SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

Because nerve growth factor (NGF) is elevated during inflammation and is known to activate the sphingomyelin signalling pathway, we examined whether NGF and its putative second messenger, ceramide, could modulate the excitability of capsaicin-sensitive adult and embryonic sensory neurons. Using the whole-cell patch-clamp recording technique, exposure of isolated sensory neurons to either 100 ng ml−1 NGF or 1 μM N-acetyl sphingosine (C2-ceramide) produced a 3- to 4-fold increase in the number of action potentials (APs) evoked by a ramp of depolarizing current in a time-dependent manner. Intracellular perfusion with bacterial sphingomyelinase (SMase) also increased the number of APs suggesting that the release of native ceramide enhanced neuronal excitability. Glutathione, an inhibitor of neutral SMase, completely blocked the NGF-induced augmentation of AP firing, whereas dithiothreitol, an inhibitor of acidic SMase, was without effect. In the presence of glutathione and NGF, exogenous ceramide still enhanced the number of evoked APs, indicating that the sensitizing action of ceramide was downstream of NGF. To investigate the mechanisms of action for NGF and ceramide, isolated membrane currents were examined. Both NGF and ceramide facilitated the peak amplitude of the TTX-resistant sodium current (TTX-R INa) by approximately 1.5-fold and shifted the activation to more hyperpolarized voltages. In addition, NGF and ceramide suppressed an outward potassium current (IK) by ≈35 %. Ceramide reduced IK in a concentration-dependent manner. Isolation of the NGF- and ceramide-sensitive currents indicates that they were delayed rectifier types of IK. The inflammatory prostaglandin, PGE2, produced an additional suppression of IK after exposure to ceramide (≈35 %), suggesting that these agents might act on different targets. Thus, our findings indicate that the pro-inflammatory agent, NGF, can rapidly enhance the excitability of sensory neurons. This NGF-induced sensitization is probably mediated by activation of the sphingomyelin signalling pathway to liberate ceramide(s), wherein ceramide appears to be the second messenger involved in modulating neuronal excitability.

Ceramides are novel second messengers that may mediate the inflammatory response, apoptosis and altered gene expression in different cell types. One component of the inflammatory response involves activation of small diameter sensory neurons which, in turn, contributes to heightened sensitivity, vasodilatation and plasma extravasation. Indeed, a number of inflammatory mediators including prostaglandins (Kress & Reeh, 1996), NGF (Lewin et al. 1993; Lewin & Mendell, 1993; Shu & Mendell, 1999a), and cytokines (Ferreira et al. 1988; Schweizer et al. 1988; Cunha et al. 1992) enhance the sensitivity of sensory neurons to noxious stimulation. Although the prostaglandin-induced sensitization results from activation of the cyclic AMP- protein kinase A (PKA) transduction cascade (Cui & Nicol, 1995; Hingtgen et al. 1995; Evans et al. 1999) the second messenger system(s) that mediate the effects of these neurotrophins and cytokines remains unknown. This raises the question of whether NGF and/or ceramide plays any role in this enhanced sensitivity of sensory neurons.

In a number of cell systems, activation of cytokine receptors results in the liberation of ceramides as second messenger signalling molecules (Schütze et al. 1994; Ballou et al. 1996). In addition, recent studies have shown that NGF binding to the low-affinity neurotrophin receptor (p75NTR) activates the sphingomyelin signalling pathway causing the liberation of ceramide (Dobrowsky et al. 1994). Ceramide is thought to act as a second messenger molecule capable of mediating multiple physiological effects and this depends on the cell type in question (Schütze et al. 1994; Ballou et al. 1996; Mathias et al. 1998). Such effects range from the regulation of cell growth and apoptosis to the immune response. Based on the notion that certain pro-inflammatory agents, such as NGF, might activate the sphingomyelin signalling pathway, we explored the idea that ceramide, acting as an intracellular messenger, alters the sensitivity of sensory neurons to excitatory stimulation. In this report we demonstrate that NGF, through a pathway that depends on activation of a neutral sphingomyelinase (SMase), and ceramide enhance the excitability of sensory neurons through the modulation of both a TTX-R INa and a voltage-dependent IK. Portions of this work have been published previously in abstract form (Zhang et al. 2000; Nicol et al. 2001).

Methods

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

Isolation and culture of embryonic and adult rat sensory neurons

The procedures for isolation and culture of embryonic rat sensory neurons have been described previously (Vasko et al. 1994). Briefly, timed-pregnant rats were rendered unconscious with CO2, and killed by cervical dislocation. Embryos (E15-E17) were removed from the uterus and placed in a dish containing calcium-free and magnesium-free Hanks’ balanced salt solution. The dorsal root ganglia were dissected from each embryo and sensory neurons were dissociated from the ganglia with 0.025 % trypsin (37 ° C, 25 min) and mechanical agitation. The cells were grown in Dulbecco's modified Eagle's medium (Life Technologies, Grand Island, NY, USA) supplemented with 2 mm glutamine, 50 μg ml−1 penicillin and streptomycin, 10 % (v/v) heat-inactivated fetal bovine serum, 50 μM 5-fluoro-2′-deoxyuridine, 150 μM uridine, and 250 ng ml−1 7S-nerve growth factor (Harlan Bioproducts for Science, Indianapolis, IN, USA). For the patch-clamp recordings, ≈150 000 cells ml−1 were plated in a poly-d-lysine-coated culture dish containing small plastic coverslips. Cultures were maintained at 37 ° C in a 5 % CO2 atmosphere and the medium was changed every second day.

Isolation of sensory neurons from adult rats used the procedures developed by Lindsay (1988) with slight modification. Briefly, male Sprague-Dawley rats (100-150 g) were rendered unconscious by placing them in a chamber filled with dry ice and then killed by cervical dislocation. The isolated spinal column was hemisected, the spinal cord was removed and the dorsal root ganglia were collected in a culture dish filled with sterilized Puck's solution. The ganglia were transferred to a conical tube with F-12 medium containing 1 mg ml−1 collagenase 1A and 2.5 mg ml−1 dispase. After a 1 h incubation at 37 °C, the tube was centrifuged for 30 s before the enzyme-containing supernatant was removed. The pellet was resuspended in F-12 medium supplemented with 250 ng ml−1 7S-nerve growth factor (Harlan Bioproducts for Science) and mechanically dissociated with fire-polished pipettes until all obvious chunks of tissues were gone. Isolated cells were plated onto plastic coverslips that previously were coated with poly-d-lysine and laminin. The cells were maintained in F-12 medium containing nerve growth factor at 37 °C and 3 % CO2 and used within 12-24 h for electrophysiological recordings. All procedures have been approved by the Animal Use and Care Committee of the Indiana University School of Medicine.

