Neonatal nociceptive circuits and dorsal horn cells are characterized by an apparent lack of inhibitory control: receptive fields are large and thresholds low in the first weeks of life. It has been suggested that this may reflect immature GABAA-receptor (GABAAR) signaling whereby an early developmental shift in transmembrane anion gradient is followed by a longer period of low Cl− extrusion capacity. To investigate whether functional GABAAR-mediated inhibition does indeed undergo postnatal regulation at the level of dorsal horn circuits, we applied the selective GABAAR antagonist gabazine to the spinal cord in anesthetized rat pups [postnatal day (P) 3 or 21] while recording spike activity in single lumbar dorsal horn cells in vivo. At both ages, blockade of GABAAR activity resulted in enlarged hind paw receptive field areas and increased activity evoked by low- and high-intensity cutaneous stimulation, revealing comparable inhibition of dorsal horn cell firing by spinal GABAARs at P3 and P21. This inhibition did not require descending pathways to the spinal cord because perforated patch-clamp recordings of deep dorsal horn neurons in P3 spinal cord slices also showed an increase in evoked spike activity after application of gabazine. We conclude that spinal GABAergic inhibitory transmission onto single dorsal horn cells “in vivo” is functional at P3 and that low Cl− extrusion capacity does not restrict GABAergic function over the normal range of evoked sensory activity. The excitability of neonatal spinal sensory circuits could reflect immaturity in other intrinsic or descending inhibitory networks rather than weak spinal GABAergic inhibition.
γ-Aminobutyric acid (GABA) is a major inhibitory neurotransmitter in the adult mammalian CNS and GABAergic interneurons in the spinal dorsal horn play an important role in the processing of tactile and nociceptive information. Spinal application of the GABAA-receptor (GABAAR) antagonist bicuculline leads to a robust disinhibition of adult rat dorsal horn neurons and a reduction in cutaneous mechanical thresholds (Cronin et al. 2004; Ishikawa et al. 2000; Reeve et al. 1998; Sivilotti and Woolf 1994). Newborn dorsal horn cells have lower mechanical thresholds and larger receptive fields than those in the adult (Andrews and Fitzgerald 1994; Fitzgerald 1985; Holmberg and Schouenborg 1996; Torsney and Fitzgerald 2002), which could result from a delayed maturation of inhibitory circuits in the neonatal dorsal horn. Indeed, recent work in vitro has demonstrated a significant increase in spontaneous and evoked GABAergic transmission within lamina II of the dorsal horn during the first two postnatal weeks (Baccei and Fitzgerald 2004).
In addition to the above postnatal changes, the modulation of intracellular chloride concentration ([Cl−]i) is likely to be critical for the maturation of inhibitory circuits in the dorsal horn. In many areas of the CNS, immature neurons have a relatively high [Cl−]i because of low levels of the neuronal KCl cotransporter KCC2, producing an ECl more positive than the resting membrane potential (Ehrlich et al. 1999; Rivera et al. 1999). Thus GABAAR activation often causes an efflux of chloride ions and depolarization of the cell, which can result in action potential generation (Ben Ari et al. 1989; Kullmann and Kandler 2001). Patch-clamp studies in spinal cord slices have shown that GABAAR activation does evoke subthreshold depolarizations in a subset of newborn lamina II neurons but these disappear by postnatal days 6 to 7 (P6–P7) and no evidence was found for a strong GABAergic excitatory drive in lamina II (Baccei and Fitzgerald 2004). It has recently been reported, however, that neonatal dorsal horn neurons have a low chloride extrusion capacity such that beyond the first postnatal week, GABAergic hyperpolarization is followed by a rebound depolarization and a rise in [Ca2+]i in vitro. It has been proposed that this not only causes disinhibition of spinal sensory circuits in the neonate but also underlies nociceptive hypersensitivity in infant rats (Cordero-Erausquin et al. 2005).
We have used a selective GABAAR antagonist to unmask the actions of endogenously released GABA to examine whether functional GABAAR-mediated inhibition within intact dorsal horn circuits is significantly weaker at early postnatal ages. We have thus examined the effect of the GABAAR antagonist gabazine on dorsal horn cell spike activity and receptive field size in response to natural cutaneous stimulation at different postnatal ages in vivo. The data clearly show that antagonism of spinal GABAARs results in a significant increase in neuronal excitability throughout the postnatal period. Using in vitro perforated patch-clamp recordings in P3 slices, we have verified that the effect of gabazine on the firing of immature dorsal horn neurons did not require the presence of descending inputs to the spinal cord. This demonstrates that intrinsic spinal GABAergic inhibitory signaling is functional from birth.
Sprague–Dawley rats of both sexes and aged P3 and P21 were used in this study. All experiments were conducted in accordance with the United Kingdom Animals (Scientific Procedures) Act of 1986.
