Present address: National Institute for Longevity Sciences, Obu, Japan.
Slow and selective death of spinal motor neurons in vivo by intrathecal infusion of kainic acid: implications for AMPA receptor-mediated excitotoxicity in ALS
Article first published online: 22 MAY 2006
Journal of Neurochemistry
Volume 98, Issue 3, pages 782–791, August 2006
How to Cite
Sun, H., Kawahara, Y., Ito, K., Kanazawa, I. and Kwak, S. (2006), Slow and selective death of spinal motor neurons in vivo by intrathecal infusion of kainic acid: implications for AMPA receptor-mediated excitotoxicity in ALS. Journal of Neurochemistry, 98: 782–791. doi: 10.1111/j.1471-4159.2006.03903.x
- Issue published online: 19 JUN 2006
- Article first published online: 22 MAY 2006
- Received January 6, 2006; revised manuscript received March 15, 2006; accepted March 20, 2006.
- amyotrophic lateral sclerosis;
- AMPA receptor;
- RNA editing
Excitotoxicity mediated by α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptors has been proposed to play a major role in the selective death of motor neurons in sporadic amyotrophic lateral sclerosis (ALS), and motor neurons are more vulnerable to AMPA receptor-mediated excitotoxicity than are other neuronal subclasses. On the basis of the above evidence, we aimed to develop a rat model of ALS by the long-term activation of AMPA receptors through continuous infusion of kainic acid (KA), an AMPA receptor agonist, into the spinal subarachnoid space. These rats displayed a progressive motor-selective behavioral deficit with delayed loss of spinal motor neurons, mimicking the clinicopathological characteristics of ALS. These changes were significantly ameliorated by co-infusion with 6-nitro-7-sulfamobenso(f)quinoxaline-2,3-dione (NBQX), but not with d(–)-2-amino-5-phosphonovaleric acid (APV), and were exacerbated by co-infusion with cyclothiazide, indicative of an AMPA receptor-mediated mechanism. Among the four AMPA receptor subunits, expression of GluR3 mRNA was selectively up-regulated in motor neurons but not in dorsal horn neurons of the KA-infused rats. The up-regulation of GluR3 mRNA in this model may cause a molecular change that induces the selective vulnerability of motor neurons to KA by increasing the proportion of GluR2-lacking (i.e. calcium-permeable) AMPA receptors. This rat model may be useful in investigating ALS etiology.
artificial cerebrospinal fluid
amyotrophic lateral sclerosis
α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid
Cu/Zn superoxide dismutase
The most common motor neuron disease, amyotrophic lateral sclerosis (ALS), is a progressive neurodegenerative disease characterized by selective upper and lower motor neuron loss that is initiated in mid-life and leads to death as a result of respiratory muscular weakness. ALS has a uniform worldwide prevalence (five cases per 100 000 individuals), of which sporadic ALS accounts for more than 90% of all cases and only the remaining 5% of cases are familial (FALS) (Roland et al. 2005). Despite the fact that a good animal model is indispensable for investigating the etiology and for developing novel therapies for a disease, few appropriate animal models for sporadic ALS have been developed. Among the animal models tested, animal lines transgenic for the mutated human Cu/Zn superoxide dismutase (SOD1) gene (Gurney et al. 1994; Nagai et al. 2001; Howland et al. 2002) are regarded as a candidate disease model for all types of ALS, but the etiology of the FALS caused by the SOD1 gene mutation is not necessarily the same as that of sporadic ALS. Indeed, the extent of neuropathological changes is not confined to motor neurons in these animals (Gurney et al. 1994), and although SOD1 gene mutations have been found in a small population of patients with sporadic ALS, no significant association has been detected between sequence variants in the SOD1 locus and sporadic ALS susceptibility or phenotype (Jackson et al. 1997; Broom et al. 2004). Furthermore, the underediting of GluR2 mRNA at the Q/R site, which is a specific molecular change in sporadic ALS motor neurons (Kawahara et al. 2004), does not occur in degenerating motor neurons in rats transgenic for mutated human SOD1 (Kawahara et al. 2006). Although SOD1 transgenic animals have been used widely, these cannot be used as a model for sporadic ALS. On the other hand, a mouse line transgenic for Ca2+-permeable artificial GluR-B(N) has been reported to develop motor neuron loss after 12 months of survival (Feldmeyer et al. 1999; Kuner et al. 2005), and therefore may be similar to sporadic ALS with respect to etiology, but this model has not been widely used.