Electrophysiology

Recordings were made using the whole-cell patch-clamp technique as previously described (Hamill et al. 1981; Evans et al. 1999). Briefly, a coverslip with the sensory neurons (embryonic neurons were used after 4-6 days in culture, adults were used within 12-24 h after isolation) was placed in a recording chamber where the neurons were bathed in normal Ringer solution of the following composition (mm): 140 NaCl, 5 KCl, 2 CaCl2, 1 MgCl2, 10 Hepes and 10 glucose, pH adjusted to 7.4 with NaOH. Recording pipettes were pulled from borosilicate glass tubing and typically had resistances of 2-5 MΩ when filled with the following solution (mm): 140 KCl, 5 MgCl2, 4 ATP, 0.3 GTP, 2.5 CaCl2, 5 EGTA (calculated free Ca2+ concentration of ≈100 nm) and 10 Hepes, at pH 7.3 adjusted with KOH. Whole-cell voltages or currents were recorded with an Axopatch 200 (Axon Instruments, Union City, CA, USA) patch-clamp amplifier; the data were acquired and analysed using pCLAMP 6.04 (Axon Instruments). The whole-cell recording configuration was established in normal Ringer solution. In the current clamp experiments, the neurons were held at their resting potentials; the depolarizing ramp was 1 s in duration. The amplitude of the ramp was adjusted to produce two to three APs under control conditions; the amplitude then remained constant throughout the recording period for each individual neuron. In voltage clamp experiments both capacitance and series resistance compensation (typically 70 %) were used; compensation was used for leak currents in TTX-R INa recordings but not in the IK recordings. For the measurements of TTX-R INa the average uncompensated series resistance was 2.6 ± 0.2 MΩ (n= 17), which yielded a maximum voltage error of −5.0 ± 0.7 mV for the peak currents obtained at −10 mV. Similarly for IK, the average uncompensated series resistance was 2.6 ± 0.2 MΩ (n= 26), which yielded a voltage error of 7.9 ± 1.3 mV for the peak currents obtained at +60 mV.

IK was isolated by superfusing the cells with 140 mmN-methyl-glucamine chloride Ringer solution (NMG, an equimolar substitution for NaCl); pH adjusted to 7.4 with KOH. The TTX-R INa was isolated by superfusing the cells with (mm): 30 NaCl, 65 NMG-Cl, 30 TEA, 0.1 CaCl2, 5 MgCl2, 10 Hepes, 10 glucose, 10 sucrose, and 500 nm TTX; pH adjusted to 7.4 with TEAOH; the recording pipette contained (mm): 110 CsFl, 25 CsCl, 10 NaCl, 5 MgCl2, 4 ATP, 0.3 GTP, 1 CaCl2, 10 EGTA, 10 glucose, and 10 Hepes, at pH 7.3 (maintained with CsOH). The membrane voltage was held at −60 mV; activation of the currents was determined by voltage steps of 300 ms or 30 ms for IK or TTX-R INa, respectively, which were applied at 5 s intervals in +5 or +10 mV increments to +60 mV. Steady-state inactivation of IK was measured by applying a 15.5 s conditioning prepulse (−100 to +40 mV in 20 mV increments) after which the voltage was stepped to +60 mV for 170 ms; a 20 s interval separated each prepulse sweep. After obtaining the control response, the superfusate was changed to the appropriate Ringer solution and cells were superfused continuously for the appropriate times. At the end of each recording, the neuron was exposed to 100 nm capsaicin. This neurotoxin was used to distinguish capsaicin-sensitive sensory neurons as these neurons are believed to transmit nociceptive information (Holzer, 1991). The results reported below were obtained from capsaicin-sensitive neurons only. All experiments were performed at room temperature (≈23 °C).

Data analysis

All values represent the mean ± standard error of the mean (s.e.m.). The excitability parameters described in Table 1 were determined by differentiating (dV/dt) the current clamp trace (sampling frequency of 250 Hz). The voltage and time at which the first AP was fired were taken as the point that exceeded the baseline value of mV ms−1 by > 20-fold. The baseline mV ms−1 was determined by averaging the points between the onset of the ramp and two data points before the take-off point of the AP. The voltage dependence for activation of TTX-R INa and IK was fitted with the Boltzmann relation G/Gmax= 1/[1 + exp(V0.5Vm)/k], where G is the conductance, Gmax is the conductance calculated for the step to +10 mV for TTX-R INa or +60 mV for IK measurements, V0.5 is the voltage for half-maximal activation, Vm is the membrane voltage, and k is a factor describing the steepness of the relation. The Boltzmann parameters were determined for each individual neuron from which the mean ±s.e.m. was calculated. Statistical differences between the control recordings and those obtained under various treatment conditions were determined by using either a paired t test, analysis of variance (ANOVA), or a repeated measures ANOVA (RM ANOVA). When a significant difference was obtained with an ANOVA, post hoc analyses were performed using a Tukey test. A one-way repeated measures ANOVA was used to determine statistical differences between the currents before and after drug treatments; in separate time control experiments IK and TTX-R INa did not vary significantly over a 20 min time period. Values of P < 0.05 were judged to be statistically significant.

Table 1.  Effects of NGF and ceramide on current clamp parameters
 Resting Vm (mV)First AP (mV)First AP (ms)Ramp depolarization rate (mV ms−1)Fold increase
  1. G, glutathione; Cer, ceramide. *P < 0.05 (paired t test).

