Article Figures & Data
Figures
- Figure 1.
Subcellular fractionation of ventral lumbar spinal cord. A, Ventral quadrant of lumbar spinal cord tissue was dissected and homogenized using a Dounce homogenizer. The whole homogenate was then centrifuged for 5 min at 5000 rcf, and the supernatant (S1) was removed. This S1 fraction was then centrifuged for 30 min at 13,000 rcf. The pellet from this fraction (P2) was then used for all subsequent Western blots. B, Plasma membrane enrichment in the P2 fraction was confirmed with N-cadherin expression, and modest synaptic enrichment (synaptoneurosome) was characterized by PSD-95, with beta-actin serving as loading control. C, Tissue-oriented coordinate grid placement for systematic cell selection. Large ventral horn neurons were systematically selected from spinal cord slices in L4–L5 region. A microscopist blind to experimental condition entered the stage coordinate location of four anatomical landmarks (central canal, anterior artery, left edge of tissue, right edge of tissue) into an Excel spreadsheet, and a custom VisualBasic macro generated a list of microscope stage coordinates that were then input into MicroManager and ImageJ software that controlled microscope stage movement. This system ensured that the orientation of the coordinate grid would always be relative to the specific orientation of the tissue. Each coordinate signified the center of an 80 × 80μm sampling window, with each coordinate spaced 100 μm apart. The blind microscopist cycled through these sampling windows at 63×. When a large neuron (cell body >40 μm) was encountered within a sampling window, the cell was centered in the frame and a stack of images was taken through the z-plane, with separate images taken through a 650 nm filter (for synaptophysin) and a 490 nm filter (for GluA1 or GluA2) at each level. D, Optical detection of synaptic AMPAR subunits. Yellow pixels produced by the overlapping of the presynaptic marker synaptophysin (red) and postsynaptic AMPAR subunit (green) puncta indicated colocalization and were quantified to determine the amount of synaptic AMPAR subunit expression. E, All image stacks were combined and deconvolved to correct for the diffusion of light using AutoQuant software.
- Figure 2.
Plasma membrane GluA1 and GluA2 phosphorylation with intermittent nociceptive stimulation delivered below complete spinal cord injury. A, Intermittent nociceptive stimulation. Rats with complete spinal transections received 6 min of intermittent nociceptive stimulation to the tail. Unstimulated controls received an equivalent period of restraint. B, Quantitative fluorescent intensity optimized for linear detection of each target band using a 1:2 dilution curve of total protein. Laser scanning intensity for each target protein was chosen based on closest linear relationship between fluorescent intensity and total protein (all R 2 > 0.98). All subsequent analyses for each target protein were run at their respective optimal scanning intensities to ensure linearity of fluorescence (see Materials and Methods for details). C, Linear quantification of GluA1 and GluA2 AMPAR subunits, 20 min poststimulation. Stimulation significantly increased GluA1 expression compared to unstimulated controls (*p < 0.05), whereas GluA2 expression was unchanged. D, The ratio of GluA1 to GluA2 subunit expression was significantly increased in stimulated animals within 20 min (Mann–Whitney U; *p < 0.05). E, F, Linear quantification of GluA1 and GluA2 AMPAR subunits, 2 h poststimulation. Stimulation had no significant effect on GluA1 or GluA2 after 2 h (p > 0.05). G, Linear intensity optimization on phosphorylated serines 831 and 880. H, Linear quantification of p-S831 and p-S880 protein expression 20 min poststimulation. Stimulation significantly increased phosphorylated serine 831 expression relative to unstimulated controls (*p < 0.05), whereas phosphorylated serine 880 was unchanged. I, Stimulation had no significant effect on p-S831 or pS880 after 2 h (p > 0.05). All bars represent mean for n = 4 subjects/per group (n = 8 for main effects, n =4 for interaction) with three independent Western blot runs per subject. Error bars represent standard error of the mean.