Motor neurons in the spinal cord are differentially more vulnerable to α-amino-3-hydroxy-5-methyl-4-isoxazole propionic acid (AMPA) receptor-mediated excitotoxicity than are other neuronal subsets in vitro owing to an increased flux of Ca2+ through AMPA receptor channels (Carriedo et al. 1996). Ca2+ influx through AMPA receptors is regulated by either the presence or the absence of a GluR2 subunit: AMPA receptors containing GluR2 have a low Ca2+ influx, whereas those containing no GluR2 have a high Ca2+ influx (Hollmann et al. 1991; Verdoorn et al. 1991; Burnashev et al. 1992). However, when AMPA receptors contain a GluR2 subunit translated from unedited mRNA, their Ca2+ permeability remains high (Sommer et al. 1991). Therefore, both deficiency in GluR2 expression and underediting of GluR2 mRNA can induce AMPA receptor-mediated neuronal death. It is therefore interesting to investigate whether the AMPA receptor-mediated slow death of motor neurons occurs in vivo and what molecular changes occur in the AMPA receptor subunits. Here we induced the selective death of motor neurons in adult rats by a continuous infusion of kainic acid (KA) into the spinal subarachnoid space. Because slowly progressive and selective neuronal death is a hallmark of degenerative neurological diseases, such a model is useful to investigate the mechanism underlying the selective death of motor neurons in ALS.
Materials and methods
Each test compound was dissolved in artificial cerebrospinal fluid (aCSF; 122 mm NaCl, 3.1 mm KCl, 5 mm NaHCO3, 0.4 mm KH2PO4, 1.3 mm CaCl2, 1.0 mm MgSO4 and 10 mm d-glucose, pH 7.4) with the pH adjusted to 7.2, and the resulting solution was used to fill an Alzet Model 2004 osmotic minipump (capacity 200 µL, pump speed 0.25 µL/ h; DURECT Corp., Cupertino CA, USA), which was incubated in sterile saline solution at 37°C overnight before operation after connection with a PE10 cannula (inner diamater, 0.25 mm; outer diamater, 0.55 mm; EICOM Corp., Tokyo, Japan). The operation was performed with slight modifications to the protocol described by Nakamura et al. (1994, 1997). In brief, male Wistar (W; body weight 180–250 g) and Fischer (F; body weight 110–160 g) rats (Japanese Oriental Yeast Co. Ltd, Shizuoko, Japan) were laminectomized at L5/6 under deep pentobarbiturate anesthesia, the free end of the PE10 cannula was inserted into the lumber spinal subarachnoid space, and the osmotic minipump was placed subcutaneously in the back.
Either W or F rats were infused with 3 mm KA (Sigma-Aldrich Corp., St Louis, MO, USA) continuously for 2 (KA-W2, KA-F2), 4 (KA-W4, KA-F4) or 8 (KA-W8) weeks. As controls, either W (aCSF-W) or F (aCSF-F) rats were infused with aCSF in the same manner for the same period of KA infusion (i.e. 2, 4 or 8 weeks). In some experiments, in order to investigate the neurotoxic mechanism, male F rats were infused intrathecally for either 2 or 4 weeks with one of the following combinations of glutamate receptor-acting drugs: 3 mm KA plus 3 mm 6-nitro-7-sulfamoyl-benzo(f)quinoxaline-2,3-dione (NBQX; TOCRIS Cookson Ltd, Bristol, UK) (KA/NBQX group); 3 mm KA plus 3 mm d(–)-2-amino-5-phosphonopentanoic acid (APV; TOCRIS NEURAMIN, Buckhurst Hill, Essex, UK) (KA/APV group); 1.5 mm cyclothiazide (CTZ; TOCRIS, Ellisville, MO, USA) (CTZ group); or 1.5 mm KA plus 1.5 mm CTZ (KA/CTZ group). In the KA/CTZ and CTZ groups, the osmotic minipump was replaced every week with one refilled with freshly prepared 1.5 mm KA plus 1.5 mm CTZ solution and 1.5 mm CTZ solution, respectively.
Animals were handled according to the protocols approved by the Institutional Animal Care and Use Committee in line with the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health.
Evaluation of movement behavior
Throughout the experimental period, the movement behavior of rats was measured each week, mainly by using a rat-specific rotarod (SN-445; Neuroscience Corp., Tokyo Japan), which constantly rotated at a speed of 16 rpm and automatically recorded the time that rats could stay on the rotarod during a 420-second period. The rotarod scores of rats were recorded on average in three separate trials. Paralysis was scored when the rat dragged one of its hindlimbs. Rats were also tested for whether they exhibited an escape response from the pain elicited by stimulation of the hindlimbs with tweezers, and whether they had urinary and/or fecal incontinence.