NGF (n=7)
 Control−55.2 ± 1.7−21.0 ± 4.2657 ± 1100.075 ± 0.021
 2 min−53.6 ± 2.6−21.4 ± 3.5481 ± 680.085 ± 0.0201.23 ± 0.08
 6 min−50.2 ± 4.2−22.9 ± 3.4332 ± 55*0.116 ± 0.025*1.71 ± 0.19*
 10 min−49.7 ± 3.8*−23.3 ± 2.8342 ± 62*0.123 ± 0.027*1.83 ± 0.27*
Ceramide (n=7)
 Control−56.4 ± 1.9−26.3 ± 3.0769 ± 710.050 ± 0.011
 6 min−52.1 ± 1.6*−23.9 ± 2.8633 ± 490.053 ± 0.0071.16 ± 0.11
 10 min−49.6 ± 1.8*−24.3 ± 2.3552 ± 50*0.057 ± 0.0081.25 ± 0.12
 20 min−44.4 ± 2.8*−23.5 ± 1.3376 ± 44*0.082 ± 0.015*1.76 ± 0.22*
GSH–NGF–ceramide (n=3)
 G-NGF 6 min−56.0 ± 3.9−20.9 ± 6.9943 ± 230.045 ± 0.004
 Cer 2 min−53.4 ± 3.8−23.2 ± 5.0715 ± 510.047 ± 0.0051.05 ± 0.03
 Cer 6 min−49.6 ± 4.2*−18.0 ± 4.8472 ± 79*0.074 ± 0.0071.70 ± 0.31
 Cer 10 min−47.2 ± 3.5*−18.9 ± 4.8388 ± 88*0.084 ± 0.009*1.93 ± 0.31*

Chemicals

C2-ceramide (N-acetyl sphingosine), dihydroceramide, and prostaglandin E2 were obtained from Cayman Chemical Co. (Ann Arbor, MI, USA). All other chemicals were obtained from Sigma Chemical Corp. (St Louis, MO, USA). C2-ceramide, prostaglandin E2, and capsaicin were dissolved in 1-methyl-2-pyrrolidinone (HPLC grade, Aldrich Chemical Co., Milwaukee, WI, USA) to obtain concentrated stock solutions. These stock solutions were then diluted with Ringer solution to yield the appropriate concentration. We have demonstrated previously that the vehicle, 1-methyl-2-pyrrolidinone, had no effect on the activation of IK (Nicol et al. 1997).

Results

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

NGF and ceramide enhance the excitability of sensory neurons

Dobrowsky et al. (1994) demonstrated that NGF binding to the low-affinity neurotrophin receptor (p75NTR) leads to the activation of SMase and increases in intracellular ceramide. Based on this observation, we explored the idea that NGF could sensitize embryonic capsaicin-sensitive sensory neurons and that this enhancement was mediated by ceramide. Treatment with 100 ng ml−1 NGF increased the number of action potentials (APs) elicited by the ramp of depolarizing current in a time-dependent manner (see Fig. 1). Under control conditions, a representative neuron fired four APs in response to the ramp (left panel A, resting Vm−51 mV), but after a 6 min exposure to NGF this neuron now fired 13 APs (right panel, resting Vm 50 mV). The results are summarized in panel B wherein treatment with NGF significantly increased the number of evoked APs from an average control value of 3.4 ± 0.2 to 13.4 ± 2.1 (n= 7, RM ANOVA) after a 6 min exposure to NGF. The increased excitability at 6 min was not accompanied by a corresponding change in the resting Vm, but by 10 min Vm had depolarized significantly. The traces below the APs illustrate the dV/dt transformation; these results were used to quantify the effects of NGF on the parameters of excitability and are summarized in Table 1. The NGF-enhanced excitability was exhibited by both a decreased latency for the firing of the first AP and an increased rate of membrane depolarization for the applied ramp. Although the number of APs and the rate of depolarization were increased with NGF, there was no change in the voltage at which the first AP was fired. Thus, an acute exposure to NGF can enhance the excitability of sensory neurons without greatly altering the resting Vm or the apparent firing threshold.

image

Figure 1. NGF and ceramide augment the excitability of embryonic sensory neurons

A, a depolarizing current ramp (600 pA) elicited APs in a representative neuron under normal control conditions. In this same neuron, a 6 min exposure to 100 ng ml−1 NGF in normal Ringer solution greatly increased the number of evoked APs (right panel). The dotted lines indicate 0 mV. The panels below show the dV/dt for each trace. B, the effect of NGF on the number of evoked APs is summarized. Ramp amplitudes ranged from 180 to 600 pA. C, ceramide increased the number of APs evoked by the ramp. The left panel summarizes the effects of 1 μM ceramide on the number of APs evoked over a time period of 20 min. The right panel demonstrates the effects of internal perfusion with bacterial SMase on the number of APs over the same time course in a different series of experiments. Ramp amplitudes ranged from 200 to 700 pA. The asterisks represent a significant (P < 0.05) increase compared to the control values.

Download figure to PowerPoint

We next examined the effects of N-acetyl sphingosine (C2-ceramide, hereafter referred to as ceramide) on the excitability of embryonic capsaicin-sensitive sensory neurons in response to the current ramp. The application of ceramide increased the number of APs evoked by the ramp in a time-dependent manner (see Fig. 1C, left panel). Of the 16 total sensory neurons, 11 exhibited an increased number of APs after exposure to ceramide; those neurons not responsive to ceramide have been excluded from the analysis. In the ceramide-sensitive neurons, the number of APs increased significantly from a control value of 2.7 ± 0.2 to 7.0 ± 1.0 and 11.4 ± 0.9 (n= 11, RM ANOVA) after 6 and 20 min treatments, respectively, with 1 μM ceramide. In a manner analogous to NGF, ceramide significantly altered the resting Vm, the firing latency, and the rate of depolarization without changing the apparent threshold (see Table 1). To examine whether the increase in APs was reversible, neurons were bathed in normal Ringer solution for 10 min after the 20 min exposure to ceramide; however, the number of APs elicited by the ramp remained elevated at 14.5 ± 1.8 (RM ANOVA, n= 4).

Because sphingomyelinase (SMase) is the enzyme that liberates ceramide from membrane phospholipids (Levade & Jaffrezou, 1999), we internally perfused neurons with bacterial SMase via the recording pipette to increase the intracellular levels of ceramide. As shown in the right panel of Fig. 1C, treatment with SMase (0.2 U ml−1) for approximately 6 min produced a significant increase in the number of APs (7.8 ± 1.1, n= 6, RM ANOVA) that was similar to that caused by ceramide. Continued internal perfusion with SMase led to a further increase in the number of APs. Although not significant, the resting Vm under control conditions was −60 ± 1 compared to −50 ± 4 mV after a 20 min exposure to SMase. In another series of experiments, exposure to the vehicle, methyl-pyrrolidinone, had no effect on the number of APs (three for both control and after 20 min, n= 3) or the resting potential (−61 ± 2 in control vs.−60 ± 4 mV 20 min after vehicle). These results suggest that both NGF and an increase in the intracellular concentration of ceramide can sensitize sensory neurons to excitatory stimulation.

