Early blockade of injured primary sensory afferents reduces glial cell activation in two rat neuropathic pain models

Abstract

Satellite glial cells in the dorsal root ganglion (DRG), like the better-studied glia cells in the spinal cord, react to peripheral nerve injury or inflammation by activation, proliferation, and release of messengers that contribute importantly to pathological pain. It is not known how information about nerve injury or peripheral inflammation is conveyed to the satellite glial cells. Abnormal spontaneous activity of sensory neurons, observed in the very early phase of many pain models, is one plausible mechanism by which injured sensory neurons could activate neighboring satellite glial cells. We tested effects of locally inhibiting sensory neuron activity with sodium channel blockers on satellite glial cell activation in a rat spinal nerve ligation (SNL) model. SNL caused extensive satellite glial cell activation (as defined by glial fibrillary acidic protein [GFAP] immunoreactivity) which peaked on day 1 and was still observed on day 10. Perfusion of the axotomized DRG with the Na channel blocker tetrodotoxin (TTX) significantly reduced this activation at all time points. Similar findings were made with a more distal injury (spared nerve injury model), using a different sodium channel blocker (bupivacaine depot) at the injury site. Local DRG perfusion with TTX also reduced levels of nerve growth factor (NGF) in the SNL model on day 3 (when activated glia are an important source of NGF), without affecting the initial drop of NGF on day 1 (which has been attributed to loss of transport from target tissues). Local perfusion in the SNL model also significantly reduced microglia activation (OX-42 immunoreactivity) on day 3 and astrocyte activation (GFAP immunoreactivity) on day 10 in the corresponding dorsal spinal cord. The results indicate that early spontaneous activity in injured sensory neurons may play important roles in glia activation and pathological pain.

Figure 1. Early DRG neuronal blockade with TTX reduced satellite glia activation after spinal nerve ligation

Sections of DRG stained for GFAP (red) and Neu-N (Green) a, b: GFAP was barely detectable in normal rat DRG. Following SNL, the expression of GFAP was dramatically increased on POD 1 (c), and remained very high on POD 3 (e) and POD 6 (g). Applying TTX locally to the axotomized DRG for 7 days starting at the time of nerve injury reduced the nerve-injury induced GFAP expression at all these time points. Scale bar = 10 µm

Figure 2. Time course of local nerve blockade effects on satellite glia activation in DRG

Top: Effect of SNL with and without nerve blockade (via perfusion of the axotomized L4 DRG with TTX) on satellite glia activation (percent of neurons surrounded by GFAP-positive satellite glia) in L4 DRG. Bottom: effect of SNI with and without nerve blockade (via bupivacaine depot placed at the nerve injury site) on glia activation in L4 DRG. #, glia activation at these time points was significantly different from control (POD 0). * glia activation significantly different from the same time point with no blockade.

Figure 3. Early DRG neuronal blockade via TTX reduced NGF expression following peripheral nerve injury

Significant expression of NGF can be found in cellular areas of normal DRG (a, b). Following SNL, NGF immunoreactivity in axotomized DRG was markedly decreased on POD 1 (c), however, it recovered back to normal on POD 3 (e), then decreased again on POD 7 (g). Qualitatively similar changes in NGF were found in the axotomized DRG blocked with TTX starting at the time of nerve injury. However, after POD 1 (d) the overall level of NGF immunoreactivity in DRG perfused with TTX (f & h) was lower than in the DRG without nerve blockade. i & j were stained as the negative control. They were processed under the same conditions as the other panels except that no anti-NGF was used. Scale bar = 10 µm.

Figure 4. Effect of SNL with and without nerve blockade (TTX applied to axotomized DRG) on NGF immunoreactivity in the axotomized L4 DRG

*, group differed significantly from control. #, significant effect of TTX for that POD.

Figure 5. Early DRG blockade via TTX reduced microglia activation caused by spinal nerve ligation in the spinal cord

In the normal spinal cord, a basal level of OX-42 is expressed in microglia (a & b). Spinal nerve ligation caused a significant increase of OX-42 expression on the side of spinal cord ipsilateral to the nerve injury (d), more so than on the contralateral side (c) on POD 3. Early DRG neuronal blockade via TTX perfusion for the first 3 days after SNL reduced spinal OX-42 expression (e & f) after SNL. The panels on the left and the right side show dorsal horn portions of the spinal cord, the sides contra- and ipsi- lateral to nerve injury, respectively. Scale bar = 50 µm.

Figure 6. Effect of perfusion of the axotomized DRG with TTX on glia activation in spinal cord

In control animals no differences were observed between contralateral and ipsilateral sides, so data have been combined into one point. Top: Microglia marker OX-42 was examined on POD 3. Bottom: Astrocyte marker GFAP was examined on POD 10. *, significantly different from control. #, significant differences between the indicated pairs of groups.

Figure 7. Early DRG blockade via local TTX perfusion reduced the expression of GFAP in the spinal cord caused by spinal nerve ligation

In the normal spinal cord, a basal level of GFAP is expressed in astrocytes (a & b). Spinal nerve ligation caused significant increase of GFAP expression on the side of spinal cord ipsilateral to the nerve injury (d), more so than on the contralateral side (c) on POD 10. Early DRG neuronal blockade via TTX perfusion for the first 7 days after SNL reduced spinal GFAP expression (e & f) observed on POD10 after SNL. Sections shown in g & h were stained as a negative control, processed in the same way as the preceding panels except that no primary antibody (anti-GFAP) was used. The panels on the left and the right side show spinal cord dorsal horn regions contra- and ipsi- lateral side to nerve injury, respectively. Scale bar = 50 µm.