P71 EXPRESSION OF THE CYSTINE/GLUTAMATE EXCHANGER IN THE RAT BRAIN
La Bella V, Valentino F & Piccoli F
Laboratory of Neurochemistry, Di.N.O.O.P., University of Palermo, Palermo, Italy
E‐mail address for correspondence: [email protected]
Background: The cystine/glutamate exchanger (antiporter xc−) has been suggested to be involved in slow oxidative excitotoxity and in the effects of beta‐N‐oxalylamino‐L‐alanine, the molecule responsible for neurolathyrism, a neurotoxic upper motor neuron disease. The xc− antiporter is involved in the intracellular transport of cystine, the rate‐limiting amino acid in the synthesis of gluthatione, an important anti‐oxidant molecule. The mouse cystine/glutamate exchanger has been cloned and shown to be composed of two distinct proteins. One of the two subunits is a novel protein, named xCT, of 502 amino acids with 12 putative trans‐membrane domains.
Objective: To study the expression of the xCT protein in the rat brain and spinal cord.
Methods: We have generated polyclonal antibodies to the xCT subunit of the mouse cystine/glutamate exchanger. The antibody was purified on an affinity column. Dot blot and Western blot analyses were performed on protein extracts from rat brain and other tissues using standard techniques. Expression of xCT was studied in several rat brain areas and spinal cord, and in different cultured cells (astrocytes, fibroblasts and neurons) using Western blot and immunocytochemical techniques. Expression of xCT was also studied in several rat tissues (lung, heart, muscle, liver, brain, spinal cord) at different developmental stages (E16, P0, P15, one month, adult).
Results: Peptide pre‐adsorption and dot‐blot experiments allowed verification of the antibody specificity. Western blot analysis of protein extracts showed that the xCT antibody stained two strong bands of about 38 and 50 kDa in all tissues and cells examined. The two isoforms are expressed in embryonic tissues and they appear to increase during development. Subcellular fractionation indicated that xCT isoforms segregate mainly in the microsomal‐mitochondrial fraction. Immunocytochemical analysis showed also a strong staining in neurons and in other cells.
Discussion: We have demonstrated that the xCT protein is expressed in the rat brain and spinal cord as two major isoforms of 38 kDa and 50 kDa, and that its expression is developmentally regulated in all tissues examined. The protein is membrane‐bound and also shows a significant expression in embryonic cultured neurons and astrocytes. This represents direct evidence that the xc− antiporter might represent an important regulator of the redox processes involving gluthatione in these cells.
The characterization of the biochemical properties and function of the xCT protein in the central nervous system may help to understand the contribution of the cystine/glutamate antiporter exchanger in the pathogenesis of certain motor neuron diseases.
P72 DANTROLENE PROTECTS MOTOR NEURONS IN CULTURE AGAINST AMPA RECEPTOR‐MEDIATED EXCITOTOXICITY
Haastert K, Jahn K, Dengler R, Grothe C, Bufler J & Grosskreutz J
Medical School Hannover, Hannover, Germany
E‐mail address for correspondence: [email protected]
Background: Chronic excitotoxicity induced by over‐stimulation of calcium‐permeable AMPA receptors (AMPAR) was shown to be a key factor in the pathogenesis of amyotrophic lateral sclerosis (ALS). The presence of AMPAR, however, is also necessary for neuronal development. AMPAR may exert such bimodal influence on neuronal survival by altering intracellular calcium homeostasis through control of calcium‐induced calcium release from the endoplasmic reticulum (ER).
Objective: The objective of this investigation was to determine a protective effect of the selective ryanodine receptor blocker dantrolene against AMPA/kainate‐induced excitotoxicity in cultured motor neurons.
Methods: Ventral horn (VH) neurons and dorsal horn (DH) neurons harvested from E14 rat lumbar spinal cord were cultured for 13 days and then incubated for 24 h with either kainate (30 µM), dantrolene (30 µM) or a combination of both (each 30 µM). SMI32‐antibodies against the non‐phosphorylated neurofilament‐H/M, and tubulin‐βIII‐ antibodies were used to estimate the number of motor neurons and total neuron counts (mean±S.E.M.).
Results: Total neuron counts remained equal in each treatment condition (240±17 VH, n = 36; 438±35.2 DH, n = 36). In VH control condition, the percentage of motor neurons in relation to total neuron count was 27%±3% (n = 9). Incubation with kainate significantly reduced the percentage of motor neurons to 14%±2% (n = 8), p<0.01. Simultaneous application of kainate and dantrolene significantly recovered the percentage of motor neurons to 27%±3% (n = 9, p<0.01). Dantrolene alone increased the percentage of motor neurons present to 40%±7% (n = 7, p<0.05). This represents a significant increase compared to the kainate condition (p<0.01) and a trend compared to the control condition (p = 0.07). The DH preparations contained a low percentage of motor neurons (5.9%±2.1%, n = 12) where no significant changes due to KA and dantrolene treatment were observed.
Conclusion: Calcium‐permeable AMPAR activate ryanodine receptors of the endoplasmic reticulum which causes a spread of the calcium signal along the surface of the ER. The resulting calcium‐induced calcium release (CICR) leads to a decrease of ER calcium concentration and to calcium uptake into mitochondria, thereby controlling protein folding and energy metabolism. In our preparation, dantrolene exerted a significant neuroprotective effect against kainite‐induced excitotoxicity and in kainate‐free conditions where physiological AMPAR activation can contribute to motor neuron death. Dantrolene may positively influence excitotoxic motor neuron degeneration by lowering energy metabolism and protein folding rates, which seem to be pathologically increased in ALS.
P73 NETWORKED SYNAPTIC AMPAR ACTIVATION AND CALCIUM HOMEOSTASIS IN CULTURED MOTOR NEURONS
Grosskreutz J, Jahn K, Haastert K, Dengler R, Grothe C & Bufler J
Medical School Hannover, Hannover, Germany
E‐mail address for correspondence: [email protected]
Background: AMPA type glutamate receptors (AMPAR) are involved in neuronal development and provide fast excitatory synaptic transmission, but they also mediate cytotoxic insults. Motor neurons are selectively vulnerable towards AMPAR‐mediated excitotoxicity, which may play a role in the pathogenesis of amyotrophic lateral sclerosis (ALS). Over‐stimulation of AMPAR by non‐desensitizing agonists leads to an alteration of intracellular calcium homeostasis in cell culture.
Objective: The objective of this study was to investigate calcium homeostasis in motor neurons under physiological stimulation of AMPAR. Motor neuron‐specific disturbances of calcium signals due to physiological AMPAR activation were compared with those of other types of neurons using calcium imaging techniques and patch‐clamp recordings.
Methods: Motor neurons (ventral horn) and dorsal horn neurons were harvested from E 14 rat spinal cord and differentiated in a co‐culture system with neonatal Schwann cells. Simultaneous patch‐clamp recordings at −80 mV membrane potential and high spatiotemporal resolution FURA‐2 calcium imaging was used to monitor membrane currents and cytosolic calcium transients in individual neurons.