Inhibition of SMase blocks the sensitization by NGF

If the sensitizing actions of NGF result from the liberation of ceramide, then inhibition of SMase should attenuate the enhanced excitability produced by NGF. To examine this, sensory neurons were pretreated with either dithiothreitol, an inhibitor of acidic SMase, or glutathione, an inhibitor of neutral SMase (Liu & Hannun, 1997; Liu et al. 1998), and then exposed to NGF. As illustrated in Fig. 2A, a 10 min pretreatment with 1 mm dithiothreitol had no effect on the number of APs evoked by the ramp under control conditions. Treatment with 100 ng ml−1 NGF in the presence of dithiothreitol produced a significant increase (RM ANOVA) in the number of APs after 6, 10, and 20 min exposures to NGF in a manner analogous to that in the absence of dithiothreitol (see Fig. 1). NGF (10 min treatment) depolarized the resting Vm to −47 ± 3 mV from the control values (−57 ± 3 mV, n= 3). In contrast to dithiothreitol, a 10 min pretreatment with 3 mm glutathione completely blocked the capacity of NGF to augment the number of APs produced by the depolarizing ramp as well as the depolarization (Fig. 2B). Neither dithiothreitol nor glutathione alone had any effect on Vm in these neurons (data not shown). It is possible that glutathione might, in some non-specific manner, prevent the neuron's capacity to be sensitized. To examine this notion, sensory neurons were pretreated sequentially with glutathione, glutathione plus NGF, and then finally with a combination of glutathione-NGF-ceramide (see Fig. 2C). Glutathione again prevented the sensitization produced by NGF (see Table 1), but in the presence of glutathione and NGF, ceramide significantly increased the number of evoked APs (RM ANOVA) and the changes in excitability parameters were similar to those observed with ceramide alone (see Table 1). Taken together, these results suggest that NGF leads to the activation of a neutral SMase which causes the release of intracellular ceramide. This second messenger molecule is then capable of enhancing the excitability of capsaicin-sensitive sensory neurons.

image

Figure 2. Inhibition of neutral, but not acidic, SMase blocks the sensitization produced by NGF in embryonic sensory neurons

A, the effects of a 10 min pretreatment with 1 mm dithiothreitol (DTT) on the number of evoked APs produced by 100 ng ml−1 NGF in normal Ringer solution. B, a 10 min pretreatment with 3 mm glutathione (GSH) completely blocked the effects of NGF. C, 1 μM ceramide (Cer) augmented the number of APs in the presence of GSH and NGF. Neurons were treated sequentially (as indicated by the horizontal bars) with 3 mm GSH for 10 min, then 100 ng ml−1 NGF for 6 min in the presence of GSH, and finally with 1 μM ceramide for 20 min in the presence of GSH and NGF. Ramp amplitudes ranged from 200 to 620 pA. The asterisks represent significant differences (P < 0.05) from the control values.

Download figure to PowerPoint

NGF and ceramide augment the TTX-R INa

The increased AP firing, depolarization and rate of depolarization after NGF or ceramide treatments raise an interesting question as to which ion currents are modulated by NGF or ceramide to enhance the excitability of capsaicin-sensitive sensory neurons. Previous work demonstrated that the pro-inflammatory prostaglandin, PGE2, augmented the TTX-resistant sodium current (TTX-R INa) in rat sensory neurons (England et al. 1996; Gold et al. 1996). To explore this notion, we examined the actions of NGF or ceramide on the TTX-R INa in adult rat sensory neurons. In these particular experiments external [Na+] was reduced to 30 mm and only neurons from adult animals were used because of space-clamp considerations. A prepulse to remove inactivation was not used in these studies so that modulation of the current would be examined at voltages that were similar to the normal resting potential recorded for these neurons (approximately −60 mV). Exposure to 100 ng ml−1 of NGF produced a time-dependent enhancement of the TTX-R INa and appeared to shift the activation to more hyperpolarized potentials (see Fig. 3). Representative control recordings are shown in the left panel of Fig. 3A whereas the right panel illustrates the effects of a 6 min treatment with NGF. NGF did not appear to alter the inactivation rates of TTX-R INa as the time constants for current recovery were not different before and after treatment (data not shown). Enhancement of the peak TTX-R INa occurred approximately 1 min after exposure to NGF (Fig. 3B). The current-voltage relation for TTX-R INa is illustrated in Fig. 3C (left panel). Under control conditions the maximum peak current occurred for the step to −10 mV and had an average value of −1.86 ± 0.28 nA (n= 5). After a 10 min exposure to NGF, the peak current significantly increased to −2.71 ± 0.20 nA and had shifted to −20 mV. To reduce the variability, the peak currents at each voltage were normalized to their respective maximum control values (middle panel). The largest control currents were obtained at −10 mV and represented 0.98 ± 0.02 of the maximum total current. A 10 min exposure to NGF caused a significant increase in the current wherein the fraction of the current was now 1.54 ± 0.25 and peaked at −20 mV. The conductance-voltage relations and their corresponding Boltzmann fits are shown in the right panel. V0.5 was shifted to more hyperpolarized values after NGF and G/Gmax was increased significantly. The Boltzmann parameters are summarized in Table 2. Thus, these results demonstrate that NGF produces a large increase in the total TTX-R INa with a shift in the activation curve to more hyperpolarized potentials.

image

Figure 3. NGF and ceramide enhance the TTX-R INa in adult sensory neurons

A, the effects of 100 ng ml−1 NGF on representative current traces under control conditions (left) compared to those after a 6 min exposure to NGF (right). The line labelled zero represents the zero current value. B, the time course of NGF's action. The peak TTX-R INa was obtained for a voltage step from −60 to −20 mV, and this step was repeated every 20 s. NGF was added at the indicated time. The asterisk represents the first time point that was significantly different from the control values. The data points represent the average obtained from three neurons. Left panel in C, the current-voltage relations obtained before and after treatment with NGF. Treatment times of 6 and 10 min produced a significant increase in the peak TTX-R INa for voltage steps between −40 and +20 mV (RM ANOVA). The membrane voltage was held at −60 mV; activation of the currents was determined by voltage steps of 30 ms that were applied at 5 s intervals in +5 or +10 mV increments to +60 mV. Middle panel in C, the normalized current-voltage relation and the effects of NGF. Peak currents were normalized to their respective control values obtained for the step to −10 mV. Significant increases were obtained for voltages between −40 and +10 mV. Right panel in C demonstrates the conductance-voltage relation; data points have been normalized to the conductance obtained at +10 mV. Left panel in D, time-dependent effects of 1 μM ceramide (Cer); the current was increased significantly (RM ANOVA) for voltages between −20 and +15 mV and −25 and +20 mV for 6 min and for both 10 and 20 min exposures, respectively. Right panel in D, the effects of ceramide on the normalized current-voltage relation. Peak currents were normalized to their respective control values obtained for the step to −10 mV. Significant increases were obtained for voltages between −15 and +5 mV and −25 and +20 mV for 10 and 20 min exposures, respectively. Asterisks represent a significant difference (P < 0.05) compared to control.