- Figure 3.
Assessment of neuronal activity and cell death markers following nociceptive stimulation. Linear Western blot quantification of cell activity and death in cytosolic fractions of ventral lumbar spinal cord, assessed at (A) 20 m, (B) 2 h, or (C) 24 h after intermittent nociceptive stimulation. ANOVA showed no significant increase in broad neuronal activity marker cFos, in other more specific markers of apoptotic cell death (cJun, cleaved caspase3) or in calcium-mediated cell death (calpain I). No significant differences were observed for beta-actin loading control (p > 0.05). Bars represent mean for n = 4 subjects/factorial group (n =12 for INS main effect; n = 8 for time main effect; n = 4 for interaction) with three independent Western blot runs per subject. Error bars represent standard error of the mean.
- Figure 4.
Synaptic GluA1 and GluA2 expression in 3-D synaptic field surrounding ventral horn neurons after intermittent nociceptive stimulation below complete spinal cord injury. Large ventral horn neurons in the L4–L5 region were assessed for colocalization of GluA1/2 (green) to synaptophysin-positive synapses (red) in nociceptive-stimulated and unstimulated spinally transected animals. A–D, Representative 3-D confocal images of ventral horn neurons 2 h after stimulation or control demonstrating an increase in GluA1 expression (A, B, green), and decreased GluA2 expression (C, D, green) after stimulation. Scale bar, 30 μm. E, Quantification of GluA1 and GluA2 through confocal stacks at 20 min poststimulation shows a significant increase in synaptic GluA1 (*p < 0.05) and no change in GluA2. F, Quantification of GluA1 and GluA2 through confocal stacks at 2 h poststimulation shows a significant increase in synaptic GluA1 and a concomitant decrease in synaptic GluA2 (*p < 0.05). Bars represent mean colocalization through confocal z-series of ventral horn neurons (124–146 cells per group assessed for GluA1, 105–154 cells per group assessed for GluA2; n = 4 subjects/per group; N = 16 rats total). Error bars represent standard error of the mean.
- Figure 5.
Ext`n of GluA1 and GluA2 on plasma membrane of ventral horn neurons after intermittent nociceptive stimulation below complete spinal cord injury. Large L4–L5 ventral horn neurons in were assessed for colocalization (yellow) of GluA1/2 (green) to synaptophysin-positive synapses (red) after nociceptive stimulation. Algorithmically selected single confocal planes of peak GluA1/synaptophysin colocalization for unstimulated (A) and stimulated groups (B), and GluA2/synaptophysin colocalization for unstimulated controls (C) and stimulated groups (D). E–H, A 2-μm-wide cutout of the confocal image containing somatic plasma membrane. I–L, Boxed plasma membrane fractions enlarged to illustrate representative differences in extrasynaptic (green) and synaptic (yellow) GluA1/2 puncta on motor neuron plasma membranes. M–P, Quantification of extrasynaptic GluA1/2 puncta and synaptic colocalization of subunit puncta with synaptophysin. M, Extrasynaptic GluA1 was significantly increased 20 min after stimulation, whereas extrasynaptic GluA2 is significantly decreased (ANOVA, *p < 0.05). N, Synaptic colocalization of GluA1 and synaptophysin was also significantly increased (*p < 0.05), whereas synaptic GluA2/synaptophysin colocalization is unaltered by stimulation. O, Two hours after stimulation, extrasynaptic GluA1 expression is unchanged between stimulated and unstimulated groups, while extrasynaptic GluA2 remains significantly decreased in response to stimulation (*p < 0.05). P, Synaptic GluA1/synaptophysin colocalization is unchanged at 2 h poststimulation, but synaptic GluA2/synaptophysin colocalization is significantly decreased in response to stimulation (*p < 0.05). Bars represent means for 124–146 cells/group for GluA1, 105–154 cells per group for GluA2; n = 4 subjects/per group, N = 16 rats total. Error bars represent standard error of the mean.