Results: In cultured ventral horn and dorsal horn neurons, spontaneous inward currents and simultaneous whole cell cytosolic fast calcium transients were observed from DIV 10 onwards, indicating a functional synaptic neuronal network. The calcium signal activity of groups of neurons was tightly synchronized. Inward currents triggered cytoplasmic calcium transients when they exceeded a threshold of ∼0.5 nA. These spontaneous inward currents and the corresponding (within 50 ms) calcium transients were completely eliminated by the specific AMPAR blocker CNQX. In contrast, the NMDAR blocker MK‐ 801 had no effect.
Conclusion: The spontaneous activity occurring under physiological conditions in cultured motor neurons and dorsal horn neurons is mediated via AMPA receptors and can therefore be used as a model for physiological AMPA receptor activation. Activity was synchronized, indicating the presence of pacemaker cells. Cytosolic calcium transients were triggered by AMPAR inward currents and spread throughout the whole neuron within ∼100 ms, indicating a fast calcium‐induced calcium release (CICR) from intracellular stores, e.g. the endoplasmic reticulum. The tight coupling between AMPAR activation and CICR may represent a molecular link between AMPAR input, energy metabolism and protein folding which are thought to be disturbed in the pathogenesis of ALS.
P74 ESTABLISHING A CELL MODEL TO STUDY DOWN‐REGULATION OF THE GLUTAMATE TRANSPORTER EAAT2 BY SOD1G93A
Peacey EE, Williams RJ & Rattray M
King's College London, London, UK
E‐mail address for correspondence: [email protected]
Background: In amyotrophic lateral sclerosis, excessive activation of post‐synaptic glutamate receptors may cause motor neuron degeneration through excitotoxicity. Loss of the glial glutamate transporter EAAT2 is a consistent observation in ALS patients and in transgenic rodent models, and EAAT2 down‐regulation may contribute to motor neuron pathology. We have previously shown that SOD1G93A causes down‐regulation of EAAT2 in primary cultures of mouse astrocytes. A recent study by Sala et al (1) demonstrated that a neuronal‐like SH‐SY5Y cell line expressed EAAT2 and showed a down‐regulation of glutamate transporter activity by SOD1G93A, suggesting that this cell line might offer a convenient cellular model to examine pathophysiological EAAT2 regulation in ALS.
Objectives: To establish a cell model to elucidate the mechanisms by which SOD1G93A causes a down‐regulation of EAAT2. In this study we wished to determine whether it would be better to use the SH‐SY5Y neuroblastoma cell line or to develop a new system for expression of recombinant EAAT2 in cell lines.
Methods: We characterized glutamate transporter expression in SH‐SY5Y cells (passage 3–8) using RT‐PCR with primers specific for each of the human glutamate transporters EAAT1–5 and Western blotting. Full length EAAT2 cDNA was derived by RT‐PCR and cloned into the mammalian expression vector pcDNA3.1.
Results: RT‐PCR showed that SH‐SY5Y cells express mRNA for EAAT3 (EAAC1). Levels of mRNA for EAAT2 were negligible, even after 40 PCR cycles. EAAT1, 4 and 5 mRNAs were not present. Western blotting showed that EAAT2 migrates as a 66 kDa band and also as a higher band representing the multimeric form of the protein. However, SH‐SY5Y cells express very low levels of EAAT‐2 immunoreactive protein compared to rat frontal cortex, which was used as a positive control.
Conclusion: We conclude that, although EAAT2 is expressed by SHSY‐5Y cells, the low levels of EAAT2 expression indicate that this cell line is not optimal for detailed biochemical studies of its down‐regulation by SOD1G93A.
Reference
- Sala G, Beretta S, Ceresa C, et al. Impairment of glutamate transport and increased vulnerability to oxidative stress in neuroblastoma SH-SY5Y cells expressing a Cu/Zn superoxide dismutase typical of familial amyotrophic lateral sclerosis. Neurochem Int. 2005;46:227–34.
P75 EFFECT OF CEREBROSPINAL FLUID FROM MND PATIENTS ON SURVIVAL OF MOTOR NEURON CULTURES: IMPORTANT ROLE OF ASTROCYTES
Barber SC1, Wood‐Allum CA1, Walsh T2, Cox LE1, Monk PN1 & Shaw PJ1
1University of Sheffield, Sheffield, and 2MND Association Care and Research Centre, Sheffield, UK
E‐mail address for correspondence: [email protected]
Background: Whether cerebrospinal fluid (CSF) contains soluble factors that contribute to the propagation of disease between motor neuron (MN) populations in motor neuron disease (MND) remains controversial. Multiple studies using different culture conditions have produced conflicting findings, suggesting cellular environment may be a factor in determining the effect of CSF exposure (1–3).
Objectives:
1. Investigate the effect of CSF from MND patients and control patients on the survival of MN‐enriched cultures.
2. Compare the effects of MND/control CSF on MNs in isolation and in the presence of glial cells.
Methods: MNs were isolated from E14 Wistar rat embryos using established procedures and were grown either in mono‐culture on laminin, or in co‐culture on confluent astrocyte feeder layers. Cultures were exposed to CSF (either an artificial CSF (aCSF), or CSF from MND/control patients) at concentrations of either 20% or 50% for 24 h before fixation. Cultures were immunostained for neurofilaments and the number of MNs present in 20 random fields was counted in triplicate cultures.
Results: In MN‐enriched cultures, neither MND nor control CSF was able to support MN survival comparable to normal culture conditions or aCSF. MN survival was significantly greater in the presence of MND CSF than control CSF (p<0.05, Student's t‐ test). Conversely in co‐cultures, control CSF supported survival equivalent to aCSF, whereas MND CSF significantly reduced survival (p<0.01). Comparison of the effects of individual CSF samples in mono‐ and co‐culture revealed an inverse correlation, where a CSF sample supporting greater survival in mono‐culture was more likely to show toxicity to MNs in co‐culture (p = 0.001).
Discussion and conclusions: The environment surrounding the MN appears to be pivotal in determining the effect of CSF on MN survival. Whilst control CSF is not sufficient to support survival when MNs are grown in isolation, trophic support from glial cells enables control CSF to support MN survival in co‐cultures. The increased ability of MND CSF to support survival of isolated MNs compared to control CSF suggests the presence of additional trophic factors, although in co‐culture toxic signalling, presumably from glia, overrides this to reduce survival.
References
- Couratier P, Hugon J, Sindou P, et al. Lancet. 1993;341:256–8.
- Iwasaki Y, Ikeda K, Shiojima T, et al. Neurol Res. 1995;17:393–5.
- Tikka TM, Vartiainen NE, Goldsteins G, et al. Brain. 2002;125:722–31.