Spinal cord samples
The rats were anesthetized deeply using diethyl ether inhalant, after which their spinal cords were quickly removed and frozen in liquid nitrogen. The samples were stored at −80°C until use. The frozen spinal cord samples were obtained from KA-W2, KA-W4 and KA-W8 (n = 5 each); KA-F2 and KA-F4 (n = 5 each); KA/NBQX (n = 5); KA/APV (n = 5); KA/CTZ (n = 5); CTZ (n = 5); aCSF-W2, aCSF-W4 and aCSF-W8 (n = 5 each); and aCSF-F2 and aCSF-F4 (n = 5 each) rats.
For morphological analysis, rats were perfused transcardially with 3.5% paraformaldehyde and 0.5% glutaraldehyde in 0.1 m phosphate-buffered saline (PBS; pH 7.6). The spinal cords of KA-W4 and KA-W8 (n = 3 each), KA-F2 and KA-F4 (n = 3 each), KA-NBQX (n = 3), KA/CTZ at 2 and 4 weeks (n = 5 each), CTZ (n = 5), aCSF-W4 and aCSF-W8 (n = 3 each), and aCSF-F2 and aCSF-F4 (n = 5 each) rats were removed and postfixed in the same fixative at 4°C for 12 h.
The lumber segments of spinal cords fixed in paraformaldehyde and glutaraldehyde were dehydrated overnight at 4°C with serial concentrations of sucrose (5%, 10%, 15%, 20%, 25% and 30%), and then rapidly frozen on dry ice. Thirty serial 20-µm-thick frozen sections were made with a cryostat (Model HM500 O; MICROM, Walldorf, Germany) and stained with either 0.1% Cresyl violet (pH 3.5) or hematoxylin and eosin. The number of neurons the diameters of which were greater than 20 µm and had identifiable nucleoli was counted in the ventral horns of the spinal cord under a light microscope, and their morphology was also observed.
To determine the morphological changes in the axons of motor neurons, L5 ventral roots of KA-W8 (n = 3), KA/CTZ (n = 3) and their respective control rats (n = 3 each) were postfixed with 2.0% paraformaldehyde and 2.0% glutaraldehyde in 0.1 m PBS, and mounted in resin. Transverse 1-µm-thick Epon-embedded sections of L5 ventral roots were made and stained with 0.1% toluidine blue and viewed under a light microscope.
Single-cell isolation was carried out using an excimer laser microdissection system (Hamamatua Photonics Ltd, Shizuoka, Japan) as previously described (Hashida et al. 2001; Kawahara et al. 2003; Sun et al. 2005). In brief, 20-µm-thick frozen sections were attached to glass slides made of artificial quartz, fixed with 100% methanol for 60 s and then stained with 0.1% toluidine blue. Thirty spinal motor neurons (Motor) were dissected free and placed in test-tubes containing 200 µL of TRIZOL reagent (Invitrogen Corp., Carlsbad, CA, USA). The substantia gelatinosa, ventral and dorsal funiculi of the spinal cord were dissected en bloc and placed in tubes containing 200 µL of TRIZOL reagent in a similar manner. All samples were stored at −20°C until use.
RNA extraction and reverse transcription
Total RNA was extracted from each tissue sample using TRIZOL reagent according to the manufacturer's instructions. Single-cell RNA extraction and reverse transcription procedures were carried out as previously described (Hashida et al. 2001; Kawahara et al. 2003; Sun et al. 2005). Reverse transcription was performed with either Ready-to-go You-Prime First-Strand beads (Amersham Biosciences Corp., Piscataway, NJ, USA) or 0.5 µm oligo(dT) primer for tissue samples, or with a Sensiscript RT Kit (QIAGEN GmbH, Hilden, Germany) using 0.5 mm oligo(dT) primer and 10 U of prime RNase inhibitor (Eppendorf AG, Hamburg, Germany) for the single-cell samples. The cDNAs were stored at −20°C until use.
Real-time quantitative PCR
The primer sets for the internal standard of GluR1–R4 and β-actin have been described elsewhere (Sun et al. 2005). Each PCR reaction was run in 50 µL of the reaction mixture containing each primer at 200 µm, 1 mm dNTP MIX, 5 µL of 10 × PCR buffer and 1 µL of Ampli Taq DNA polymerase (Applied Biosystems, Roche Molecular Systems Inc., Branchburg, NJ, USA). The PCR amplification conditions were 95°C for 2 min, followed by 35 cycles of 95°C for 30 s, 60°C for 30 s and 72°C for 1 min. The PCR products were then purified with a Zymoclean Gel DNA Recovery Kit (Zymo Research, Orange, CA, USA). The concentration of PCR products was determined with a spectrophotometer (Nano DropTM ND-1000; Nano Drop Technologies Inc., Wilmington, DE, USA). We prepared serial dilutions with 108−106, 106−104, 104−103, 103−102 and 103−101 copies per 2 µL, which contained 100 ng of herring sperm DNA in Tris-EDTA buffer (pH 8.0).