Download figure to PowerPoint

Table 2.  Effects of NGF and ceramide on the Boltzmann parameters for TTX-R INa
 G/GmaxV0.5 (mV)k (mV)
  1. *P < 0.05 (paired t test).

NGF (n=5)
 Control1.00 ± 0−20.1 ± 3.25.6 ± 0.9
 10 min1.39 ± 0.16−26.3 ± 4.1*4.6 ± 1.1*
Ceramide (n=4)
 Control1.00 ± 0−17.0 ± 3.57.4 ± 0.3
 20 min1.25 ± 0.09*−20.1 ± 3.9*8.0 ± 0.6

Similar to NGF, ceramide enhanced the amplitude of the TTX-R INa and appeared to shift the activation to more hyperpolarized potentials. The effects of ceramide on the voltage-dependent activation of TTX-R INa are summarized in the left panel of Fig. 3D. Under control conditions, the peak TTX-R INa obtained at −10 mV had an average value of −3.13 ± 1.07 nA (n= 4). After 10 and 20 min exposures to 1 μM ceramide, the peak TTX-R INa shifted to −15 mV and increased significantly to −3.76 ± 1.32 and −3.95 ± 1.23 nA, respectively (RM ANOVA). Because of the large variability in the peak TTX-R INa, the currents were normalized to their respective maximum control values (right panel of Fig. 3D); ceramide (after 20 min) significantly increased the average peak TTX-R INa by 1.25 ± 0.13-fold compared to the controls with a 5 mV leftward shift in the peak voltage. When the increase in TTX-R INa was assessed irrespective of voltage (to eliminate the voltage variabilities at which Imax occurs, two cells at −15 mV, one at −10 mV, and one at 0 mV), ceramide significantly augmented the peak current by 1.21 ± 0.08, 1.28 ± 0.09 and 1.34 ± 0.07-fold after 6, 10 and 20 min exposures (n= 4, RM ANOVA). These increases were not different from one another indicating that the maximal effects of ceramide were obtained after approximately 6 min. The effects of ceramide on the conductance-voltage relation were similar to those observed for NGF (data not shown); the results were fitted by the Boltzmann relation for each neuron wherein ceramide significantly increased the value of G/Gmax and V0.5 was shifted leftward (see Table 2). Thus, these results indicate that ceramide, like NGF, augmented both the amplitude and conductance of TTX-R INa and also shifted the activation voltage to more hyperpolarized values.

To ascertain whether this enhancement was specific to ceramide, we examined the actions of an inactive analogue of ceramide, dihydroceramide, on TTX-R INa. The peak TTX-R INa was not affected by 1 μM dihydroceramide (20 min exposure, data not shown) at −10 mV (control, −2.01 ± 0.61 vs. dihydroceramide, −2.14 ± 0.57 nA, n= 3). In a separate series of normal control recordings, the peak TTX-R INa obtained at −10 mV remained unchanged over a recording period of 20 min (control −2.33 ± 0.86 vs.−2.24 ± 0.72 nA after 20 min, n= 4). Also, time or dihydroceramide had no effect on the value of G/Gmax or the slope factor, k (data not shown). These experiments indicate that neither time nor treatment with dihydroceramide had any effect on the peak TTX-R INa at −10 mV. Thus, treatment with the inactive analogue of ceramide did not affect this current in the same manner as observed with ceramide.

NGF and ceramide suppress a delayed rectifier-like IK

The excitability of sensory neurons is regulated by the modulation of both TTX-R INa (England et al. 1996; Gold et al. 1996) and an outward potassium current(s) (Nicol et al. 1997; Evans et al. 1999). In another series of experiments, we examined the capacity of NGF or ceramide to modify outward potassium current(s) (IK) in capsaicin-sensitive sensory neurons. As illustrated in Fig. 4, treatment with NGF (100 ng ml−1) produced a time-dependent suppression of IK in embryonic rat sensory neurons. Panel A shows representative current traces obtained under control conditions (left panel) and after a 20 min treatment with NGF wherein IK was reduced by 37 % (middle). The IK sensitive to NGF was obtained by subtraction of the currents remaining after NGF from their control traces (shown in the right panel). The NGF-sensitive IK exhibited rapid activation with a slow relaxation suggesting that NGF acts on a delayed rectifier-like IK. The time dependence for suppression of IK by NGF is shown in Fig. 4B wherein a significant reduction in IK was observed after 2 min. The current-voltage relation (Fig. 4C, left) demonstrates that NGF (20 min treatment) reduced the average peak value of IK from 3.89 ± 0.16 (n= 4) to 2.66 ± 0.24 nA for the step to +60 mV. The conductance-voltage relation is summarized in the right panel; the values were fitted by the Boltzmann relation. Treatment with NGF significantly reduced G/Gmax by approximately 30 % without altering V0.5 or k (see Table 3). Similar results for the suppression of IK by NGF were obtained from sensory neurons isolated from adult rats and are illustrated in Fig. 4D. Again, only G/Gmax was altered significantly by NGF (see Table 3). In a separate series of experiments, the recovery of IK was examined after a 6 min exposure to NGF, which produced a suppression of 24 ± 3 % (n= 5) of IK at +40 mV. After 10 and 15 min periods of washout with normal Ringer solution the extent of suppression had not lessened and 18 ± 4 % (n= 5) and 22 ± 4 % (n= 4) inhibition remained, respectively. In one cell held for a 30 min washout, IK recovered by 7 % (from 22 to 15 % blockage). These observations are similar to those obtained in current clamp recordings with ceramide. Such findings indicate that the NGF-mediated suppression of IK is not readily reversed. These results indicate that NGF suppressed IK in both embryonic and adult sensory neurons to similar extents without modification of the voltage dependence of activation and that the NGF-sensitive IK exhibited slow relaxation kinetics.

image

Figure 4. NGF suppresses a voltage-dependent IK in embryonic and adult sensory neurons