In vivo electrophysiology
Rats were anesthetized with a single intraperitoneal (ip) injection of 2.0–2.5 g/kg urethane (Sigma, Dorset, UK) in saline, then tracheotomized, and secured above a heated blanket with ear and hip bars. The heart rate was monitored throughout by electrocardiogram. The lumbar spinal cord was exposed and held stable with a rostral vertebral clamp and the dura removed.
Extracellular recordings were made with 10-μm-tip glass-coated tungsten microelectrodes, lowered onto the cord surface under microscopic vision, then through the cord in 2- or 10-μm steps by a microdrive. Cells were classed as superficial (laminae I and II; depth ≤200 μm from the dorsal surface at P3, ≤300 μm at P21) or deep (laminae III, IV, V). Stroking of the plantar skin of the hind paw was used as a search stimulus. Single spikes were isolated and background activity noted before the receptive field (RF) was mapped with a pointed cotton swab and calibrated nylon hairs [von Frey hairs (vFh)]. Baseline firing to 3-s threshold and suprathreshold (three hairs above threshold) vFh stimulation was recorded for ≥10 min, before 5 μl of 20 μM gabazine (SR-95531; Sigma) in saline was applied topically to the exposed cord. This likely represents a maximal concentration of gabazine because previous in vitro work from our laboratory and others has documented that lower (3–10 μM) concentrations of gabazine abolished synaptic GABAAR-mediated currents in both the immature and mature dorsal horn (Chery and De Koninck 1999). Spontaneous activity, RF area (measured as a percentage of the total plantar area), and response to threshold and suprathreshold vFh were measured ≤60 min after gabazine application. To be classed as showing an increase or a decrease in responsiveness after gabazine, the change from baseline had to be ≥20% and be sustained over a minimum of two consecutive time points. The effect of gabazine on threshold and suprathreshold vFh was similar and thus only data from the suprathreshold stimulation are presented.
Response to noxious heat was tested in six separate cells by applying steady 4-s jets of 48°C water to the center of the RF both before (baseline) and 10 min after application of gabazine. A 30°C jet was used to control for activation of pressure and touch receptors. A minimum of two baseline measurements (at both 30 and 48°C) were taken for each cell to ensure a stable baseline and to rule out the possibility of sensitization with repeated application of noxious heat. We counted the total number of spikes fired in response to each stimulus, with the length of the window for spike counting kept constant within each cell. Heat-evoked firing rates (Hz) were monitored in 0.5-s bins.
Recordings were fed into Chart software (AD Instruments, Chalgrove, Oxfordshire, UK) and data were analyzed in Prism 3.0 (GraphPad Software, San Diego, CA) and Minitab (Minitab, State College, PA). Animals were killed with an overdose of sodium pentobarbital (ip) at the end of the experiment.
Patch-clamp recordings in spinal cord slices
Neonatal Sprague–Dawley pups (P2–P3) were given an overdose of halothane (5% in medical oxygen) and decapitated. The spinal column was quickly removed and placed in an ice-cold dissection solution consisting of (in mM): 250 sucrose, 2.5 KCl, 25 NaHCO3, 1.0 NaH2PO4, 6 MgCl2, 0.5 CaCl2, 25 glucose, and continuously bubbled with 95% O2-5% CO2. The spinal cord was immersed in low melting point agarose (3% in above solution; GibcoBRL; Paisley, UK) and a sagittal slice (400–600 μm) was cut using a modified Vibroslice tissue slicer (HA-752; Campden Instruments, Leicester, UK), preserving the entire L4–L5 dorsal root entry zone with the L4 and L5 dorsal root ganglia and sciatic nerve attached. The slice was transferred to a chamber filled with oxygenated dissection solution for 30 min then allowed to recover for a minimum of 1 h at room temperature in an oxygenated artificial cerebrospinal fluid (aCSF) solution containing (in mM): 125 NaCl, 2.5 KCl, 25 NaHCO3, 1.0 NaH2PO4, 1.0 MgCl2, 2.0 CaCl2, and 25 glucose.
Dorsal horn neurons were visualized with IR-DIC, and perforated patch-clamp recordings were obtained as described previously (Baccei and Fitzgerald 2004) using an electrode solution consisting of (in mM): 130 K-gluconate, 10 KCl, 10 HEPES, 1.0 EGTA, 0.1 CaCl2, and 2.0 MgATP, pH 7.2 (305 mOsm) with gramicidin added at a final concentration of 25 μg/ml. Synaptic responses were evoked by electrical stimulation of the sciatic nerve (1 mA, 1 ms at 0.033 Hz) with a suction electrode connected to a constant-current stimulator (NeuroLog system; Digitimer, Hertfordshire, UK).
