumlu RS (2001) Regulation of neurotrophin-induced axonal responses via Rho GTPases. J Comp Neurol 438:377-387. Park YH, Kantor L, Guptaroy B, Zhang M, Wang KK, Gnegy ME (2003) Repeated amphetamine treatment induces neurite outgr owth and enhanced amphetamine-stimulated dopamine release in rat pheochromocytoma cells (PC12 cells) via a pr otein kinase Cand mitogen activated protein kinase-depe ndent mechanism. J Neurochem 87:1546-1557. Peri S, Navarro JD, Amanchy R, Kristiansen TZ, Jonnalagadda CK, Surendranath V, Niranjan V, Muthusamy B, Gandhi TK, Gronborg M, Ibar rola N, Deshpande N, Shanker K, Shivashankar HN, Rashmi BP, Ramya MA, Zhao Z, Chandrika KN, Padma N, Harsha HC, Yatish AJ, Kavitha MP, Menezes M, Choudhury DR, Suresh S, Ghosh N, Saravana R, Chandran S, Krishna S, Joy M, Anand SK, Madavan V, Joseph A, Wong GW, Schiemann WP, Constantinescu SN, Huang L, Khosravi-Far R, Steen H, Tewari M, Ghaffari S, Blobe GC, Dang CV, Garcia JG, Pevsner J, Jensen ON, Roepstorff P, Deshpande KS, Chinnaiyan AM, Hamosh A, Chakravarti A, Pandey A (2003) Development of human protein reference database as an initial platform for approaching systems biology in humans. Genome Res 13:2363-2371. Peterson BZ, DeMaria CD, Adelman JP, Yue DT (1999) Calmodulin is the Ca2+ sensor for Ca2+ -dependent inactivation of L-t ype calcium channels. Neuron 22:549-558. Petruska JC, Mendell LM (2004) The many functions of nerve growth factor: multiple actions on nociceptors. Neuros ci Lett 361:168-171.
152 Pike BR, Flint J, Dutta S, Johnson E, Wang KKW, Hayes RL (2001) Accumulation of nonerythroid alpha II-spectrin and calpain-clea ved alpha II-spectrin breakdown products in cerebrospinal fluid after traumatic brai n injury in rats. J Neurochem 78:1297-1306. Pike BR, Zhao X, Newcomb JK, Posmantur RM Wang KK, Hayes RL (1998) Regional calpain and caspase-3 proteolysis of alpha-spectrin after traumatic brain injury. Neuroreport 9:2437-2442. Pike BR, Flint J, Dave JR, Lu XC, Wang KKW, Tortella FC, Hayes RL (2004) Accumulation of calpain and caspase-3 pr oteolytic fragments of brain-deri ved alphaII-spectrin in cerebral spinal fluid after middle cerebral artery o cclusion in rats. J Cereb Blood Flow Metab 24:98-106. Pineda JA, Wang KKW, Hayes RL (2004) Biom arkers of proteolytic damage following traumatic brain injury. Brain Pathol 14:202-209. Pohorecki R, Becker GL, Reilly PJ, Landers DF (1990) Ischemic brain injury in vitro: protective effects of NMDA receptor antagonists and calmidazolium. Brain Res 528:133-137. Povlishock JT, Buki A, Koiziumi H, Stone J, Okonkwo DO (1999) Initiating mechanisms involved in the pathobiology of traumatically induced axonal injury and interventions targeted at blunting their progres sion. Acta Neurochir Suppl 73:15-20. Quach TT, Duchemin AM, Rogemond V, Aguera M, Honnorat J, Be lin MF, Kolattukudy PE (2004) Involvement of collapsin response mediator proteins in the neurite extension induced by neurotrophins in dorsal root ga nglion neurons. Mol Cell Neurosci 25:433-443. Reddy VS, Ali GS, Reddy AS (2002) Genes enc oding calmodulin-binding proteins in the Arabidopsis genome. J Biol Chem 277:9840-9852. Revenu C, Athman R, Robine S, Louvard D (2004) The co-workers of actin filaments: from cell structures to signals. Nat Rev Mol Cell Biol 5:635-646. Riento K, Ridley AJ (2003) Rocks: multifunctional kinases in cell behaviour. Nat Rev Mol Cell Biol 4:446-456. Riss TL, Moravec RA (2004) Use of multiple as say endpoints to investigate the effects of incubation time, dose of toxin, and plating dens ity in cell-based cytotoxicity assays. Assay Drug Dev Technol 2:51-62. Ruben L, Goodman DB, Rasmussen H (1980) Identif ication of calmodulin in mitochondria from rat liver: a possible role in regulation of th e oligomycin sensitive ATPase. Ann N Y Acad Sci 356:427-428. Sahuquillo J, Poca MA, Amoros S (2001) Curre nt aspects of pathophysiology and cell dysfunction after severe head injury. Curr Pharm Des 7:1475-1503.
