Intrathecal infusion of a Ca(2+)-permeable AMPA channel blocker slows loss of both motor neurons and of the astrocyte glutamate transporter, GLT-1 in a mutant SOD1 rat model of ALS

Abstract

Elevated extracellular glutamate, resulting from a loss of astrocytic glutamate transport capacity, may contribute to excitotoxic motor neuron (MN) damage in ALS. Accounting for their high excitotoxic vulnerability, MNs possess large numbers of unusual Ca(2+)-permeable AMPA channels (Ca-AMPA channels), the activation of which triggers mitochondrial Ca(2+) overload and strong reactive oxygen species (ROS) generation. However, the causes of the astrocytic glutamate transport loss remain unexplained. To assess the role of Ca-AMPA channels on the evolution of pathology in vivo, we have examined effects of prolonged intrathecal infusion of the Ca-AMPA channel blocker, 1-naphthyl acetylspermine (NAS), in G93A transgenic rat models of ALS. In wild-type animals, immunoreactivity for the astrocytic glutamate transporter, GLT-1, was particularly strong around ventral horn MNs. However, a marked loss of ventral horn GLT-1 was observed, along with substantial MN damage, prior to onset of symptoms (90-100 days) in the G93A rats. Conversely, labeling with the oxidative marker, nitrotyrosine, was increased in the neuropil surrounding MNs in the transgenic animals. Compared to sham-treated G93A animals, 30-day NAS infusions (starting at 67+/-2 days of age) markedly diminished the loss of both MNs and of astrocytic GLT-1 labeling. These observations are compatible with the hypothesis that activation of Ca-AMPA channels on MNs contributes, likely in part through oxidative mechanisms, to loss of glutamate transporter in surrounding astrocytes.

1. Progressive motor neuron degeneration in G93A rats: attenuation by Ca-AMPA channel blockade

A. Motor neuron morphology: Photomicrographs (200x) show ventral horn regions of lumbar spinal cord sections derived from 97±2 day old rats (upon termination of the 30 day intrathecal infusions), stained for Nissl substance. In contrast to the healthy appearing MNs in the control (WT) condition, in transgenic animals treated with saline infusion (Tg-S), there is moderate MN injury characterized variably by cell shrinkage and fragmentation, eccentric nuclei, and microglial infiltration. The infusion of NAS into the transgenic animals (Tg-NAS) decreased MN injury. In symptomatic transgenic animals (Tg-ES), MN injury was far greater than in the presymptomatic animals, with >70% MN loss or severe injury. Bar = 100 μm. B. Quantification of injury: Graph shown quantification of MN injury, based upon cell counts from 7 independent experiments (12-13 animals, > 10 slices counted / animal each condition). # indicates difference from WT, * indicates difference from Tg-S by 2-tailed t test (p<0.01).

2. Marked changes in astrocytes surrounding ventral horn MNs in presymptomatic G93A rats

Lumbar spinal cord sections were obtained from 97±2 day old wild type and G93A transgenic rats and stained for glial fibrillary acidic protein (GFAP, green) and SMI-32 (red, A), 3-nitrotyrosine (B), or the astrocytic glutamate transporter, GLT-1 (green) along with SMI-32 (red, C). Note the increased number of reactive astrocytes near transgenic MNs indicated by strong GFAP labeling. Further note the nitrotyrosine labeling immediately surrounding transgenic MNs, consistent with the possibility that the labeling is the result of ROS produced within MNs (the image is shown in pseudocolor, with white and red highest and purple lowest in order to highlight the gradient in labeling intensity). Finally note the rim of strong GLT-1 labeling surrounding MNs in wild type animals and the distinct loss of labeling surrounding MNs in the transgenics.

3. Progressive astrogliosis and nitrotyrosine labeling of G93A spinal cord

A. Photomicrographs show the ventral horn region of lumbar spinal cord sections derived from 97±2 day old rats (upon termination of the 30 day intrathecal infusions), stained for glial fibrillary acidic protein (GFAP; 400x) or for 3-nitrotyrosine (NT; 400x, inserts 40x). Note the marked increase in numbers of GFAP-positive reactive astrocytes in transgenic animals (Tg-S), the modest attenuation of this astrogliosis in the presence of NAS (Tg-NAS), and the sharp further increase in astrogliosis in symptomatic (Tg-ES) animals. The irregular areas of NT staining in wild type animals (WT) largely represent capillary endothelium. Note the marked increase in NT labeling in the transgenic animals (Tg-S), initially most prominent within and surrounding ventral horn MNs, with a further increase in staining of neuropil and of reactive astrocytes throughout the gray matter in symptomatic (Tg-ES) animals. Arrow shows early staining in ventral horn of transgenic animals. Bar = 50 μm. B. Quantification of labeling: Graphs show quantification of GFAP and NT labeling. GFAP labeling was quantified as the number of distinct GFAP-positive astrocytes per high power (400x) field, whereas NT staining was quantified as the mean labeling intensity in 25 μm zones surrounding each MN (GFAP counts based upon 6 independent experiments, > 20 fields for WT, Tg-S and Tg-NAS conditions; 3 experiments, 12 fields for Tg-ES. NT values based upon cell counts from 3 independent experiments, >70 surround regions for WT and Tg-S conditions; 3 animals, >70 regions for Tg-ES condition), # indicates difference from WT, * indicates difference from Tg-S by 2-tailed t test (p<0.01).

4. Loss of GLT-1 in ventral horn astrocytes of G93A rats and preservation by NAS

A. GLT-1 immunofluorescence: Photomicrographs show the ventral horn region of lumbar spinal cord sections derived from 97±2 day old rats (upon termination of the 30 day intrathecal infusions), stained for GLT-1 (40, 400x). Note the strong GLT-1 labeling surrounding MNs and throughout ventral horn MN clusters in wild type (WT) animals, in contrast to the preferential loss of ventral horn labeling, with preservation of dorsal horn labeling in transgenic animals (Tg-S). Also note the marked preservation of the ventral horn labeling with NAS treatment (Tg-NAS). Arrows show clusters of ventral horn motor neurons with strong astrocytic GLT-1 labeling; arrowhead shows persistent strong GLT-1 labeling in dorsal horn of untreated transgenic animals. Bar = 500 μm (left column) or 50 μm (right column). B. Quantification of labeling: Graph shown quantification of GLT-1 labeling in 5 μm zones surrounding each MN (compiled from 4 independent experiments, > 100 surround regions for WT, Tg-S, and Tg-NAS conditions; 3 animals, >60 regions for Tg-ES condition). # indicates difference from WT, * indicates difference from Tg-S by 2-tailed t test (p<0.01).