Figure 1.
Opening a spinal gate for nociception.
(A) The classical gate control theory: Afferent nociceptive fibers (Aδ- and C-fibers) directly excite transmission neurons via excitatory synapse E1. Collaterals of nociceptive afferents inhibit inhibitory neurons through a proposed inhibitory synapse I1. Activity in Aδ- and C-fibers during nociception would thus depress activation of the inhibitory interneurons. This would lead to a reduced (↓↓) pre- (I2) or postsynaptic inhibition (not shown) of nociceptive transmission neurons in the spinal dorsal horn. The existence of the proposed inhibitory synapse (I1) has, however, not been shown experimentally. (B) The present and previous studies [10], [16] have demonstrated monosynaptic excitatory input (E3) to GABAergic neurons from Aδ- and C-fiber afferents. We tested the hypothesis, that in neuropathy global excitatory drive (E2 and E3) including direct excitation from presumably nociceptive Aδ- and C-fibers to GABAergic neurons is impaired. This would lead to a reduced inhibition at I2 (↓↓) and thus open the spinal gate for nociception (↑↑ E1). E2: other excitatory input from local circuits or descending pathways. (color code: green: inhibitory synapse; red: excitatory synapse)
Figure 2.
Time course of thermal (A) and mechanical (B) withdrawal thresholds after CCI.
Withdrawal thresholds in CCI (n = 19) and sham-operated (n = 17) animals for both ipsi- and contralateral hindpaws. Tests were performed on days 2 and 1 before operation and on days 1, 3, 5, 7 and 9 after surgery. (A) Thermal hyperalgesia is indicated by a significant reduction in the paw withdrawal latency. (B) Mechanical hyperalgesia is indicated by a significant reduction of the 50% paw withdrawal threshold. Data are expressed as mean ± s.e.m. (* P < 0.05; ** P < 0.01, two-way ANOVA).
Figure 3.
Effects of CCI on mEPSC rates and amplitudes recorded from EGFP-labeled, GABAergic neurons.
Representative traces of mEPSCs at a holding potential of -75 mV in sham- (A) and CCI-treated (B) animals (scale bar: 10 s; 50 pA). (C) CCI treated mice exhibited a reduced rate in mEPSCs recorded from GABAergic neurons (n = 34) as compared to sham-operated animals (n = 26 neurons; *** P < 0.001, Mann-Whitney rank sum test). Data: mean ± s.e.m. (D) Corresponding cumulative probability plot of inter-event-intervals of mEPSCs. (E) Amplitudes of mEPSCs were not significantly changed in CCI treated mice (n = 34 neurons) as compared to sham control (n = 26 neurons, P>0.1, Mann-Whitney rank sum test). Data: mean ± s.e.m. (F) Corresponding cumulative probability plot of mEPSC amplitudes.
Figure 4.
3D reconstruction of dendritic spines of GABAergic neurons.
(A, B) Representative examples of calcein filled GABAergic neurons from a sham-operated mouse (A) and a CCI-operated mouse (B). (A1, B1) 2D projection of original images of filled neurons (scale bars: 20 µm, left = caudal, right = rostral). (A2, B2) Shadow projection of same neurons. (A3, A5, A7, B3, B5, B7) dendritic segments at higher magnification (scale bars: 5 µm). (A4, A6, A8, B4, B6, B8) color coded reconstruction of dendritic spines (red) for morphological analysis (scale bars: 5 µm).
Table 1.
Density and morphological parameters of dendritic spines on GABAergic neurons.
Figure 5.
Immunohistochemical triple stainings to quantify excitatory synapses on GABAergic neurons in lamina II of the spinal dorsal horn.
Confocal images depicting co-staining of GAD67 (green) with the postsynaptic marker PSD-95 (blue) and the presynaptic marker synaptophysin (red). Bottom right, white arrowheads in enlarged boxed areas indicate colocalization of the three antigens. Scale bars: 10 µm (2 µm for enlarged boxes).
Table 2.
Quantification of colocalization of GAD67 with the postsynaptic marker PSD-95 and the presynaptic marker synaptophysin in spinal lamina II.
Figure 6.
Paired-pulse ratio recorded from EGFP-labeled, GABAergic neurons following dorsal root stimulation is increased after CCI treatment.
(A) Overlaid monosynaptically evoked Aδ-fiber EPSCs during 10 Hz stimulation (scale bars: 10 ms, 50 pA). (B) Averaged traces of paired-pulse recordings following Aδ-fiber stimulation at 50 ms interstimulus interval (scale bars: 20 ms, 100 pA). (C) Paired-pulse ratio following Aδ-fiber stimulation was increased in CCI treated mice at 50 ms interstimulus interval (sham n = 19, CCI n = 20, ** P < 0.01, two-way ANOVA) and at 300 ms interstimulus interval (sham n = 21, CCI n = 20, * P < 0.05, two-way ANOVA). Data: mean ± s.e.m. (D) Overlaid monosynaptically evoked C-fiber EPSCs during 1 Hz stimulation (scale bars: 10 ms, 50 pA). (E) Averaged traces of paired-pulse recordings following C-fiber stimulation at 300 ms interstimulus interval (scale bars: 100 ms, 100 pA). (F) Paired-pulse ratio following C-fiber stimulation was increased in CCI treated mice at 300 ms interstimulus interval (sham n = 13, CCI n = 13, ** P < 0.01, two-way ANOVA) and at 500 ms interstimulus interval (sham n = 16, CCI n = 16, * P < 0.05, two-way ANOVA). Data: mean ± s.e.m.
Figure 7.
mEPSCs and PPRs recorded from EGFP-labeled, GABAergic neurons with BAPTA included in recording pipette.
mEPSCs from GFP-labeled neurons with intracellular BAPTA did not show statistically significant differences between CCI-treated and sham-operated mice in rate (A: sham n = 23; CCI n = 21, P>0.1, Mann-Whitney rank sum test) and amplitude (C: sham n = 23; CCI n = 21, P>0.1, Mann-Whitney rank sum test) – compare with Fig. 3. B and D: Corresponding cumulative probability plots of inter-event-intervals (B) and amplitudes (D) of mEPSCs. (E) and (F) Paired-pulse ratios were not different in CCI-treated mice as compared to sham-operated control mice in none of the interstimulus intervals tested neither following Aδ-fiber stimulation (sham: n = 13, CCI: n = 16; E) nor C-fiber stimulation (sham: n = 19, CCI: n = 12; F) (P>0.1 for all groups, two-way ANOVA). Data: mean ± s.e.m.
Table 3.
Bath application of drugs did not change mEPSC rates.
Figure 8.
c-Fos expression in GABAergic neurons.
Upper row: c-Fos expression after noxious heat stimulus in a sham-treated mouse colocalizes with EGFP-labeled GABAergic neurons. Lower row: In a CCI-operated mouse c-Fos can be detected in fewer EGFP-labeled neurons as compared to a sham-treated mouse. (scale bar: 50 µm, left = lateral, right = medial) Border between dorsal horn and white matter is indicated by a solid line, lamina borders by a dashed line. Lamina II is indicated by L II.
Table 4.
c-Fos expression in GABAergic neurons following a noxious heat stimulus was reduced in CCI-treated mice.