153 Saris NE, Carafoli E (2005) A historical review of cellular calcium handling, with emphasis on mitochondria. Biochemistry (Mosc) 70:187-194. Sasaki Y, Suzuki M, Hidaka H (2002) The nove l and specific Rho-kinase inhibitor (S)-(+)-2methyl-1-[(4-methyl-5-isoquinoline)sulfonyl]-hom opiperazine as a probing molecule for Rho-kinase-involved pathway. Pharmacol Ther 93:225-232. Schwartz M (2004) Rho signalling at a glance. J Cell Sci 117:5457-5458. Schweitzer J, Becker CG, Schachner M, Beck er T (2005) Expression of collapsin response mediator proteins in the nervous system of embryonic zebrafish. Gene Expr Patterns 5:809-816. Sebok A, Nusser N, Debreceni B, Guo Z, Santos MF, Szeberenyi J, Tigyi G (1999) Different roles for RhoA during neurite initiation, elongation, and regeneration in PC12 cells. J Neurochem 73:949-960. Selzer ME (2003) Promotion of axonal regeneration in the inju red CNS. Lancet Neurol 2:157166. Shen X, Valencia CA, Szostak JW, Dong B, Liu R (2005) Scanning the human proteome for calmodulin-binding proteins. Proc Natl Acad Sci U S A 102:5969-5974. Shevchenko A, Schaft D, Roguev A, Pijnappe l WW, Stewart AF (2002) Deciphering protein complexes and protein interaction networks by tandem affinity purification and mass spectrometry: analytical perspec tive. Mol Cell Proteomics 1:204-212. Shewan D, Dwivedy A, Anderson R, Holt CE ( 2002) Age-related changes underlie switch in netrin-1 responsiveness as growth cones advance along visual pathway. Nat Neurosci 5:955-962. Smillie KJ, Cousin MA (2005) Dynamin I phosphoryl ation and the control of synaptic vesicle endocytosis. Biochem Soc Symp:87-97. Smith DH, Meaney DF, Shull WH (2003) Diffuse a xonal injury in head trauma. J Head Trauma Rehabil 18:307-316. Sola C, Barron S, Tusell JM, Serratosa J (1999) The Ca2+/calmodulin signaling system in the neural response to excitab ility. Involvement of neuronal and glial cells. Prog Neurobiol 58:207-232. Somlyo AP, Somlyo AV (2000) Signal transduc tion by G-proteins, rhokinase and protein phosphatase to smooth muscle and non-muscle myosin II. J Physiol 522 Pt 2:177-185. Stenmark P, Ogg D, Flodin S, Flores A, Ko tenyova T, Nyman T, Nordlund P, Kursula P (2007) The structure of human collapsin response me diator protein 2, a regulator of axonal growth. J Neurochem 101:906-917.