The procedure was performed as described previously (Kawahara et al. 2003; Sun et al. 2005). In brief, quantitative PCR was performed with a LightCycler System (Roche Diagnostics, Mannheim, Germany). The PCR primers and probes for GluR1, GluR4 and β-actin comprised approximately 180–250 base pairs and matched sequences in the 3′ untranslated regions (UTR) in order to avoid the flip/flop and GluR4c alternative splicing sites. Moreover, phosphate groups were attached to the 3′ end of the LCRed640-containing probes to prevent probe extension. A set of standards and cDNA samples was amplified in duplicate in a reaction mixture (20 µL of total volume) containing 2 µL of 10 × FastStart Taq DNA polymerase, 4 mm MgCl2, each primer at 0.5 mm, 0.2 µm fluorescein isothiocyanate (FITC)-containing probe and 0.4 µm LCRed640-containing probe. Herring sperm DNA solution was included as a negative control in each series of reactions. The reactions were initiated by the activation of FastStart Taq DNA polymerase at 95°C for 10 min, and amplification was achieved by running 45 cycles of denaturation at 95°C for 1 s, annealing at 58°C for 10 s, and extension at 72°C for 12 s.
Differences between groups were evaluated by Mann–Whitney U-test. Significance was assumed at values of p < 0.05.
Behavioral and morphological changes in the rats infused with KA
Because infusion of 1.5 mm KA did not cause any change in the rat for up to 8 weeks of continuous infusion, and because infusion of 4.5 mm KA induced death as a result of convulsion shortly after operation (data not shown), we adopted the concentration of 3.0 mm KA except in the experiments in which KA was co-infused with CTZ. As compared with the aCSF-infused control rats (aCSF-W/F), which maintained a score of 420 s, the rotarod scores of the KA-infused rats significantly decreased in a time-dependent manner after 2 weeks of KA administration in both the KA-W and the KA-F rat groups, reaching a seventh of the control score after 8 weeks (KA-W: week 0, 420 s, n = 15; week 2, 237.3 ± 40.4 s, mean ± SEM, n = 15; week 4, 205.4 ± 45.9 s, n = 10; week 8, 67.2 ± 19.0 s, n = 5) (KA-F: week 0, 420 s, n = 10; week 2, 179.7 ± 48.4 s, n = 10; week 4, 176.2 ± 40.1 s, n = 5) (p < 0.001; Fig. 1). Moreover, rats infused with KA developed complete paralysis of the hindlimbs by the end stage, whereas none of the rats showed either sensory deficits or urinary incontinence (data not shown).
The motor neurons of the KA-F4 and the KA-W8 rats that displayed selective motor dysfunction exhibited morphological changes, including cytoplasmic microvacuolation, loss of Nissl substance and accumulation of glial cells around the degenerating neurons (Figs 2a and c), whereas neurons in the dorsal horn, including small neurons in the substantia gelatinosa of the same rats, appeared morphologically normal (Figs 2a and b). Axons in the fifth lumber ventral root of the KA-W8 rats were severely damaged and displayed a reduction in large myelinated fibers as compared with control rats (Figs 2d and e). Moreover, the number of large motor neurons was significantly reduced in the ventral horn of KA-F4 and KA-W8 rats (KA-F4, 4.8 ± 0.2; aCSF-F4, 6.2 ± 0.2; KA-W8, 3.2 ± 0.1; aCSF-W8, 6.0 ± 0.2; mean ± SEM; n = 5) (p < 0.001; Fig. 2f) but not in that of either KA-F2 or KA-W4 rats (p > 0.1; Fig. 2f). Taken together, these findings show that the KA-infused rats exhibited selective impairment in motor function, which is probably a result of the selective loss of motor neurons in the spinal cord.
Behavioral and morphological changes in the rats co-infused with KA and antagonists
Rats that were infused with KA and NBQX, a potent antagonist for AMPA and KA receptors, showed significantly higher rotarod scores after 3 weeks as compared with those infused with KA alone, whereas those infused with KA and APV, an NMDA receptor antagonist, did not differ from the KA-infused control rats (Fig. 3). By contrast, rats co-infused with CTZ, an AMPA receptor desensitization blocker, and the lower concentration of KA (1.5 mm KA + 1.5 mm CTZ) exhibited very low rotarod scores after 1 week, hence the co-infusion of CTZ and KA seemed to induce a more rapid decline of motor function as compared with the infusion of a higher concentration of KA alone (3 mm KA) (at week 1, p < 0.05; Fig. 3). Furthermore, rats that were infused with CTZ alone (1.5 mm CTZ) showed slightly, if any, and only transiently, lower rotarod scores as compared with aCSF-infused control rats (Fig. 3).