P76 IN VITRO PROPERTIES OF MICROGLIAL CELLS FROM G93A SOD1 TRANSGENIC MICE AND THEIR NON‐TRANSGENIC LITTERMATES
Sargsyan S, Monk P & Shaw P
The University of Sheffield, Sheffield, UK
E‐mail address for correspondence: [email protected]
Background: Microglial cells are increasingly capturing scientific attention as possible generators or propagators of motor neuronal injury in ALS. The hypothesis of the non‐cell autonomous damage to motor neurons proposes that the properties of neighbouring glial cells are altered by the presence of mutant SOD1 in SOD1‐related ALS cases, and that these alterations are toxic to the motor neurons in the vicinity. Microglial cells, being the most immunocompetent cells of the CNS parenchyma, are more likely to secrete cytotoxic products, to which motor neurons may display a selective vulnerability.
Objectives: The objective of this study was to investigate whether the properties of microglial cells from the G93A SOD1 ALS mouse models compared to their non‐transgenic littermates differ in vitro. The cellular properties investigated were: 1) TNF‐α cytokine production; 2) cell number; 3) cell size; 4) cellular ramification; 5) F4/80 receptor expression; and 6) phagocytosis.
Methods: Primary microglial cells were obtained from motor regions of 1 to 2‐day‐old mouse pup cortices. The cells were cultured in vitro until they reached 30 days of age, then stimulated with bacterial lipopolysaccharide (LPS) for 24 h and incubated with labelled latex beads for 1 h. After stimulation, TNF‐α production was measured using an ELISA kit. Morphological properties were analysed by immunocytochemistry and fluorescent microscopy.
Results: After LPS stimulation, both transgenic and non‐transgenic microglial cells showed increases in TNF‐α production, index of ramification, cell size and phagocytic ability. LPS caused a significant reduction in G93A SOD1 microglial cell numbers compared to that of the non‐transgenic cells; fold changes were 0.912 for the non‐transgenic and 0.478 for transgenic cells, p = 0.038. The levels of expression of the inflammatory receptor F4/80 with LPS stimulation were nearly doubled in the transgenic populations. We did not observe significant differences between the two cell types in TNF‐α secretion (when corrected for cell number), cell size, index of ramification or phagocytic ability.
Discussion and conclusions: We describe here some of the in vitro properties of resting and activated wild‐type and G93A SOD1 primary microglial cells. Our results show that at 30 days in vitro, wild‐ type or G93A SOD1 transgenic microglia do not differ in morphological characteristics investigated in this study. However, the sensitivity of transgenic microglial cells to environmental cues may be increased, as application of LPS significantly reduced the cell numbers in transgenic populations and showed a trend towards increased F4/80 expression. The mechanisms involved in the cell number reduction are currently under investigation.
P77 DIETARY FLAVONOIDS AT LOW CONCENTRATIONS ACTIVATE ERK, A PRO‐SURVIVAL SIGNAL IN NEURONS
Rainey‐Smith S, Fahmi AI, Williams RJ & Rattray M
King's College London, London, UK
E‐mail address for correspondence: [email protected]
Background: There are many studies that report neuroprotective actions of dietary‐derived flavonoids. As well as their classical hydrogen‐donating antioxidant activity, recent evidence suggests that flavonoids also exert modulatory effects on intracellular signalling pathways including ERK, a component of the pro‐survival mitogen‐activated protein kinase signalling cascade.
Objectives: We sought to characterize which of three key dietary flavonoids is likely to be the most potent neuroprotective agent by measuring their ability to regulate ERK phosphorylation. Representatives of different flavonoid classes were tested at concentrations relevant to those obtained through the diet.
Methods: Since flavonoid effects are dependent on cell type we used three culture systems: primary mouse cortical neurons, the neuron‐like SH‐SY5Y neuroblastoma cell line and human primary dermal fibroblasts (promocells). Cells were exposed for 15 min to the flavonoids over a concentration range of 10 nM to 10 µM. Flavonoids used were hesperetin (a flavonone), epicatechin (a flavonol) and quercetin (a flavonol). The cells were then lysed and Western blotting carried out, using a phospho‐specific antibody to detect and quantify activated ERK1/2 (pERK).
Results: In neurons, hesperetin was the most efficacious flavonoid examined, potently increasing pERK at lower concentrations (peak effect at 100 nM). Hesperetin had similar effects in SH‐SY5Y cells and in primary fibroblasts. In all cells examined, the concentration response to hesperetin was bell‐shaped: at higher concentrations (greater than 1 µM), hesperetin caused a relative reduction in pERK. Epicatechin showed more modest increases in ERK phosphorylation compared to hesperetin. The effects of quercetin were highly dependent on cell type. In neurons, most concentrations of quercetin caused a net down‐regulation in pERK compared to controls, whereas in SH‐SY5Y cells, quercetin increased pERK levels, even at the highest concentration examined (10 µM).
Discussion and conclusions: Low concentrations of hesperetin, epicatechin and quercetin all caused increased pERK in each cell type, but with different concentration‐response relationships that are cell type dependent. Our data predict that hesperetin, a flavonone which is particularly enriched in citrus fruit, is likely to be a neuroprotective flavonoid. The steps linking ERK phosphorylation and cytoprotection remain to be clearly elucidated. Data to examine the mechanisms underlying flavonoid‐induced neuroprotection will be useful when considering dietary supplementation and modification as a potential therapeutic strategy in ALS.
P78 UNDERSTANDING THE MECHANISM OF IMPAIRED STRESS RESPONSES IN MOTOR NEURONS
Taylor DM1, Minotti S1, De Koninck P2 & Durham HD1
1Montreal Neurological Institute, McGill University, Montreal, and 2Neurobiologie Cellulaire, Centre de Recherche Universite Laval Robert‐Giffard, Quebec, Canada
E‐mail address for correspondence: [email protected]
Background: Aggregation of abnormal protein occurs in familial and sporadic ALS, yet motor neurons fail to mount a defensive heat shock response due to a high threshold for activating the transcription factor, Hsf1 (1). Lowering this threshold in a stress‐sensitive manner has therapeutic potential in several neurodegenerative diseases. Binding of Hsf1 to heat shock elements (HSE) is not sufficient to initiate transcription of heat shock genes and various phosphorylation events are thought to control activation. Within the activation domain of Hsf1, phosphorylation of S303 by GSK3β and S307 by ERK1 decrease activation in cultured cells (2). Phosphorylation of S230 has been considered a requirement for induction of HSP expression and was induced experimentally by CaMKIIα (3).
Objective: This study tested the hypothesis that aberrant phosphorylation of key residues in the activation domain of Hsf1 is responsible for the high threshold for activating Hsf1.
Methods: Expression of stress‐inducible Hsp70 was evaluated in motor neurons and glia of dissociated murine spinal cord cultures and in mouse embryonic fibroblasts in the presence and absence of heat shock (43°C for 1 h; 6 h recovery). Phosphorylation of key residues was modified by pre‐treatment with the GSK3β inhibitor, lithium chloride (25 mM) or transfection with plasmid encoding a nuclear or cytoplasmic isoform of CaMKIIα or dominant negative ERK1.