154 Tabb DL, McDonald WH, Yates JR, 3rd (2002) DT ASelect and Contrast: tools for assembling and comparing protein identifications from shotgun proteomics. J Proteome Res 1:21-26. Tahimic CG, Tomimatsu N, Nishigaki R, Fukuhara A, Toda T, Kaibuchi K, Shiota G, Oshimura M, Kurimasa A (2006) Evidence for a role of Collapsin response mediator protein-2 in signaling pathways that regulat e the proliferation of non-ne uronal cells. Biochem Biophys Res Commun 340:1244-1250. Takahashi TT, Austin RJ, Roberts RW (2003) mRNA display: ligand discovery, interaction analysis and beyond. Trends Biochem Sci 28:159-165. Takei K, Yoshida Y, Yamada H (2005) Re gulatory mechanisms of dynamin-dependent endocytosis. J Biochem (Tokyo) 137:243-247. Tang BL (2003) Inhibitors of neuronal regeneration: mediators and signaling mechanisms. Neurochem Int 42:189-203. Thienpont LM, Van Uytfanghe K, Marriot J, Stokes P, Siekmann L, Kessler A, Bunk D, Tai S (2005) Metrologic traceability of total thyroxine measurements in human serum: efforts to establish a network of reference measur ement laboratories. Clin Chem 51:161-168. Tojima T, Ito E (2004) Signal transduction ca scades underlying de novo protein synthesis required for neuronal morphogenesis in diffe rentiating neurons. Prog Neurobiol 72:183193. Tompa P, Buzder-Lantos P, Tantos A, Farkas A, Szilagyi A, Banoczi Z, Hudecz F, Friedrich P (2004) On the sequential determinants of calpain cleavage. J Bi ol Chem 279:20775-20785. Tsunekawa S, Takahashi K, Abe M, Hiwada K, Ozawa K, Murachi T (1989) Calpain proteolysis of free and bound forms of calponin, a troponin T-like protein in smooth muscle. FEBS Lett 250:493-496. Uchida Y, Ohshima T, Sasaki Y, Suzuki H, Yanai S, Yamashita N, Nakamura F, Takei K, Ihara Y, Mikoshiba K, Kolattukudy P, Honnorat J, Goshima Y (2005) Semaphorin3A signalling is mediated via sequential Cdk5 and GSK3be ta phosphorylation of CRMP2: implication of common phosphorylating mechanism underlying axon guidance and Alzheimer's disease. Genes Cells 10:165-179. van Dalen JJ, Gerendasy DD, de Graan PN, Sc hrama LH, Gruol DL (2003) Calcium dynamics are altered in cortical neurons lacking the ca lmodulin-binding protein RC3. Eur J Neurosci 18:13-22. Verma DP, Hong Z (2005) The ins and outs in me mbrane dynamics: tubulation and vesiculation. Trends Plant Sci 10:159-165. Wahl S, Barth H, Ciossek T, Aktories K, Mu eller BK (2000) Ephrin-A5 induces collapse of growth cones by activating Rho and Rho kinase. J Cell Biol 149:263-270.