The morphometry of large neurons demonstrated that the decrease in the number of motor neurons was significantly less extensive in the KA/NBQX group than in the KA group after 4 weeks (KA, 4.8 ± 0.2; KA/NBQX, 5.6 ± 0.1; mean ± SEM; n = 5; p < 0.001), but not after 2 weeks (Fig. 4a). Although the concentration of KA in the KA/CTZ group was half of that in the KA-alone group, after 4 weeks the number of motor neurons in the KA/CTZ group tended to be lower than that in the KA group, and was significantly lower than that in the aCSF control and CTZ alone group (Fig. 4a). Moreover, motor neurons in the KA/CTZ group exhibited more severe morphological changes than did motor neurons in the KA-F4 (Fig. 4b) and KA-W8 groups (Fig. 2c), including numerous large cytoplasmic vacuoles with condensation of Nissl substance (Fig. 4c). After 4 weeks, the ventral roots of the KA/CTZ rats displayed marked axonal degeneration with loss of fibers as compared with the control rats (Figs 4d and e), which appeared rather more severe than the changes seen in the KA-W8 group (Fig. 2e).
Neuronal mRNA expression profile of AMPA receptor subunits
In motor neurons of the KA-infused groups (KA-W2, KA-W4, KA-W8, KA-F2 and KA-F4), the quantities of total AMPA receptor subunit mRNA (GluRs) expressed relative to the β-actin baseline were significantly greater than those in the respective aCSF control groups (aCSF-W2, aCSF-W4, aCSF-W8, aCSF-F2 and aCSF-F4) (Fig. 5a). Among the AMPA receptor subunits GluR1, GluR2, GluR3 and GluR4, only the quantity of GluR3 mRNA relative to β-actin was significantly increased in all of the KA-infused groups as compared with their respective aCSF control groups (KA-W vs. aCSF-W: 2 weeks, 39.6 ± 3.8 × 10−3 vs. 25.2 ± 2.1 × 10−3; 4 weeks, 41.8 ± 4.6 × 10−3 vs. 21.5 ± 1.0 × 10−3; 8 weeks, 44.8 ± 3.8 × 10−3 vs. 25.8 ± 1.8 × 10−3; KA-F vs. aCSF-F: 2 weeks, 33.8 ± 1.2 × 10−3 vs. 21.7 ± 2.0 × 10−3; 4 weeks, 46.3 ± 2.2 × 10−3 vs. 26.3 ± 1.4 × 10−3; mean ± SEM; n = 5 in each) (p < 0.001; Fig. 5a). Because the quantities of GluR1, GluR2 and GluR4 did not change in any of the KA-infused groups, the increase in total AMPA receptor subunit mRNA in these groups was caused by the selective increase in GluR3 mRNA. This molecular change resulted in a significant decrease in the proportion of GluR2 mRNA relative to the total AMPA receptor subunit mRNA in all of the KA-infused groups (Fig. 5b). By contrast, there was no difference in the mRNA expression profile of AMPA receptor subunits including GluR3 mRNA between the KA-infused rats and their controls in either the substantia gelatinosa or the funiculus of the spinal cord (Figs 5c and d).
The motor selective behavioral and neuropathological changes induced in this rat model are hallmarks of the clinicopathological changes seen in ALS. In addition, the delayed and progressive nature of these changes mimics ALS, suggesting that the KA-infused rat is a clinicopathologically appropriate model for ALS. The fact that motor deficit preceded the neuropathological changes by between 2 and 4 weeks may indicate the dysfunction of motor neurons prior to death and the slow death-inducing process. In addition, the results with Wistar and Fischer rats indicate that it takes around 4 weeks to induce the death of motor neurons by this method, and the motor neuron vulnerability is slightly different among the rat strains.
Our co-infusion experiments strongly suggest that the neuronal degeneration observed was mediated by AMPA receptors, in agreement with the results of KA toxicity in cultured rat hippocampal neurons (Ohno et al. 1997). The observation that co-infusion of CTZ exacerbated KA toxicity lends further support to a mechanism of AMPA receptor-mediated neurotoxicity. CTZ is a desensitization blocker of AMPA receptors, particularly of the flip splice variants (Partin et al. 1994), and does not exhibit neurotoxicity by itself as is observed in cultured neurons (May and Robison 1993; Brorson et al. 1995) but enhances KA-induced neurotoxicity with a leftward shift of the KA dose–response curve in cultured neurons (Ohno et al. 1998). Because both the flip and flop variants of each AMPA receptor subunit are expressed in rat spinal motor neurons (Tölle et al. 1993), CTZ probably exacerbates the neurotoxicity of KA via its desensitization effects on AMPA receptors.