Block of GABAARs increases responsiveness of dorsal horn cells in vivo at both P3 and P21
The functional state of GABAAR-mediated inhibition during the postnatal period was tested using extracellular recording of dorsal horn cells in vivo at P3–P4 (n = 20 for mechanical sensitivity; plus n = 6 for noxious heat) and P20–P22 (n = 12). These were stable single-cell recordings over a period of ≤60 min to allow baseline and repeated post-drug testing. The proportions of superficial (presumed laminae I and II) and deep (presumed laminae III, IV, and V) cells sampled for mechanical sensitivity were similar for the two age groups: 8/20 cells at P3 and 4/12 cells at P21 were superficial. Single cells were isolated and baseline activity characterized before spinal application of the GABAAR antagonist gabazine (5 μl of 20 μM).
Figure 1A shows the typical effect of gabazine on dorsal horn cell receptive fields (RFs) and mechanical responsiveness. Baseline RF areas at P3 tended to be relatively larger than those at P21 (Fitzgerald 1985; Torsney and Fitzgerald 2002). Gabazine caused an increase in both RF area and number of spikes evoked by a mechanical stimulus [three von Frey hairs (vFh) above threshold] in the majority of cells at both P3 and P21 (Fig. 1B). The mean vFh response increased from 9.1 ± 1.5 to 15.8 ± 3.0 spikes at P3 and from 21.7 ± 3.8 to 37.4 ± 8.0 spikes at P21. Within each age group, there was a significant difference between baseline and peak gabazine response for each measure (Wilcoxon matched-pairs tests: P < 0.01 for all tests). No cells at any age showed a decreased response after gabazine to either parameter. The proportions of responding cells were similar across age groups: 14/19 (74%) P3 cells and 8/12 (67%) P21 cells showed an increase in RF area; 13/20 (65%) P3 cells and 8/11 (73%) P21 cells showed an increased mechanically evoked response (χ2 test, P > 0.2). The mean percentage changes from baseline in RF area or mechanical response after gabazine application were not different for P3 and P21 (RF area: 191 ± 23% at P3 and 178 ± 39% at P21; vFh spikes 183 ± 18% at P3 and 183 ± 26% at P21; t-test, P > 0.75). The excitatory effect of gabazine was apparent by 10 min post-application for both age groups, and a recovery back toward baseline for at least one of the parameters was seen in 9/14 P3 cells and 6/8 P21 cells. In one cell, the drug was applied a second time, 35 min after the first application, illustrating the repeatability of the excitatory effect (Fig. 1C). Background firing developed or increased after gabazine in 6/20 P3 cells but only 1/12 P21 cell, and was unaffected in the remainder. At both ages, there were no significant differences in the effect of gabazine between superficial and deep dorsal horn neurons in any of the parameters examined (data not shown).
In six separate cells at P3, the effect of gabazine was tested on the responses to prolonged noxious heating (48°C for 4 s) of the receptive field (see Fig. 2). The total number of spikes fired in response to heat increased after gabazine from 47.3 ± 17.9 to 81.2 ± 26.5 spikes. In 4/6 cells at P3, the maximum heat-evoked firing rate was increased by gabazine, from 12.5 ± 3.8 to 20.0 ± 5.6 Hz. Afterdischarge duration more than doubled after gabazine in each of these four cells. The remaining two cells were unaffected. In all cases, firing to a 30°C water jet (to control for mechanical effects of the stimulation) was minimal (1.2 ± 1.2 spikes before and 1.8 ± 1.4 spikes after gabazine).
Block of GABAARs facilitates primary afferent-evoked action potential discharge in neonatal deep dorsal horn neurons in vitro
To exclude the possibility that the functional GABAAR-mediated inhibition seen at early postnatal ages is dependent on an intact nervous system, perforated patch-clamp recordings were obtained from deep dorsal horn neurons (located 450–700 μm from the edge of the dorsal white matter) in spinal cord slices at P2–P3. In the presence of aCSF, sciatic nerve stimulation evoked short-latency action potential (AP) discharge followed by a barrage of excitatory postsynaptic potentials (EPSPs). In 10 of 11 cells examined, bath application of gabazine (10 μM) increased the number of APs evoked during a 5-s period after stimulation, as a result of increased summation of EPSPs (Fig. 3A). An abolition of AP discharge occurred in the presence of the drug in one neuron (data not shown). Overall, gabazine caused a 271 ± 48% increase in the number of spikes per stimulus compared with control (n = 11; P < 0.05, Kruskal–Wallis test; Fig. 3B); averages of 1.28 ± 0.22 APs were discharged in aCSF and 3.08 ± 0.49 APs in the presence of gabazine. In the majority of dorsal horn neurons, the evoked firing returned to baseline levels on washout of the drug (to 121 ± 21% of control; see Fig. 3B).
