Interlaminar difference in chloride extrusion capacity in the spinal dorsal horn

We have previously shown that a decrease in KCC2 activity in the spinal dorsal horn (DH) following peripheral nerve injury weakens Cl−-mediated hyperpolarization, compromising local inhibitory control and causing pain hypersensitivity (Coull et al. Nature 2003, 2005). We have also shown that neurons with weaker Cl- extrusion capacity at certain developmental stages are more prone to a collapse in hyperpolarizing inhibition in conditions of repetitive activity (Cordero-Erausquin et al. J. Neurosci. 2005). Here, we asked whether specific DH cell populations show differences in Cl− extrusion capacity in normal conditions and thus may be more or less sensitive to Cl− accumulation when challenged. Spinal cord slices were obtained from adult rats. Whole cell patch clamp recordings were obtained from lamina I (LI) and II (LII) neurons using low (9 mM) or high (29 mM) Cl−-containing pipettes. The reversal potential of GABA (EGABA) was extrapolated from I-V curves obtained by puffing GABA (1 mM, 30 ms) at different membrane potentials in presence of CNQX, APV and TTX. After recording, some neurons were filled with Lucifer yellow to allow laminar localization. Recordings from LI neurons in high Cl− condition yielded measured EGABA more negative than the expected value of -37 mV (-45±1 mV, n=38). Values were distributed along a wide range (-35 mV to -59 mV), indicative of a substantial heterogeneity in Cl− extrusion capacity between cells. Furosemide (100 µM), a blocker of KCC2 and NKCC1, brought EGABA close to the expected value (-39±2 mV; n=15, p<0.001), while bumetanide (10µM), a more specific blocker of NKCC1, had no effect, indicating that Cl- transport was predominantly KCC2-mediated. In LII, EGABA measured using high Cl- pipettes was more negative than in LI (-55±2 mV, n=14; range -41 mV to -69 mV; p<0.01) and it shifted close to the expected value in presence of furosemide, suggesting higher expression of KCC2 at this level. This hypothesis was supported by quantitative KCC2 immunoreactivity analysis. Quantification was performed by measuring the mean fluorescence intensity in lamina I and II. Signal was significantly higher in lamina II with a lamina II/lamina I ratio of 146%±4%. The interlaminar difference was reduced when low Cl- pipettes were used (LI: -73±2 mV, n=6; LII: -74±1 mV, n=8; P=0.5). Gramicidine-perforated patch showed no differences in EGABA between LI and LII. Consistent with their weaker Cl-extrusion capacity, LI neurons showed a more prominent collapse of IPSC amplitude when bombarded by repetitive stimulation and this collapse was accelerated in the presence of benzodiazepines, likely due to enhanced Cl-load via GABAA channels. Thus, GABAA and glycine-mediated inhibition is more labile in LI than LII when challenged by sustained input, making them more prone to activity dependent sensitization and thus an important substrate for the development of pain hypersensitivity.