A, the effects of a 20 min treatment with 100 ng ml−1 NGF on IK obtained from a representative embryonic sensory neuron. Traces on the left represent control recordings, the middle traces were recorded after exposure to NGF, and the right illustrates the NGF-sensitive current. Neurons were held at −60 mV; the traces shown were obtained for voltage steps from −80 to +60 mV in 20 mV increments. The line labelled zero represents the zero current value. B, the time course for the suppression of IK by NGF. The peak IK was obtained for a voltage step from −60 to +40 mV; the step was repeated every 20 s. The asterisk represents the first point that is significantly different from the control values. Left panel in C, effects of NGF on the current-voltage relation for IK in embryonic neurons are summarized. The NGF-induced reduction of IK after 2 min was significant for voltage steps between 0 and +60 mV and after 10 and 20 min for steps between −30 and +60 mV (RM ANOVA). The conductance-voltage relation is illustrated in the right panel of C where the data points have been fitted by the Boltzmann relation. The suppression of G/Gmax after 2 min was significant for voltage steps between 0 and +60 mV and after 10 and 20 min for voltage steps between −30 and +60 mV (RM ANOVA). Left panel in D summarizes the effects of NGF on the current-voltage relation for IK in adult neurons. The conductance-voltage relation is illustrated in the right panel; the data points have been fitted by the Boltzmann relation and are shown as the continuous lines. NGF had no significant effect at 2 min, but did produce a significant reduction in both IK and G/Gmax for voltage steps positive to −30 mV for 10 and 20 min exposures.

Download figure to PowerPoint

Table 3.  Effects of NGF and ceramide on the Boltzmann parameters for IK
 G/GmaxV0.5 (mV)k (mV)
  1. *P < 0.05 (paired t test).

NGF (n=4)
 Control1.01 ± 0.01−7.8 ± 1.513.5 ± 0.1
 20 min0.69 ± 0.04*−4.9 ± 1.116.2 ± 0.7
NGF, adult (n=4)
 Control1.02 ± 0.025.6 ± 5.317.3 ± 0.3
 20 min0.69 ± 0.08*5.3 ± 4.518.6 ± 1.0
100 nM ceramide embryonic (n=5)
 Control0.96 ± 0.02−10.2 ± 1.411.4 ± 0.7
 20 min0.77 ± 0.03*−12.8 ± 2.011.1 ± 0.4
1 μM ceramide, embryonic (n=7)
 Control0.98 ± 0.013.3 ± 2.913.6 ± 0.7
 20 min0.61 ± 0.04*−4.2 ± 2.3*13.2 ± 0.9
10 μM ceramide, embryonic (n=5)
 Control0.98 ± 0.01−4.7 ± 2.111.1 ± 0.7
 20 min0.52 ± 0.04*−7.4 ± 1.711.5 ± 0.7
10 μM ceramide, adult (n=4)
 Control0.98 ± 0.02−0.3 ± 5.214.7 ± 1.1
 20 min0.30 ± 0.03*−11.1 ± 1.3*15.2 ± 0.9

Ceramide also produced a time-dependent suppression of outward potassium current(s) IK in capsaicin-sensitive sensory neurons. The effects of ceramide on the current- voltage relation obtained in embryonic capsaicin-sensitive sensory neurons are summarized in Fig. 5A (left panel). After a 20 min exposure to 1 μM ceramide, the fraction of the remaining IK obtained at +60 mV was 0.62 ± 0.04 (n= 7) compared to the respective control currents. The right panel represents the conductance-voltage relation wherein the conductance values were fitted by the Boltzmann relation. Ceramide (1 μM) caused a significant shift in V0.5 and a reduction in G/Gmax for the step to +60 mV whereas the value of k was unaffected (see Table 3). The extent of inhibition by ceramide was very similar whether the peak or the sustained IK at the end of the voltage step was assessed. For example, 1 μM ceramide produced a 38 ± 4 % inhibition of the peak IK; similarly, the decrease was 39 ± 4 % for the sustained IK. Similarly, in those experiments examining steady-state inactivation of IK (see Methods and below), the ceramide-induced suppression of IK measured at the peak (44 % reduction) was quite similar to the sustained IK (49 % reduction) measured at the end of the 15 s conditioning prepulse. Inhibition of IK by ceramide was concentration dependent (data not shown). A concentration of 10 nm ceramide (n= 3) had no effect on IK whereas 100 nm, 1, and 10 μM significantly decreased IK by 17.2 ± 2.5 (n= 5), 37.7 ± 3.8 (n= 7), and 45.0 ± 4.7 % (n= 5), respectively. The inhibition produced by 1 and 10 μM was significantly greater than that at 100 nm; there was no difference between 1 and 10 μM (ANOVA). In a separate series of experiments, 1 μM dihydroceramide had no effect on the peak IK at +60 mV (data not shown). Under control conditions, the average current was 2.25 ± 0.38 nA and after a 20 min exposure to dihydroceramide the average current remained unchanged (2.19 ± 0.35 nA, n= 3). These results indicate that ceramide inhibited the total IK by ≈35 % with a hyperpolarizing shift, although small, in the voltage dependence for activation.

image

Figure 5. Ceramide suppresses a voltage-dependent IK in both embryonic and adult sensory neurons

Left panel in A summarizes the current-voltage relation for IK under control conditions and after 10 and 20 min exposures to 1 μM ceramide (C2). The conductance-voltage relation is illustrated in the right panel; the data points have been fitted by the Boltzmann relation and are shown as the continuous lines. Treatment with ceramide for either 10 or 20 min caused a significant reduction in IK and G/Gmax for voltage steps positive to −10 mV. B, the effects of a 20 min treatment with 10 μM ceramide on IK obtained from a representative adult sensory neuron. Traces on the left represent control recordings, the middle traces were recorded after exposure to ceramide, and the right panel illustrates the ceramide-sensitive current. Neurons were held at −60 mV; the traces shown were obtained for voltage steps from −80 to +60 mV in 20 mV increments. The line labelled zero represents the zero current value. Left panel in C summarizes the current-voltage relation for IK under control conditions and after 2, 10 and 20 min exposures to 1 μM ceramide. IK was decreased significantly for voltages positive to +20 mV for the 10 min treatment whereas the suppression was significant for voltages positive to −10 mV for the 20 min exposure. The conductance-voltage relation is illustrated in the right panel; the data points have been fitted by the Boltzmann relation and are shown as the continuous lines. G/Gmax was significantly decreased for voltages positive to +30 mV for the 2 min application and positive to −10 mV for both the 10 and 20 min exposures.