We and other researchers have investigated the neurotoxic effects of various glutamate receptor agonists on spinal neurons in vivo and found that glutamate receptor agonists induce the degeneration of different subsets of spinal neurons depending on the route and the duration of administration (Kwak et al. 1992; Nakamura et al. 1994; Kwak and Nakamura 1995a,b; Hirata et al. 1997; Corona and Tapia 2004). Although intrathecal infusion of KA for a brief period induces long-standing damage in spinal motor neurons, it also severely affects interneurons (Kwak and Nakamura 1995b). Because interneurons appeared to be intact after two months of intrathecal KA infusion in the present study, it is likely that in accordance with the results in the cultured neurons (Terro et al. 1998), the subtoxic dose of KA that we used here may have selectively activated motor neurons without non-specifically activating interneurons.
An increase in intracellular Ca2+ concentration owing to an influx through Ca2+-permeable AMPA receptors has been demonstrated to play a pivotal role in AMPA receptor-mediated neuronal death in cultured motor neurons (Carriedo et al. 1996; Van Den Bosch et al. 2000). It is likely therefore that long-term administration of KA will induce an increase of Ca2+ influx through AMPA receptors, thereby causing the death of motor neurons. Here, we found that the expression of GluR3 mRNA expression was persistently increased in motor neurons, at least from week 2 of KA infusion when no motor neuron death was detected. Up-regulation of GluR3 mRNA was observed only in motor neurons and not in either dorsal horn neurons or in white matter, which morphologically remained intact throughout the experimental period, suggesting that this molecular change is probably caused by the long-term activation of motor neurons by KA. An increase in GluR3 mRNA has been reported in the motor neurons of mice transgenic for mutated human SOD1 (SOD1G93 A mice) (Spalloni et al. 2004), and these motor neurons display an increased vulnerability to excitotoxicity (Spalloni et al. 2004). In addition, the survival of these mice can be prolonged by the administration of GluR3 antisense protein nucleic acid (Rembach et al. 2004). These findings suggest that an increase in GluR3 mRNA in motor neurons is tightly associated with the mechanism underlying the selective degeneration of motor neurons.
GluR2 plays a critical role both in controlling the assembly and trafficking of AMPA receptors in hippocampal neurons (Sans et al. 2003), and in regulating Ca2+ permeability (Hollmann et al. 1991; Verdoorn et al. 1991; Burnashev et al. 1992). Furthermore, cultured neurons derived from GluR2-deficient mice are more vulnerable to excitotoxicity than those derived from wild-type animals (Iihara et al. 2001; Van Damme et al. 2005), and mice deficient for GluR2 RNA editing at the Q/R site die young as a result of premature neuronal death (Brusa et al. 1995). Therefore, an increase in GluR3 mRNA may result in a reduction in the relative proportion of GluR2 among AMPA receptor subunits, thereby increasing the proportion of GluR2-lacking, Ca2+-permeable AMPA receptors among the functional AMPA receptors expressed in the motor neurons of KA-infused rats (Fig. 5b). Indeed, the survival of human SOD1G93 A transgenic mice can be prolonged by the over-expression of GluR2 (Tateno et al. 2004), whereas mice that are additionally deficient for GluR2 show decreased survival as compared with those transgenic for human SOD1G93 A alone (Van Damme et al. 2005). Thus, it seems likely that long-term infusion of KA induces the degeneration of motor neurons via an AMPA receptor-mediated mechanism by increasing the proportion of Ca2+-permeable AMPA receptors in motor neurons; however, how KA increases GluR3 mRNA expression, or rather how KA selectively increases this expression in motor neurons while keeping that in dorsal neurons and white matter cells unchanged, remains to be elucidated. Because rat motor neurons express significantly lower quantities of GluR2 mRNA and higher quantities of GluR3 mRNA, as compared with other neuronal subsets (Sun et al. 2005), the up-regulation of GluR3 mRNA that presumably reflects an increased level of GluR3 protein may lead to a more marked reduction in the proportion of GluR2, and in turn to a higher proportion of GluR2-lacking, Ca2+-permeable AMPA receptors in motor neurons after KA infusion. The characteristics of motor neurons with a low relative abundance of GluR2 may explain, in part, the selective vulnerability of these cells to AMPA receptor-mediated excitotoxicity.