Results: Lithium chloride failed to activate Hsf1 in heat shocked motor neurons, despite its effectiveness in non‐neuronal cells. Motor neurons in spinal cord cultures lacked endogenous expression of a nuclear CaMKIIα isoform, which would be required to phosphorylate DNA‐bound Hsf1, but over‐expression of nuclear or cytoplasmic CaMKIIα did not result in expression of Hsp70 following heat shock.
Conclusions: Our studies so far indicate that failure of phosphorylation or dephosphorylation of serine residues in the regulatory domain of Hsf1 is not responsible for the impaired heat shock response in motor neurons. Induction of Hsp70 in motor neurons by the Hsp90‐binding agents (steroid receptors, geldanamycin) suggests that interactions of Hsf1 with Hsp90 multichaperone complexes in the nucleus might underlie the high threshold for activation of heat shock gene expression in neurons.
Acknowledgement: This work was supported by CIHR/MDAC/ALS and ALSA.
References
- Batulan Z, Shinder GA, Minotti S, et al, Journal of Neuroscience. 2003;23:5789–98.
- Chu BY, Soncin F, Price BD, Stevenson MA, Calderwood SK. Journal of Biological Chemistry. 1996;271:30847–57.
- Holmberg CI, Hietakangas V, Mikhailov A, et al, EMBO Journal. 2001;20:3800–10.
P79 UP‐REGULATION OF HSP70 PREVENTS S‐NITROSOGLUTATHIONE (GSNO)‐INDUCED APOPTOSIS IN MOTOR NEURONAL CELLS EXPRESSING MUTANT CU/ZN SUPEROXIDE DISMUTASE (SOD1) VIA AN INHIBITION OF NF‐κB
Kim HJ1, Park JH1, Kim MY1, Cho AS1, Lee JS2, Seo JS2, Sung JJ3, Kim M3 & Lee KW3
1Department of Neurology, and Clinical Research Institute, Seoul National University Hospital, Seoul, 2Department of Biochemistry and Molecular Biology, Seoul National University College of Medicine, Seoul, and 3Department of Neurology, Seoul National University Hospital, Seoul, South Korea
E‐mail address for correspondence: [email protected]
Background and Objectives: Recent evidence suggests that nitric oxide (NO) toxicity is one of the primary mechanisms of motor neuronal degeneration. We previously reported that GSNO‐mediated toxicity decreased viability of motor neuronal cells expressing mutant SOD1 (1). Hsp70 is a well‐known heat shock protein that is cytoprotective against apoptosis in neurodegenerative disorders. In this study, we attempted to test the protective effect of Hsp70 on mutant SOD1 against NO‐mediated toxicity.
Methods: Motor neuron neuroblastoma hybrid (VSC 4.1) cells were stably transfected with pMFG.puro vector containing Hsp70 cDNA and analyzed by staining or Western blotting using an Hsp70 antibody to investigate overexpression of Hsp70. To test the effect of Hsp70 on constitutively expressing mutant SOD1, VSC4.1 cells co‐expressing Hsp70 and mutant SOD1 (A4V) were made. These cells were treated with GSNO and the viability was determined by 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay. In order to investigate the features of cell death, caspase 3 activity was assayed in treated cells. To determine the target for the protective effect of Hsp70, the expression of the active p65 NF‐κB subunit was measured by Western blotting.
Results: GSNO decreased the viability of cells expressing mutant SOD1 (A4V) in a dose‐dependent manner. GSNO‐induced mutant cell death was preceded by an increase of caspase 3 activation. Overexpression of Hsp70 in mutant cells blocked the GSNO‐induced elevation of caspase 3 activity. In addition, up‐regulation of Hsp70 prevented the expression and nuclear translocation of NF‐κB produced by GSNO.
Conclusions: Our data showed an anti‐apoptotic effect of Hsp70 against NO toxicity in motor neurons expressing mutant SOD1. This process seemed to involve a reduction of caspase 3 activation and levels of the active p65 NF‐κB subunit. Our data suggest that Hsp70 may act as a suppressor of apoptosis and oxidative stress in motor neurons expressing FALS‐linked mutant SOD1.
Reference
- Hyun-Jung Kim, Manho Kim, Sung Hun Kim, et al. Neuroreport 2002;13:1131–5.
P80 TETANUS TOXIN FRAGMENT C AND BCL‐2 FUSION PROTEIN PREVENTS DEATH OF SERUM‐ AND NGF‐WITHDRAWN PC12 CELLS
Watanabe Y1, Matsuba T2, Nakano T1, Kitayama M1, Yasui K1, Doi K1, Fukada Y1, Morita R1, Tanaka Y2 & Nakashima K1
1Department of Neurology, Institute of Neurological Sciences, Faculty of Medicine, Tottori University, Yonago, and 2Division of Microbiology, Department of Microbiology and Pathology, Faculty of Medicine, Tottori University, Yonago, Japan
E‐mail address for correspondence: [email protected]
Background: Tetanus toxin produced by Clostridium tetani is taken up by nerve endings at the neuromuscular junction and is retrogradely transported to the neuronal cell body. The C terminal region of the tetanus toxin (TTC) is in itself innocuous and has the same ability to be transported as the whole toxin. We explored whether the fragment of toxin can be applied to deliver a beneficial molecule to neuronal cells.
Objectives: Bcl‐2, an anti‐apoptotic protein, is reported to have prevented neuronal death in several experimental models in vitro as well as in vivo. We wished to fuse TTC with Bcl‐2 in order to examine whether the fusion protein bound to neuron‐like cells and whether it showed anti‐apoptotic effect. Theoretically, if the fusion protein retains the characteristics of both molecules, it will exhibit an anti‐apoptotic effect selectively in neurons.
Methods: The coding region of TTC was fused with Bcl‐2 cDNA in the sequence Bcl‐2 TTC or TTC Bcl‐2. The constructs were introduced into prokaryotic expression vectors, which added a His tag signal sequence at the N terminus of the gene product. The fusion proteins were expressed in Escherichia coli. The recombinant proteins were purified by immobilized metal affinity chromatography. In binding experiments, PC12 cells were differentiated by NGF, then the fusion protein was added to the cells. Binding capability was observed immunohistochemically. Whether exposure of the fusion protein had an anti‐apoptotic effect was examined by measuring its effect on the death of PC12 cells caused by withdrawal of growth factors (serum and NGF).
Results: TTC‐Bcl2 and Bcl2‐TTC fusion proteins were produced in E. coli and purified, although the purity of the protein was not sufficient. Among the purified proteins, Bcl2‐TTC but not TTC‐Bcl2 exhibited neuronal binding ability. Furthermore, Bcl2‐TTC prevented neuronal death of PC12 cells induced by serum and NGF deprivation to some extent. To improve the yield and purity of the fusion protein, we deleted the N terminal portion of TTC from Bcl2‐TTC, which we refer to as Bcl2‐halfTTC (Bcl2‐hTTC). Bcl2‐hTTC was produced efficiently in E. coli and its purity was better than that of Bcl2‐TTC. Bcl2‐hTTC retained the same propensity as Bcl2‐TTC with regard to neuronal binding and anti‐apoptotic properties.