- Figure 6.
Effect of CP-AMPAR antagonist on impaired adaptive sensorimotor performance following intermittent nociceptive stimulation. A, INS/spinal cord training paradigm. Rats with complete thoracic spinal transection received 6 min of INS to the tail followed by intrathecal administration of the CP-AMPAR antagonist Naspm (10 mm). Spinal instrumental training task began 20 min later. B, Vehicle-treated subjects failed to exhibit a progressive increase in response duration over time, indicative of the INS-induced impairment in spinal adaptation. Naspm increased response duration over time, indicating that blocking CP-AMPAR activity protects against INS-induced maladaptive spinal plasticity. ANOVA revealed a significant increase in response duration over time in the Naspm-treated group compared to vehicle-treated animals, n = 12 subjects/per group (repeated measures, p < 0.05). Error bars represent standard error of the mean.
- Figure 7.
A, Theoretical pathway underlying INS-induced maladaptive plasticity. Following afferent intermittent nociceptive stimulation, increased glutamate release engages postsynaptic AMPA receptors. Calcium influx via CP-AMPARs activates the calcium detectors PKC and/or CamKI phosphorylating the serine 831 site on GluA1 AMPAR subunit. Serine 831 phosphorylation increases the open probability of AMPARs, creating a feedforward loop that leads to membrane insertion of extrasynaptic CP-AMPARs. These receptors are trafficked laterally to the synaptic membrane, further strengthening this excitatory connection. B, Conceptual model of CP-AMPAR effects on spinal plasticity after SCI. CP-AMPAR activity critically shapes synaptic strength and use-dependent spinal cord plasticity after injury. Peripheral stimulation below the injury engages CP-AMPAR-mediated calcium influx, activating intracellular modulators of synaptic plasticity and strengthening excitatory tone to promote adaptive spinal training. However, CP-AMPARs are hyper-responsive to peripheral input (eg, limb positioning; skin stimulation) and are easily overdriven, resulting in synaptic saturation that overwhelms the capacity for adaptive spinal cord learning. As CP-AMPAR activity further increased, excitotoxicity and cell death may occur. Therapeutic intervention to decrease CP-AMPAR over-activity normalizes the balance of synaptic GluA1 and GluA2, and restores optimal adaptive plasticity.
Tables
Results Type of Test Effect Size (Eta Squared) Observed Power a GluA1 Western blot, Time × Stimulation interaction ANCOVA 0.231 0.940 b GluA2 Western blot, Time × Stimulation interaction ANCOVA 0.025 0.178 c GluA1: GluA2 ratio t test 0.592 0.694 d p-S831 Western blot, Time × Stimulation interaction ANCOVA 0.273 0.975 e p-S880 Western blot, Time × Stimulation interaction ANCOVA 0.067 0.406 f Western blot cell death markers ANCOVA <0.023 <0.170 g GluA1 neuropil expression, main effects of time and stimulation ANOVA Effect of time: 0.066 Effect of stimulation: 0.053 Effect of time: 1.000 Effect of stimulation: 1.000 h GluA2 neuropil expression, Time × Stimulation interaction ANOVA 0.036 0.991 i GluA1 extrasynaptic membrane expression, Time × Stimulation interaction ANOVA 0.016 0.847 j GluA1 synaptic expression, main effects of time and stimulation ANOVA Effect of time: 0.111 Effect of stimulation: 0.015 Effect of time: 1.000 Effect of stimulation: 0.800 k GluA2 extrasynaptic membrane expression, main effects of time and stimulation ANOVA Effect of time: 0.030 Effect of stimulation: 0.125 Effect of time: 0.973 Effect of stimulation: 1.000 l GluA2 synaptic expression, Time × Stimulation interaction ANOVA 0.035 0.988 m Response duration, sensorimotor learning task, Time × Drug interaction ANOVA, repeated measures 0.069 0.991
In this issue