Another factor influencing the Ca2+ permeability of AMPA receptors is the status of GluR2 mRNA editing at the Q/R site. Recently, we have demonstrated that a significant reduction in RNA editing of GluR2 at the Q/R site occurs, in a disease-specific and neuronal class-selective manner, in the motor neurons of patients with sporadic ALS (Kawahara et al. 2004, 2006). In addition, the number of neurons was significantly reduced in the ventral horn of the spinal cord of mice transgenic for artificial Ca2+-permeable GluR-B(N) minigenes (Feldmeyer et al. 1999; Kuner et al. 2005). It is likely that this molecular change is closely relevant to the pathoetiology of ALS; however, we did not find any reduction in GluR2 editing efficiency in the motor neurons of KA-infused rats in this study (data not shown). Therefore, although long-term activation of AMPA receptors per se induces the death of motor neurons both in vitro and in vivo, the underlying mechanism is probably an alteration in the relative proportion of AMPA receptor subunits, and not a reduction in RNA editing. In addition to GluR3 up-regulation (Spalloni et al. 2004), a lack of reduction in GluR2 RNA editing (Kawahara et al. 2006) suggests the participation of an AMPA receptor-mediated neuronal death mechanism similar to that present in rats with mutated SOD1-associated familial ALS (ALS1). In this respect, the KA-infused rat is a model for ALS1 rather than sporadic ALS. There seems to be different Ca2+-permeable AMPA receptors that mediate neuronal death, including those caused by underedited GluR2 and those caused by a lack of GluR2. It seems likely that the underediting of GluR2 mRNA specifically seen in sporadic ALS motor neurons (Kawahara et al. 2004) is not caused by long-term AMPA receptor activation but rather by a defect intrinsic to motor neurons in patients with sporadic ALS.
We thank Dr Jun Shimizu for valuable discussion. This study was supported, in part, by a grant-in-aid for Scientific Research on Priority Areas from the Ministry of Education, Culture, Sports, Science and Technology of Japan (14017020, 15016030 and 16015228 to SK).
- 2004) No association of the SOD1 locus and disease susceptibility or phenotype in sporadic ALS. Neurology 63, 2419–2422. , , , , , , , , and (
- 1995) AMPA receptor desensitization predicts the selective vulnerability of cerebellar Purkinje cells to excitotoxicity. J. Neurosci. 15, 4515–4524. , , and (
- 1995) Early-onset epilepsy and postnatal lethality associated with an editing-deficient GluR-B allele in mice. Science 270, 1677–1680. , , , , , and (
- 1992) Calcium-permeable AMPA-kainate receptors in fusiform cerebellar glial cells. Science 256, 1566–1570. , , , , , , and (
- 1996) Motor neurons are selectively vulnerable to AMPA/kainate receptor-mediated injury in vitro. J. Neurosci. 16, 4069–4079. , and (
- 2004) AMPA receptor activation, but not the accumulation of endogenous extracellular glutamate, induces paralysis and motor neuron death in rat spinal cord in vivo. J. Neurochem. 89, 988–997. and (
- 1999) Neurological dysfunctions in mice expressing different levels of the Q/R. site-unedited AMPAR subunit GluR-B. Nat. Neurosci. 2, 57–64. , , , , , , , , , , et al. (
- 1994) Motor neuron degeneration in mice that express a human Cu,Zn superoxide dismutase mutation. Science 264, 1772–1775. , , , , , , , , , , et al. (
- 2001) Single cell analysis of CAG repeat in brains of dentatorubral-pallidoluysian atrophy (DRPLA). J. Neurol. Sci. 190, 87–93. , , , , , , , and (
- 1997) AMPA receptor-mediated slow neuronal death in the rat spinal cord induced by long-term blockade of glutamate transporters with THA. Brain Res. 771, 37–44. , , , and (
- 1991) Ca2+ permeability of KA-AMPA-gated glutamate receptor channels depends on subunit composition. Science 252, 851–853. , and (
- 2002) Focal loss of the glutamate transporter EAAT2 in a transgenic rat model of SOD1 mutant-mediated amyotrophic lateral sclerosis (ALS). Proc. Natl Acad. Sci. USA 99, 1604–1609. , , , , , , , , , , et al. (
- 2001) The influence of glutamate receptor 2 expression on excitotoxicity in Glur2 null mutant mice. J. Neurosci. 21, 2224–2239. , , , , , , , and (
- 1997) Copper/zinc superoxide dismutase 1 and sporadic amyotrophic lateral sclerosis: analysis of 155 cases and identification of a novel insertion mutation. Ann. Neurol. 42, 803–807. , , , , and (