Discussion and conclusions: Bcl2‐TTC and Bcl2‐hTTC are anti‐apoptotic in vitro. This intriguing nature of the fusion proteins could be a potent therapeutic agent in various neurodegenerative diseases, such as amyotrophic lateral sclerosis.
P81 TAUROURSODEOXYCHOLIC ACID (TUDCA), A BILE ACID, INHIBITS GSNO‐INDUCED APOPTOSIS BY MODULATING REACTIVE OXYGEN SPECIES (ROS) PRODUCTION IN MOTOR NEURONAL CELLS EXPRESSING MUTANT CU/ZN SUPEROXIDE DISMUTASE (SOD1)
Sung JJ2, Kim HJ1, Min JH2, Park JH1, Kim SM2, Cho AS1, Kim MY1, Kim M1 & Lee KW1
1Department of Neurology, and Clinical Research Institute, Seoul National University, Seoul, and 2Departnent of Neurology, Seoul National University Hospital, Seoul, South Korea
E‐mail address for correspondence: [email protected]
Background: The dihydroxy bile acid ursodeoxycholic acid (UDCA) and its taurine conjugated derivative, tauroursodeoxycholic acid (TUDCA), are hydrophilic bile acids which are normally produced endogenously in humans at a very low level. Recent evidence has shown that UDCA and its conjugated derivatives play a role in modulating apoptosis in both hepatic and non‐hepatic experimental models including neurodegenerative diseases.
Objectives: In this study, we examined whether TUDCA is neuroprotective in motor neuron‐neuroblastoma hybrid cells (VSC4.1) expressing mutant SOD1 (A4V, G93A) against NO toxicity.
Methods: To test the effect of TUDCA on GSNO‐mediated toxicity, TUDCA was pre‐incubated in motor neuronal cells expressing wild‐type (wt) or mutant (A4V, G93A) SOD1 and then treated with exogenous nitric oxide (S‐nitrosoglutathione; GSNO) for 24 h. Cell viability was determined by 3‐(4,5‐dimethylthiazol‐2‐yl)‐2,5‐diphenyltetrazolium bromide (MTT) assay. To investigate the characteristics of neuronal cell death, cells treated with GSNO were stained with Hoechst 33342 and were assayed for caspase 3 activity. To examine the free radical formation, the concentration of peroxides was measured by flow cytometry using fluorescence emitted by DCFH oxidation. Bax expression and poly (ADP‐ribose) polymerase (PARP) cleavage were also analyzed.
Results: GSNO‐induced mutant cell death was significantly reduced by TUDCA, which decreased the number of Hoechst 33342‐positive mutant cells and caspase 3 activity. Moreover, TUDCA prevented GSNO‐increased ROS generation, Bax expression and PARP cleavage.
Discussion and conclusions: The results suggest that TUDCA may contribute to the protection of motor neurons from degeneration caused by SOD1 mutations through anti‐apoptotic and anti‐oxidant mechanisms. Therefore, it may provide a potentially useful treatment in patients with some forms of amyotrophic lateral sclerosis.
P82 THE ROLE OF ENDOPLASMIC RETICULUM‐STRESS IN MOTOR NEURON DEGENERATION IN ALS MICE
Kieran DM & Prehn JHR
Department of Physiology, The Royal College of Surgeons, Dublin, Ireland
E‐mail address for correspondence: [email protected]
Background: Motor neuron degeneration in ALS has been proposed to occur by an active process of programmed cell death, and indeed many studies have shown the involvement of the mitochondrial apoptotic pathway. Less is known about the involvement of other organelles, such as the endoplasmic reticulum (ER), in motor neuron degeneration in ALS. ER stress occurs when misfolded proteins accumulate in the lumen of the ER. This initially results in the up‐regulation of cytoprotective ER‐resident chaperones, however if ER stress is severe and prolonged, cells eventually activate a caspase‐dependent cell death programme. We have previously shown that the pro‐apoptotic protein PUMA is necessary and sufficient for ER‐stress‐induced cell death in neurons.
Objectives: In this study we examined the expression of ER stress markers during disease progression in mutant SOD1 mice, and we also looked at the involvement of the BH3‐only pro‐apoptotic protein PUMA in mutant SOD1 motor neuron degeneration.
Methods: Using quantitative PCR and Western blotting, as well as immunohistochemistry, we examined the expression of ER stress markers and BH3‐only proteins in tissue from mutant SOD1 mice during disease progression.
Results: Our studies have shown that up‐regulation of ER‐resident chaperones Grp78 and Grp94 is evident during early disease progression, when the up‐regulation of other chaperones is becoming evident. Spinal cord sections immunostained with antibodies to Grp78 and Grp94 showed a large increase in expression in motor neurons. We also found an increase in the expression of the transcription factor CHOP during disease progression. The expression of the BH3‐only protein PUMA increased progressively during disease progression, and interestingly was elevated during the early stage of disease progression, compared to wild‐type littermates.
Conclusions: Our results suggest that ER‐stress is actively involved in motor neuron degeneration in mutant SOD1 mice.
P83 A STUDY ON THE INTRA‐MITOCHONDRIAL EFFECTS OF FALS‐SOD1S: IS THERE A CORRELATION WITH THE BIOPHYSICAL CHARACTERISTICS OF THE PROTEINS ?
Ferri A1, Cozzolino C2, Crosio C3, Nencini M2, Casciati A2, Russo C2, Rotilio G4 & Carrì MT4
1Istituto di Neuroscienze CNR, Sez. Psicobiologia e Psicofarmacologia, Rome, 2Fondazione S.Lucia IRCCS, Rome, 3Università di Sassari, Dipartimento di Scienze Fisiologiche Biochimiche e Cellulari, Sassari, and 4Dipartimento di Biologia, Università di Roma Tor Vergata, Rome, Italy
E‐mail address for correspondence: [email protected]
Background: Degeneration of mitochondria represents a pathological hallmark in ALS. A significant fraction of FALS‐associated mutant SOD1 is targeted to the mitochondria, where it forms aggregates containing proteins involved in the control of apoptosis. Proper folding of wild‐type SOD1 depends upon the binding of metal ions and/or the redox state of specific cysteine residues. The disulphide status of SOD1 seems to play a pivotal role in the monomer‐dimer equilibrium, stabilizing the quaternary structure of the protein.
Objectives: Data in cell or animal systems have been obtained only for a few mutant SOD1s, and no exhaustive analysis has been performed. In order to analyse the biochemical properties (propensity to form aggregates, monomer/dimer equilibrium, redox state of cysteine residues) of different SOD1 categories in the mitochondrial compartment of a neuronal environment, we have built a collection of inducible cell lines, derived from the mouse motor neuronal line NSC34, expressing a wide panel of SOD1 mutants.