Download figure to PowerPoint

We also examined the capacity of ceramide to inhibit IK in adult sensory neurons; these results are presented in Fig. 5B and C. As in the NGF studies, the total IK obtained at the more depolarized voltages was of greater amplitude in the adult neurons compared to the embryonic neurons. Representative current traces are shown in Fig. 5B where a 20 min exposure to 10 μM ceramide suppressed a large portion of the total IK. The ceramide-sensitive IK (right panel) exhibited the characteristics of a delayed rectifier like current. The inhibitory effects of 10 μM ceramide are summarized in Fig. 5C wherein the average IK was reduced from 7.24 ± 2.43 nA to 2.37 ± 0.86 nA (n= 4) after a 20 min exposure to ceramide. The right panel illustrates the effects of ceramide on the conductance-voltage relation. In these adult sensory neurons, 10 μM ceramide produced a large reduction in G/Gmax at +60 mV with a significant shift in V0.5 to a more hyperpolarized voltage (see Table 3). The findings obtained in the adult sensory neurons are quite similar to those in the embryonic neurons, suggesting that the sphingomyelin-signalling pathway exists at these different developmental stages.

Characterization of the ceramide-sensitive IK

To determine the nature of the IK inhibited by ceramide, the currents remaining after exposure to ceramide were subtracted from their respective control currents to yield the ceramide-sensitive IK. The results obtained for three concentrations of ceramide are illustrated in Fig. 6. Panel A shows representative current traces obtained for 1 μM (left) and 10 μM (right) ceramide. These currents activated rapidly and exhibited little relaxation during the voltage step, except for the more depolarized steps in 10 μM. A steady-state inactivation protocol was used to distinguish different subpopulations of the ceramide-sensitive IK upon activation to +60 mV (see Methods). Using this inactivation protocol, subtracted traces obtained from the same two neurons in Fig. 6A are shown in Fig. 6B. The ceramide-sensitive currents obtained for both 1 and 10 μM showed little time-dependent relaxation even for the most hyperpolarized conditioning prepulse (−100 mV), suggesting that the current inhibited by ceramide is a delayed rectifier-type IK. The current-voltage relation for the activation of ceramide-sensitive IK is summarized in Fig. 6C. These currents began to activate at −10 mV; the amplitudes of these sensitive currents (measured at +60 mV) became larger as the ceramide concentration increased. The steady-state inactivation properties of the ceramide-sensitive IK are shown in Fig. 6D. For conditioning prepulses more depolarized than −20 mV, the ceramide-sensitive current appeared to be largely inactivated; this also is shown in the traces of Fig. 6B. There appeared to be a large increase in the amount of IK that is available for inhibition by ceramide at the more hyperpolarized prepulses. For example, at a conditioning prepulse of −100 mV, the available peak IK approximately doubled for each 10-fold increase in the concentration of ceramide, but for the prepulse to +40 mV the available currents were approximately the same regardless of concentration. In support of our contention that the ceramide-sensitive IK is a delayed rectifier type, we found that pretreatment with the K+ channel blocker, tetraethylammonium (TEA), completely abolished the inhibitory effects of ceramide (data not shown). Treatment with 10 mm TEA significantly reduced the peak value of IK from a control value of 2.02 ± 0.21 to 0.91 ± 0.12 nA (at +60 mV, n= 3). In the presence of TEA, a 20 min exposure to 1 μM ceramide had no effect (0.89 ± 0.12 nA). Similar results were obtained using the steady-state inactivation protocol where, in the presence of TEA, ceramide produced no additional inhibition of IK (measured as the peak value) for either the −100 or +40 mV prepulses.

image

Figure 6. The ceramide-sensitive IK for three concentrations of ceramide

These ceramide-sensitive currents were obtained in embryonic neurons by subtracting the currents remaining after exposure to ceramide from their respective control currents. A, representative current traces for the ceramide-sensitive current obtained for 1 μM (left) and 10 μM (right) ceramide (C2) using the activation protocol. Neurons were held at −60 mV, and the traces shown were obtained for voltage steps from −80 to +60 mV in 20 mV increments. The line labelled zero represents the zero current value. B, ceramide-sensitive currents obtained using the steady-state inactivation. The initial portion of each trace represents the current at the very end of the conditioning prepulse (−100 to +40 mV), and the voltage step was to +60 mV for all traces. The line represents the zero current value. C, summary of the current-voltage relation for activation of the ceramide-sensitive IK for three different concentrations. D, steady-state inactivation properties of the ceramide-sensitive IK.

Download figure to PowerPoint

Prostaglandin E2 produces additional inhibition of IK after ceramide

We have demonstrated previously that the inflammatory prostaglandin, PGE2, enhanced the excitability of rat sensory neurons, in part, through a cyclic AMP/PKA-mediated inhibition of a delayed rectifier-like IK (Nicol et al. 1997; Evans et al. 1999). To determine whether ceramide was acting on the same IK as PGE2, we examined whether PGE2 could produce a further suppression of IK after exposure to ceramide. Figure 7 illustrates representative current traces for these various treatments as well as the ceramide- and PGE2-sensitive IK. Both of the sensitive currents exhibited little time-dependent relaxation, again suggesting that these currents are delayed rectifier types of IK. Figure 8 summarizes the effects of ceramide and PGE2 on IK. Panel A (left) demonstrates that a 20 min exposure to 1 μM ceramide significantly reduced IK from a control value of 2.99 ± 0.47 nA (measured at +60 mV, n= 6) to 1.81 ± 0.22 nA. At this point 1 μM PGE2 was applied to the same neurons in the presence of ceramide. After an additional 20 min exposure to ceramide and PGE2, IK was further decreased to 1.17 ± 0.10 nA. The right panel of Fig. 8A shows the conductance-voltage relation wherein ceramide suppressed G/Gmax by 38 ± 4 % and PGE2 caused an additional decrease of 20 ± 4 % of IK compared to its control values.

image

Figure 7. Sequential inhibitory effects of ceramide and PGE2 on IK

Representative traces obtained from the same embryonic sensory neuron are shown in the top row. The left panel demonstrates the recording under control conditions, the middle panel shows the current remaining after 1 μM ceramide (C2), and the right panel the current remaining after 1 μM PGE2 (E2) in the presence of ceramide. The ceramide- and PGE2-sensitive IK were obtained by subtraction and are shown in the left and right panels, respectively, of the bottom row. The current calibration bar corresponds to their respective panels. Neurons were held at −60 mV; the traces shown were obtained for voltage steps from −80 to +60 mV in 20 mV increments.