- 2003) Human spinal motoneurons express low relative abundance of GluR2 mRNA: An implication for excitotoxicity in ALS. J. Neurochem. 85, 680–689. , , , , , , and (
- 2004) Glutamate receptors: RNA editing and death of motor neurons. Nature 427, 801. , , , , and (
- 2006) Underediting of GluR2 mRNA, a neuronal death inducing molecular change in sporadic ALS, does not occur in motor neurons in ALS1 or SBMA. Neurosci. Res. 54, 11–14. , , , , , , and (
- 2005) Late-onset motoneuron disease caused by a functionally modified AMPA receptor subunit. Proc. Natl Acad. Sci. USA 102, 5826–5831. , , , , , , , , , , et al. (
- 1995a) Acute and late neurotoxicity in the rat spinal cord in vivo induced by glutamate receptor agonists. J. Neurol. Sci. 129, 99–103. and (
- 1995b) Selective degeneration of inhibitory interneurons in the rat spinal cord induced by intrathecal infusion of acromelic acid. Brain Res. 702, 61–71. and (
- 1992) Acromelic acid, a novel kainate analogue, induces long-lasting paraparesis with selective degeneration of interneurons in the rat spinal cord. Exp. Neurol. 116, 145–155. , , and (
- 1993) Cyclothiazide treatment unmasks AMPA excitotoxicity in rat primary hippocampal cultures. J. Neurochem. 60, 1171–1174. and (
- 2001) Rats expressing human cytosolic copper-zinc superoxide dismutase transgenes with amyotrophic lateral sclerosis: associated mutations develop motor neuron disease. J. Neurosci. 21, 9246–9254. , , , , , , and (
- 1994) Late-onset selective neuronal damage in the rat spinal cord induced by continuous intrathecal administration of AMPA. Brain Res. 654, 279–285. , and (
- 1997) Concentration-dependent changes in motor behavior produced by continuous intrathecal infusion of excitatory amino acids in the rat spinal cord. Brain Res. Protocol. 1, 385–390. , , and (
- 1997) Kainate excitotoxicity is mediated by AMPA- but not kainate-preferring receptors in embryonic rat hippocampal cultures. Neurochem. Int. 31, 715–722. , , , and (
- 1998) Characterization of cyclothiazide-enhanced kainate excitotoxicity in rat hippocampal cultures. Neurochem. Int. 32, 265–271. , , , and (
- 1994) Cyclothiazide differentially modulates desensitization of alpha-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid receptor splice variants. Mol. Pharmacol. 46, 129–138. , and (
- 2004) Antisense peptide nucleic acid targeting GluR3 delays disease onset and progression in the SOD1 G93A mouse model of familial ALS. J. Neurosci. Res. 77, 573–582. , , , , , , , , , , et al. (
- 2005) Hereditary and acquired motor neuron diseases, in Merritt's Neuology, 11th edn (RolandL., ed.), pp. 861–870. Lippincott, Williams & Wilkins, Philadelphia. , and (
- 2003) Aberrant formation of glutamate receptor complexes in hippocampal neurons of mice lacking the GluR2 AMPA receptor subunit. J. Neurosci. 23, 9367–9373. , , , , , , , , and (
- 1991) RNA editing in brain controls a determinant of ion flow in glutamate-gated channels. Cell 67, 11–19. , , and (
- 2004) Cu/Zn-superoxide dismutase (GLY93 – >ALA) mutation alters AMPA receptor subunit expression and function and potentiates kainate-mediated toxicity in motor neurons in culture. Neurobiol. Dis 15, 340–350. , , , , , and (
- 2005) Expression profile of AMPA receptor subunit mRNA in single adult rat brain and spinal cord neurons in situ. Neurosci. Res. 52, 228–234. , , , and (
- 2004) Calcium-permeable AMPA receptors promote misfolding of mutant SOD1 protein and development of amyotrophic lateral sclerosis in a transgenic mouse model. Hum. Mol. Genet. 13, 2183–2196. , , , , , , , , , , et al.(
- 1998) Mild kainate toxicity produces selective motoneuron death with marked activation of CA (2+)-permeable AMPA/kainate receptors. Brain Res. 809, 319–324. , , , and (
- 1993) The differential expression of 16 NMDA and non-NMDA receptor subunits in the rat spinal cord and in periaqueductal gray. J. Neurosci. 13, 5009–5028. , , , and (
- 2005) GluR2 deficiency accelerates motor neuron degeneration in a mouse model of amyotrophic lateral sclerosis. J. Neuropathol. Exp. Neurol. 64, 605–612. , , , and (
- 2000) Ca (2+)-permeable AMPA receptors and selective vulnerability of motor neurons. J. Neurol. Sci. 180, 29–34. , , , and (
- 1991) Structural determinants of ion flow through recombinant glutamate receptor channels. Science 252, 1715–1718. , , , and (