Methods: The cDNAs coding for 12 FALS‐SOD1s were expressed in NSC34‐derived cell lines under the control of the inducible Tet‐ON promoter, with or without a short C‐terminal myc tail. For isolation of mitochondria, NSC34‐derived cell lines were homogenized and subjected to sequential centrifugations at different speeds. The crude mitochondrial pellet was loaded onto a discontinuous gradient for further purification. Western blot detection of accessible FALS‐SOD1 cysteines was performed on lysates from cytosolic and mitochondrial fractions. The samples were incubated in the presence of Mal‐PEG for covalent modification of accessible cysteines. The addition of Mal‐PEG to accessible cysteines increased the subunit mass of SOD1 by 5 kDa modification. Total and mitochondrial glutathione content was assayed in all NSC34‐derived cell lines. Low molecular weight thiols were separated by HPLC.
Results: In a motor neuronal cell environment some FALS‐SOD1s retain partial or full superoxide dismutase activity, whereas others do not. Mutants of the active site are present essentially in monomeric form. WT SOD1 is barely detectable in the mitochondrial compartment, while the fraction of FALS‐SOD1s localized in the mitochondria displays a decrease in cysteine accessibility. Moreover, most mutants do induce a decrease in the GSH/GSSG ratio inside mitochondria and this data parallels a marked decrease in complex I activity (NADH‐ubiquinone oxidoreductase).
Conclusions: In agreement with the hypothesis that mitochondria are the site of primary damage in SOD1‐linked FALS, the expression of FALS‐SOD1 mutants induces a selective alteration in mitochondrial GSH/GSSG ratio and a failure in the respiratory activity of complex I. These alterations are related to the mitochondrial localization and cysteine oxidation state of FALS‐SOD1s.
P84 ANALYSIS OF FALS‐ AND NON‐FALS LINKED SOD1 GLYCINE‐93 VARIANTS EXPRESSED IN MOTOR NEURONAL CELLS
Turner BJ, Atkin JD, Farg MA, Rembach A & Cheema SS
Brain Injury and Repair Group, Howard Florey Institute, University of Melbourne, Melbourne, Australia
E‐mail address for correspondence: [email protected]
Background and objectives: Over 100 mutations in the SOD1 gene are implicated in familial forms of amyotrophic lateral sclerosis (FALS), yet predict proteins with diverse biochemical and structural properties despite a similar pathogenesis. Mutation of glycine residues occurs frequently in SOD1‐linked FALS. In particular, codon 93 is point mutated to six possible variants in FALS, suggesting this glycine residue is critical for maintenance of native structure or function. Thus, glycine‐93 may assume a local protein conformation forbidden to any other amino acid substitution. To test this hypothesis, murine motor neuron‐like (NSC‐34) cells expressing every possible SOD1 missense variant at position 93 were analysed for key biochemical parameters, including aggregation propensity, solubility, enzyme activity and pro‐oxidant chemistry.
Methods: NSC‐34 cells were transfected with pEGFP‐N1 vectors containing human wild‐type (wt), FALS‐linked (G93A, G93C, G93D, G93R, G93S, G93V) or non‐FALS linked (G93E, G93F, G93H, G93I, G93K, G93L, G93M, G93N, G93P, G93Q, G93R, G93S, G93T, G93V, G93W, G93Y) mutant SOD1 cDNAs. Cell transfectants were analysed using fluorescence microscopy, immunoblotting, zymography, dichlorofluorescein (DCF) oxidation and viability assays. SOD1 mutants were also visualized using molecular modelling and correlated with cell culture and biochemical data.
Results and discussion: The SOD1 glycine‐93 mutants exhibited differential electrophoretic mobilities, activities, DCF oxidation and toxicity. However, cytoplasmic inclusion formation, impaired solubility and enhanced proteolysis were common to all 19 variants of SOD1. Thus, no clear relationship between aggregation and the substituted side chain properties (e.g. size, charge, polarity, hydrophobicity) was evident in mutants. These results suggest that preservation of glycine‐93 is essential to maintain native SOD1 structure while preventing unfavourable self‐association.
P85 IDENTIFICATION OF DIFFERENT ISOFORMS OF THE SOD1 PROTEIN IN A CELLULAR MODEL OF AMYOTROPHIC LATERAL SCLEROSIS
Di Poto C1, Salvini R2, Bardoni A2, Passadore I2, Corato M3, Cereda C3 & Ceroni M4
1Neurological Institute IRCCS, ‘C. Mondino’, Department of Biochemistry, University of Pavia, Pavia, 2Department of Biochemistry, University of Pavia, Pavia, 3Neurological Institute IRCCS, ‘C. Mondino’, Pavia, and 4Neurological Institute, ‘C. Mondino’, Department of Neurological Science, University of Pavia, Policlinico of Monza, Pavia, Italy
E‐mail address for correspondence: [email protected]
Background: Familial amyotrophic lateral sclerosis (FALS) has been associated with mutations in the SOD1 gene encoding the Cu/Zn superoxide dismutase enzyme (SOD1). It has been hypothesized that the disease is not caused by a modification of the dismutase activity but by a gain of a new toxic function. The pathogenic mechanism is currently unknown, but studies made in other neurodegenerative diseases (1) suggest that the SOD1 protein is a major target of post‐transcriptional modification.
Objectives: We have analysed, by proteomics tools, changes in expression and post‐transcriptional modifications of the SOD1 protein in human neuroblastoma SH‐SY5Y cells transfected with human wild‐type and two different mutated SOD1 genes: H46R and G93A.
Methods: Protein extracts from each line were separated by 2D gel electrophoresis and protein spots were detected by the Coomassie and silver stains and analysed by PD‐Quest software. SOD1 protein was identified by using a combination of mass spectrometry and Western blotting analysis. For the identification of the different isoforms of SOD1 protein, the gels of each line were transferred on to PVDF membrane. The membranes were probed with anti‐human SOD1 antibody.
Results: SOD1 two‐dimensional immunoblotting displayed four isoforms of the SOD1 protein with isoelectric points (pI) of 5.0, 5.7, 6.0 and 6.3 respectively. In particular, the SH‐SY5Y line showed only three isoforms (pI 5.0, 5.7 and 6.3) while the isoform at pI = 5.7 is a characteristic of the transfected cell lines. The analysis of the two‐dimensional gels showed an increase of the total quantity of the SOD1 isoforms in the transfected lines compared to the untransfected cells. Data for the expression of total SOD1 protein were confirmed by one‐dimensional Western blotting. Two‐dimensional gels also demonstrated that the isoforms show differences of expression.
Conclusions: The transfected cell lines express an isoform of the SOD1 protein not present in the untransfected cell line. Moreover, the four isoforms are differentially expressed in the WT, G93A and H46R lines. These data indicate that the overexpression of the SOD1 protein in the transfected lines induces post‐transcriptional modification of the protein as suggested by the appearance of a new isoform.