155 Walmsley AR, Mir AK (2007) Targeting the NogoA signalling pathway to promote recovery following acute CNS injury. Curr Pharm Des 13:2470-2484. Wang KC, Koprivica V, Kim JA, Sivasanka ran R, Guo Y, Neve RL, He Z (2002) Oligodendrocyte-myelin glycopr otein is a Nogo receptor ligan d that inhibits neurite outgrowth. Nature 417:941-944. Wang KK (2000a) Calpain and caspase: can you te ll the difference? Trends Neurosci 23:20-26. Wang KK, Nath R, Raser KJ, Hajimohammad reza I (1996) Maitotoxin induces calpain activation in SH-SY5Y neuroblastoma cells a nd cerebrocortical cult ures. Arch Biochem Biophys 331:208-214. Wang KK, Ottens A, Haskins W, Liu MC, Kobei ssy F, Denslow N, Chen S, Hayes RL (2004) Proteomics studies of traumatic brai n injury. Int Rev Neurobiol 61:215-240. Wang KK, Villalobo A, Roufogalis BD (1989a) Calmodulin-binding proteins as calpain substrates. Biochem J 262:693-706. Wang KKW (2000b) Calpain and caspase: can you tell the difference? Trends Neurosci 23:2026. Wang KKW, Villalobo A, Roufogalis BD (1989b) Calmodulin-binding proteins as calpain substrates. Biochem J 262:693-706. Wang LH, Strittmatter SM (1996) A family of rat CRMP genes is differentially expressed in the nervous system. J Neurosci 16:6197-6207. Wang LH, Strittmatter SM (1997) Brain CRMP forms heterotetramers similar to liver dihydropyrimidinase. J Neurochem 69:2261-2269. Waterhouse N, Kumar S, Song Q, Strike P, Sparrow L, Dreyfuss G, Alnemri ES, Litwack G, Lavin M, Watters D (1996) Heteronuclear ri bonucleoproteins C1 and C2, components of the spliceosome, are specific targets of inte rleukin 1beta-converting enzyme-like proteases in apoptosis. J Biol Chem 271:29335-29341. Weitzdoerfer R, Fountoulakis M, Lubec G (2001) Aberrant expression of dihydropyrimidinase related proteins-2,-3 and -4 in fetal Down syndrome brain. J Neural Transm Suppl:95-107. Xia Z, Choi EJ, Wang F, Blazyns ki C, Storm DR (1993) Type I calmodulin-sensitive adenylyl cyclase is neural specific. J Neurochem 60:305-311. Xia Z, Storm DR (2005) The role of calmodulin as a signal integrator for synaptic plasticity. Nat Rev Neurosci 6:267-276. Xie F, Zheng B (2007) White matter inhibitors in CNS axon regeneration failure. Exp Neurol.
156 Yamaguchi Y, Katoh H, Yasui H, Mori K, Negi shi M (2001) RhoA inhibits the nerve growth factor-induced Rac1 activation through Rho-associated kinase -dependent pathway. J Biol Chem 276:18977-18983. Yamamoto H, Fukunaga K, Goto S, Tanaka E, Miyamoto E (1985) Ca2+, calmodulin-dependent regulation of microtubule formation via phospho rylation of microtubuleassociated protein 2, tau factor, and tubulin, and compar ison with the cycl ic AMP-dependent phosphorylation. J Neurochem 44:759-768. Yamashima T (2004) Ca2+-dependent proteases in ischemic neuronal death: a conserved 'calpain-cathepsin cascade' from nema todes to primates. Cell Calcium 36:285-293. Yamashita T, Fujitani M, Yamagishi S, Hata K, Mimura F (2005) Multiple signals regulate axon regeneration through the nogo receptor complex. Mol Neurobiol 32:105-112. Yamashita T, Higuchi H, Tohyama M (2002) The p75 receptor transduces the signal from myelin-associated glycoprotein to Rho. J Cell Biol 157:565-570. Yap KL, Kim J, Truong K, Sherman M, Yuan T, I kura M (2000) Calmodulin target database. J Struct Funct Genomics 1:8-14. Ybe JA, Brodsky FM, Hofmann K, Lin K, Liu SH Chen L, Earnest TN, Fletterick RJ, Hwang PK (1999) Clathrin self-assembly is mediated by a tandemly repeated superhelix. Nature 399:371-375. Yoshida H, Watanabe A, Ihara Y (1998) Collaps in response mediator pr otein-2 is associated with neurofibrillary tangles in Alzhei mer's disease. J Biol Chem 273:9761-9768. Yoshimura T, Kawano Y, Arimura N, Kawabata S, Kikuchi A, Kaibuc hi K (2005) GSK-3beta regulates phosphorylation of CRMP-2 and neuronal polarity. Cell 120:137-149. Yuasa-Kawada J, Suzuki R, Kano F, Ohka wara T, Murata M, Noda M (2003) Axonal morphogenesis controlled by antagonistic roles of two CRMP subtypes in microtubule organization. Eur J Neurosci 17:2329-2343. Yukawa K, Tanaka T, Bai T, Ueyama T, Owa da-Makabe K, Tsubota Y, Maeda M, Suzuki K, Kikutani H, Kumanogoh A (2005) Semaphorin 4A induces growth cone collapse of hippocampal neurons in a Rho/Rho-kinase-d ependent manner. Int J Mol Med 16:115-118. Zhang S, Ehlers MD, Bernhardt JP, Su CT, Huga nir RL (1998) Calmodulin mediates calciumdependent inactivation of N-methyl-D -aspartate receptors. Neuron 21:443-453. Zhang Z, Ottens AK, Golden EC, Hayes RL, Wang KK (2006a) Using Calmodulin-Affinity Capture to Study the Rat Brain Calmodulin Bi nding Proteome and Its Vulnerability to Calpain and Caspase Proteoly sis. J Neurotrauma 1:125-129.