The present results demonstrate, for the first time, that block of spinal GABAAR activation in vivo elicits a profound disinhibition of dorsal horn neuronal firing throughout the early postnatal period. In fact, gabazine increased firing to a similar extent at P3 and P21, suggesting that GABAergic inhibition may be relatively mature by P3. A more stringent comparison of the strength of GABAergic inhibition at P3 and P21 would require a comprehensive analysis of the potency of intrathecal gabazine at the two ages. However, because previous work has demonstrated that the potency of gabazine is independent of possible developmental changes in receptor subunit composition (Ebert et al. 1997), it seems unlikely that the potency is significantly different between P3 and P21. Thus it appears that functional GABAergic inhibition matures in the spinal cord earlier than in the rest of the CNS, which is consistent with earlier reports demonstrating reduced excitability of embryonic (E18.5) motoneurons in the presence of GABA (Hubner et al. 2001).
A strong GABAergic inhibition of dorsal horn neurons at early postnatal ages may seem surprising because the presence of GABAergic depolarizations in newborn dorsal horn neurons in vitro (Baccei and Fitzgerald 2004) makes it tempting to predict that functional GABAAR-mediated inhibition would be significantly weaker in these cells compared with older neurons in which GABA exclusively evokes membrane hyperpolarization. However, it should be noted that in addition to evoking hyperpolarization, the activation of GABAARs reduces neuronal excitability by increasing membrane conductance and shunting subsequent excitatory inputs (Bormann et al. 1987). This shunting component would be expected to be present in newborn dorsal horn cells regardless of the [Cl−]i and resultant ECl. It has been well documented that the efficacy of excitatory synaptic transmission in the newborn dorsal horn is low and increases significantly during the early postnatal period. For example, electrical activation of C-fibers fails to induce APs in dorsal horn neurons in vivo until about P10 and spontaneous EPSCs are appreciably upregulated during the first ten postnatal days (Baccei et al. 2003). It is possible that the GABAAR-mediated increase in conductance is sufficient to effectively shunt the existing excitatory inputs onto neonatal dorsal horn neurons during the first few postnatal days despite the absence of strong hyperpolarization.
Recent work has demonstrated that, although the developmental shift in chloride homeostasis is completed by P7, lamina I dorsal horn neurons did not reach their full Cl− extrusion capacity until the third postnatal week (Cordero-Erausquin et al. 2005). Although we cannot exclude the possibility that lamina I cells may have a distinct Cl− homeostasis compared with the deeper cells we studied, the present work suggests that the reduced Cl− extrusion ability may still be sufficient to maintain a strong inhibitory tone onto neonatal dorsal horn neurons because a similar degree of disinhibition of dorsal horn cell firing by gabazine was observed at P3 and P21 in vivo. In addition, in vitro perforated patch-clamp recordings of P3 dorsal horn neurons also revealed a significant increase in primary afferent evoked firing when spinal GABAARs were blocked. It should be noted that, although GABAAR-mediated miniature IPSCs decay at a slower rate in the immature dorsal horn (Keller et al. 2004), recent studies have also demonstrated a lower frequency of spontaneous IPSCs and weaker primary afferent-evoked IPSCs during the first postnatal days in spinal cord slices (Baccei and Fitzgerald 2004). As a result, during the spinal processing of sensory stimuli under normal conditions, the intracellular Cl− load imposed on neonatal dorsal horn cells may be substantially lower than that seen in adult neurons, thus requiring a lower efficiency of Cl− extrusion to allow for adequate neuronal inhibition.
However, we cannot rule out the possibility that during intense or prolonged stimulation under pathological conditions, chloride influx could be too great for the Cl− extrusion mechanisms present at P3 to maintain the correct [Cl−]i, leading to a reduction or loss of inhibition. Indeed, the polarity of GABAergic currents recorded from dorsal horn neurons can reverse during high-frequency stimulation (20–100 Hz) of local synaptic inputs (Cordero-Erausquin et al. 2005). Sustained firing of dorsal horn neurons in vivo could also potentially result in an acute downregulation of KCC2 mRNA or protein levels, as has been demonstrated to occur in the hippocampus by a BDNF-dependent mechanism (Rivera et al. 2004).
These results show that the increased excitability of tactile and nociceptive circuits in the newborn is unlikely to arise from insufficient GABAergic inhibitory activity. The large dorsal horn cell and reflex receptive fields, low thresholds, and long duration responses in the newborn rat may instead result from subtle changes in synaptic organization or indeed reflect differences in other types of local inhibitory synapses (e.g., glycinergic, opioidergic), or immature descending inputs to the developing spinal cord.
This work was supported by the Wellcome Trust and the United Kingdom Medical Research Council. L. Bremner is a participant in the Wellcome Trust four-year Neuroscience PhD program.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2006 by the American Physiological Society