Reference
- Choi et al. J Biol Chem. 2005;280:11648–55
P86 ALS2/ALSIN REGULATES RAC‐PAK SIGNALLING AND NEURITE OUTGROWTH
Tudor EL, Perkinton MS & Miller CCJ
Institute of Psychiatry, London, UK
E‐mail address for correspondence: [email protected]
Background: Mutations in ALS2 are causative for some recessive familial forms of ALS. The structure of ALS2 predicts that it functions as a guanine nucleotide exchange factor (GEF). GEFs regulate the activity of members of the Ras superfamily of GTPases. ALS2 contains three putative GEF domains, the central region containing Dbl and pleckstrin homology domains that are found in GEFs for Rho family members. The recessive nature and types of mutations in affected families strongly suggest that a loss of ALS2 function is the primary cause of disease. As such, a proper understanding of the molecular mechanisms by which mutant ALS2 induces motor neuron disease requires insight into its function.
Objectives: The objectives were to gain insight into ALS2 GEF function.
Methods: The subcellular distribution of ALS2 in neurons was examined by immunocytochemical analyses. The effects of ALS2 on the activities of Rho family members and PAK1, and on neurite outgrowth were all investigated using cellular assays.
Results: Activity assays revealed that ALS2 stimulated the activity of Rac but not Rho or Cdc42. Rac regulates the activity of PAK family kinases and ALS2 also stimulated PAK1 activity. Mutation of the ALS2 Rac GEF domain abrogated these effects. During development, Rac‐PAK signalling functions to regulate actin dynamics in the growth cone so as to control neurite outgrowth. ALS2 localized within growth cones of neurons and expression of ALS2 significantly promoted neurite outgrowth by approximately 1.5‐fold. Disruption of the ALS2 Rac GEF domain and expression of dominant negative Rac but not dominant Rab5 abrogated the stimulatory effect of ALS2 on neurite outgrowth.
Discussion: We demonstrate that ALS2 is present within neuronal growth cones and that it functions as a GEF for Rac so as to stimulate Rac‐Pak signalling and neurite outgrowth. Mutations in ALS2 may therefore induce disease by compromising proper development of motor neurons, making them more susceptible to later toxic insults.
P87 AMYLOID PRECURSOR PROTEIN OVEREXPRESSION IN A CELLULAR MODEL OF AMYOTROPHIC LATERAL SCLEROSIS.
Crestini A1, Piscopo P1, Bonfiglio E1, Pieri M2, Zona C2 & Confaloni A1
1Istituto Superiore di Sanità, Rome, and 2University of Rome, Tor Vergata, Rome, Italy
E‐mail address for correspondence: [email protected]
Background: Some studies indicate that amyloid precursor protein (APP), the transmembrane precursor of β‐amyloid peptide, increases as a consequence of central nervous system insults (i.e. cholinergic lesions, hypoxic‐ischaemic injury, traumatic brain injury) at an early stage after damage. Recently, it was also observed that APP is increased in the anterior horn motor neurons of ALS patients. Also, mutations in SOD1 associated with amyotrophic lateral sclerosis (ALS) confer enhanced pro‐oxidative enzyme activities. Thus, mutant SOD1‐mediated ALS conforms to classic neurodegenerative diseases such as Alzheimer's disease (AD), by a common, not well known, neurodegenerative pathway, which causes self‐aggregating and neurotoxic proteins to accumulate in the central nervous system.
Objectives: The present study examines the potential link between pro‐oxidative G93A SOD1 and the expression of some components influencing amyloidogenesis. In fact, a causal relationship between oxidative stress and AD is well known. With this aim, we examined whether the expression of some Alzheimer's genes was modulated in cortical cell cultures of transgenic ALS mice.
Methods: Three different cultures were prepared, murine, expressing human SOD1 (hSOD1) and G93A mutated hSOD1 cells. A real time PCR method was used to analyse the APP, PSEN1 and PSEN2 expression. Total RNA from each sample was used to generate cDNA by reverse‐transcriptase and random primers. For quantitative PCR, primers and probe mixtures specific for each gene were used. The relative quantification was performed using the comparative CT method and 18s as an endogenous reference.
Results: Our data show a significant overexpression of APP in cortical G93A cultures compared to hSOD1 cultures, which were used as controls (1.8±0.1 vs. 1±0.1; p<0.05). An increase of PSEN1 and PSEN2 mRNA levels was also evident, but without any statistical significance.
Discussion and conclusions: Our results indicate a different transcriptional response of APP with respect to presenilins induced by the mutated SOD. Increased APP could correspond to the brain β‐amyloid accumulation in mutant SOD1 according to Turner et al. These findings indicate that G93A SOD1 in the cortical cultures may serve to up‐regulate APP synthesis and indirectly contribute to β‐amyloid deposition by means of higher substrate availability for the proteolytic enzymes of the amyloidogenic pathway.
P88 GENE EXPRESSION PROFILE OF SPINAL MOTOR NEURONS IN SOD1 G93A MOUSE MODEL
Ferraiuolo L, Heath P, Holden H, Baptista M, Kirby J & Shaw P
University of Sheffield, Sheffield, UK
E‐mail address for correspondence: [email protected]
Background: Amyotrophic lateral sclerosis is one of the most common adult onset neurodegenerative diseases, characterized by progressive and selective degeneration of the upper and lower motor neurons in the motor cortex, brain stem and spinal cord. The causative mechanism of this cell‐specific loss is still unknown.
Objective: The aim of this study is to investigate the changes in gene expression profiles of degenerating spinal motor neurons isolated from human SOD1 G93A and SOD1 WT mice and non‐ transgenic littermates at 60, 90 and 120 days in order to identify which pathways are involved in the development of the neurodegenerative process.
Methods: Approximately 1000 motor neurons were isolated from lumbar spinal cord. RNA was extracted using Picopure kit (Arcturus), amplified using the RiboAmp Amplification kit (Arcturus) and labelled using the BioArray High Yield RNA Transcript Labelling Kit (Enzo). 10 µg cRNA was applied to the Affy MOE430A GeneChip, and data analysis was performed using ArrayAssist System (Iobion).
Results: Initial studies comparing motor neurons from 120‐day‐old SOD1 G93A, SOD1 WT and their respective non‐transgenic littermate mice show a significant change in expression of 241 genes of the 14,000 arrayed, when applying criteria of changes of two‐fold or higher. One hundred and fifty of the genes examined are down‐regulated, and 91 are up‐regulated. The genes have been categorized according to their molecular function, and include genes involved in the transcription process, antioxidant and stress responses, apoptosis, and protein degradation. Consistent with results obtained by the analysis of the NSC34 cell line transfected with vector expressing human G93A SOD1 (1), the transgenic mice also show a marked degree of transcriptional repression. Interestingly, genes involved in antioxidant activity and stress response are significantly decreased, underlining the important role that oxidative stress plays in ALS. No significant differences have been found comparing SOD1 WT mice with their littermates.
Discussion and conclusions: We aim to present gene expression changes identified in earlier stages of disease, in order to find out which cellular functions are initially impaired and to determine which processes are involved in motor neuronal degeneration. Microarray technology combined with laser‐capture microdissection is the most appropriate approach to understanding the motor neuron‐specific expression profile related to the degeneration process occurring in ALS, as demonstrated by Jiang et al. (2). It enables the isolation of single cell types allowing us to overcome bias of motor neuron loss, reactive astrogliosis and other cellular reactions that interfere with the analysis of motor neuron‐specific transcripts.