157 Zhang Z, Ottens AK, Larner SF, Kobeissy FH, Williams ML, Hayes RL, Wang KK (2006b) Direct Rho-associated kinase inhibiton i nduces cofilin dephosphorylation and neurite outgrowth in PC-12 cells. Cell Mol Biol Lett 11:12-29. Zhang Z, Ottens AK, Sadasivan S, Kobeissy FH, Fang T, Hayes RL, Wang KK (2007) Calpainmediated collapsin response mediator protein1, -2, and -4 proteolysi s after neurotoxic and traumatic brain injury. J Neurotrauma 24:460-472. Zhao XH, Jin WL, Ju G (2007) An in vitro study on the involvement of LINGO-1 and Rho GTPases in Nogo-A regulated differentiation of oligodendrocyte precu rsor cells. Mol Cell Neurosci 36:260-269. Zheng B, Ho C, Li S, Keirstead H, Steward O, Tessier-Lavigne M (2003) Lack of enhanced spinal regeneration in Nogo-de ficient mice. Neuron 38:213-224. Zheng B, Lee JK, Xie F (2006) Genetic mouse m odels for studying inhibitors of spinal axon regeneration. Trends Neurosci 29:640-646. Zhou Y, Su Y, Li B, Liu F, Ryder JW, Wu X, Gonzalez-DeWhitt PA, Gelfanova V, Hale JE, May PC, Paul SM, Ni B (2003) Nonsteroid al anti-inflammatory drugs can lower amyloidogenic Abeta42 by inhi biting Rho. Science 302:1215-1217. Zhu H, Bilgin M, Bangham R, Hall D, Casamayor A, Bertone P, Lan N, Jansen R, Bidlingmaier S, Houfek T, Mitchell T, Miller P, Dean RA Gerstein M, Snyder M (2001) Global analysis of protein activities using prot eome chips. Science 293:2101-2105.
158 BIOGRAPHICAL SKETCH Zhiqun Zhang was born in 1975 in Linxiang, mi d-southern of China. She received a Bachelor of Science degree in medicine from Nanjing University (Nanjing, China) in 1997. In 1999 she received a Master of Science degree in medicine from the same university. After graduation, she moved to Suzhou and became a phys ician in the Department of Nephrology at SuZhou Medical School Hospital. In 2003, she jo ined the Interdisciplinary Program in Biomedical Sciences at the UF College of Medicine. From 2004, she began her doctoral study under the guidance of Dr. Kevin K. W. Wang in the Departments of Psychiatry and Neuroscience. Under the guidance of her me ntor Dr. Kevin K. W. Wang, Zhiqun earned a number of research honors including the Outs tanding Research Award from the College of Medicine, and the Bryan W. Robinson Neurolog ical Foundation Hornable Mention Achievement Award (2006). She also presented her research work at the National Institute of Mental Health Research Festival in 2006. During her research st udies, she was the first author of four published journal articles, one of which was featured in an editorial commentary in the journal Calcium Binding Proteins.