References
- Kirby J, Halligan E, Baptista M, et al. Mutant SOD1 alters the motor neuronal transcriptome: implications for familial ALS. Brain. 2005;128:1686–706.
- Jiang Y, Yamamoto M, Kobayashi Y, et al. Gene expression profile of spinal motor neurons in sporadic amyotrophic lateral sclerosis. Ann Neurol. 2005;57:236–51.
P89 DEVELOPMENT OF A THREE‐DIMENSIONAL TISSUE‐ENGINEERED IN VITRO MODEL TO STUDY MOTOR NEURON DISEASES
Gingras M1, Durham HD2 & Berthod F1
1Laval University, Quebec, and 2McGill University, Montreal, Canada
E‐mail address for correspondence: [email protected]
Background: Motor neuron disorders such as amyotrophic lateral sclerosis are very complex diseases and their development and progression involve many cellular processes and many cell types. Current in vitro models include neuronal cell lines, primary cell cultures and three‐dimensional (3‐D) tissue slices. Primary cell cultures usually contain mixed cell types and dividing support cells rapidly overcome neurons, while tissue slices do not allow the addition or removal of specific cells. Thus, a good in vitro model for the study of motor neuron disorders in a 3‐D physiological environment would fill the gap between elementary models of cell culture and complex models of transgenic mice.
Objectives: Taking advantage of our background in tissue engineering (1,2), our aim was to develop a highly physiological 3‐D model that could facilitate the study of motor neuron diseases in vitro. Specific objectives were to purify motor neurons and to incorporate them into a remodelled 3‐D biomaterial (sponge) containing an extracellular matrix and environment that could allow the culture and neurite extension of motor neurons.
Methods: We seeded mouse fibroblasts alone or with Schwann cells onto collagen and chitosan sponges and cultured them in the presence of ascorbic acid. We isolated motor neurons from embryonic mouse spinal cords and purity was determined by staining of Hb9 followed by FACS analysis. Motor neurons were added freshly to the sponges that were further cultured with or without neurotrophic factors (BDNF, GDNF, CNTF and NT‐3). Histological analysis was performed and neurite outgrowth was assessed by immunofluorescence staining of neurofilaments.
Results: The technique yielded more than 90% of Hb9‐positive motor neurons. Culture of the sponges seeded with mouse fibroblasts or mouse fibroblasts and Schwann cells allowed a good production of extracellular matrix. This suited the purified motor neurons well (which are usually difficult to grow in culture), as they survived in the models and extended neurites in the sponges. However, the addition of neurotrophic factors was required in order to induce a deeper extension of motor neuron neurites into the models.
Discussion and conclusions: We developed a 3‐D tissue‐engineered in vitro model that mimics the physiological environment of motor neurons and allows their culture and neurite extension. We will further complete the model by adding and co‐culturing support cells that are possibly implicated in the pathogenesis of motor neuron diseases (astrocytes, microglial cells, endothelial cells). Thus, this model represents a powerful tool for the study of motor neuron disorders.
Acknowledgements: This study was supported by the Muscular Dystrophy Association and by the Fonds de la Recherche en Santé du Québec.
References
- Gingras M, Bergeron J, Déry J, et al. FASEB J. 2003;17:2124–6.
- Gingras M, Paradis I, Berthod F. Biomaterials. 2003;24:1653–61.
P90 BONE MARROW STROMAL CELLS PROMOTE PROLIFERATION OF POTENTIAL NEURAL STEM CELLS IN ORGANOTYPIC SPINAL CORD SLICE CULTURES
Tsuji S, Kikuchi S, Tashiro J, Shinpo K & Sasaki H
Hokkaido University, Sapporo, Japan
E‐mail address for correspondence: [email protected]
Background: Recent reports have suggested that transplantation of bone marrow stromal cells (BMSCs) could rescue neurons in ischemic brains and in injured spinal cords of rodent models. They also promoted proliferation of stem cells in the subventricular zone (SVZ). However, details of the mechanism remain unclear. Clinical trials are now underway among human ALS patients, although a basic understanding has not yet been established.
Objectives: To investigate the effect of BMSCs on spinal cord, we established slice cultures from rat spinal cord that were maintained with BMSCs in the same wells (bi‐culture).
Methods: BMSCs were obtained from adult Sprague‐Dawley rats and cultured in plastic flasks. The spinal cords were isolated from the same strain on postnatal day 6, and cut into 400 µm‐thick transverse slices. The slices were cultured on membranes inserted in six‐well‐plates on the bottom of which BMSCs had been seeded in advance.
Results: The slice cultures without BMSCs increased their thickness and showed marked gliosis especially in the margin of slices from 10 days in vitro. However, cultures with BMSCs did not become thick. The protein content of bi‐cultured slices was half that of conventional cultures. Immunohistochemical analysis of bi‐cultures by GFAP staining showed less glial response. Furthermore, nestin‐positive cells proliferated around the margin of slices. They had taken up BrdU soon after slice‐preparation.
Discussion and conclusion: The results of this study suggest that BMSCs can promote the proliferation of nestin‐positive cells instead of gliosis in spinal cords. Those newly divided, nestin‐positive cells could be potential neural stem cells, though we have yet to confirm whether they can differentiate into neurons in the next step. One of the practical problems for regenerative therapy is the source of stem cells. BMSCs could be a useful tool to solve this problem.
P91 IDENTIFICATION OF COMPOUNDS CAPABLE OF REGENERATING MOTOR AXONS BY SMALL MOLECULE SCREENING PROGRAM
Haenggeli C, Coccia C & Rothstein JD
Johns Hopkins University, Baltimore, USA
E‐mail address for correspondence: [email protected]
Background: ALS is characterized by the loss of specific motor neuron pools and the associated degeneration of motor axons. A large number of studies have focused on finding ways to prevent motor neuron cell death. Finding treatments able to rescue motor neuron soma would be very promising. However, to assure the complete recovery of motor function, the discovery of compounds capable of regenerating motor nerves is critical.
Objectives: The aim of this study was to identify molecules capable of inducing motor axon growth.
Methods: To do this we treated post‐natal rat spinal cord organotypic cultures with drugs from a library of 1040 Food and Drug Administration‐approved drugs and nutritionals. Tissue cultures were incubated for 7 and 14 days with 10 μM of each drug (n = 10 cultures/drug), then collected and stained for neurofilament‐bearing axons with SMI‐32. All cultures were blindly rated for axon outgrowth by qualitative and quantitative scales. Glial‐derived neurotrophic factor (GDNF)‐treated cultures served as positive control, while DMSO‐treated cultures served as a vehicle control. All positive hits were re‐evaluated by treating the organotypic cultures with different doses of drugs.
Results and conclusions: Multiple compounds were found to induce motor neuron axon growth in vitro. To validate the potential of these agents in vivo, we will assess them in various paradigms of acute axonal injury such as facial nerve axotomy and sciatic nerve crush.