Download figure to PowerPoint

image

Figure 8. PGE2 can further inhibit IK after exposure to ceramide in embryonic sensory neurons

Left panel in A, summary of the current-voltage relation for sequential exposures to 1 μM ceramide (C2; 20 min application) and then ceramide and 1 μM PGE2 (E2) for 10 and 20 min (total times of 30 and 40 min, respectively). IK was reduced significantly by ceramide compared to the control (normal Ringer solution) for voltages positive to −10 mV; PGE2 produced a significant inhibition compared to the ceramide values for voltages positive to 0 mV for the 10 min exposure and −20 mV for the 20 min exposure (RM ANOVA). Right panel in A, the conductance-voltage relation for the same neurons represented in the left panel. The data points have been fitted by the Boltzmann relation and are shown as the continuous lines. The suppressions by ceramide and ceramide plus PGE2 were significantly different from control currents for voltages positive to −20 mV and for ceramide compared to ceramide plus PGE2 for voltages positive to −10 mV. B, the sequential effects of ceramide and PGE2 on the steady-state inactivation properties of IK. The left panel shows the inhibitory effects on the peak values of IK obtained for the step to +60 mV as a function of the prepulse voltage. The right panel represents the currents for the step to +60 mV after normalization to the peak IK obtained under control conditions for the prepulse to −100 mV.

Download figure to PowerPoint

Another series of experiments examined the possibility that this additional suppression of IK by PGE2 resulted from merely a longer exposure to ceramide rather than a specific effect of PGE2. Sensory neurons were exposed to 1 μM ceramide for a total period of 40 min. After 20 min, IK was significantly decreased by 34 ± 6 % (2.66 ± 0.39 nA in control vs. 1.71 ± 0.22 nA with ceramide, n= 4, data not shown) and was similar to that described above. An additional 20 min exposure to ceramide (40 min total) reduced IK to 1.40 ± 0.14 nA. To minimize cellular variability, the current amplitudes for ceramide plus PGE2 and ceramide (40 min) were normalized to the IK remaining after the 20 min exposure to ceramide. For this case, the extra exposure to ceramide reduced IK by only an additional 17 ± 3 % whereas exposure to PGE2 decreased IK by an additional 33 ± 4 %. This additional block of IK by PGE2 was significantly different for voltages positive to +20 mV (t test). To further define the target of action of ceramide and PGE2, the effects of these agents on the properties of IK obtained with the steady-state inactivation protocol were examined (see Fig. 8B). Compared to the control values, both ceramide and ceramide plus PGE2 significantly inhibited the peak IK obtained for the step to +60 mV for all pre-conditioning voltages, but there was no additional decrease produced by ceramide plus PGE2 compared to ceramide alone (RM ANOVA). These results suggest that ceramide and PGE2 might be acting on different types of IK, although the activation and recovery kinetics indicate that these currents are likely to be subtypes of the delayed rectifier family. The inhibition of these IK also appears not to involve a modulation of the inactivation properties.

Discussion

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

In this report, we demonstrate that NGF and ceramide directly enhance the capacity of capsaicin-sensitive sensory neurons to generate APs. NGF and ceramide increased the rate at which the ramp depolarized the neuron and could thus account for the decreased latency of firing. Although the number of evoked APs was increased after treatment, the apparent threshold for AP firing was not changed. These results suggest that NGF and ceramide do not alter the membrane properties that set the firing voltage but rather those properties that determine how rapidly it achieves that voltage. The focus of future studies will examine how the modulation of TTX-R INa and IK (and perhaps other currents) is integrated to alter the firing of the AP and thus the coding properties of these NGF-sensitive sensory neurons. NGF and ceramide modify a TTX-R INa and IK in both embryonic and adult sensory neurons in a manner that is consistent with the characteristics of the augmented neuronal excitability. Because similar effects were observed in both adult and embryonic neurons this demonstrates that the embryonic neurons can be a useful model system in which to study the effects of inflammatory agents on neuronal excitability.

In sensory neurons and other cell types, long-term exposure to NGF can increase the current density or expression levels of sodium channels (Mandel et al. 1988; Omri & Meiri, 1990; Aguayo & White, 1992; Fjell et al. 1999). To date, there is little if any information on whether acute administration of NGF affects sodium channels. Our observations indicate that both NGF and ceramide rapidly enhanced the TTX-R INa by ≈1.3-fold, but this may be lower than actual values because a prepulse to remove inactivation was not used, thus some portion of TTX-R INa may have been inactivated. Also, NGF or ceramide shifted activation of TTX-R INa to more hyperpolarized voltages. We believe that this shift did not result from a loss of space clamp in these neurons because the current-voltage relation remained graded with depolarization. Both the slope factor, k, and the reversal potential for TTX-R INa (calculated to be +28 mV) did not change over the entire recording period. Furthermore, after a 20 min exposure to normal recording solution or dihydroceramide the peak TTX-R INa did not shift to more hyperpolarized voltages nor were G/Gmax or k altered. These findings suggest that the clamp was maintained during the recording period and that this current was stable throughout our period of recording.

Recent work by Rush et al. (1998) suggests that in adult rat sensory neurons there might be three different types of TTX-R INa. Molecular studies originally described the PN3/SNS subtype (Akopian et al. 1996; Sangameswaran et al. 1996) and more recent studies have found another distinct TTX-R subtype known as SNS2 (Tate et al. 1998) or NaN (Dib-Hajj et al. 1998b). The mRNA levels for SNS and NaN decreased after axotomy with corresponding changes in the currents (Dib-Hajj et al. 1996, 1998b; Cummins & Waxman, 1997; but see Novakovic et al. 1998). The decreased expression of SNS after axotomy could be reversed by chronic treatment (10-12 days) with NGF (Dib-Hajj et al. 1998a) suggesting that NGF plays an important role in the expression of SNS. Interestingly, inflammation produced by complete Freund's adjuvant increased the expression levels of either PN3 (see Porreca et al. 1999) or SNS2 (Tate et al. 1998) rather than decreasing them as observed with the axotomy model. The physiological significance as to why injury causes a decrease whereas inflammation causes an increase in TTX-R sodium channel expression remains to be determined. Because these studies examined sodium channel expression over long periods of treatment, i.e. days, rather than the short time periods as in our study, it is difficult to speculate upon which TTX-R subtypes NGF/ceramide might act.

We found that NGF and ceramide suppressed an outward IK that was probably a delayed rectifier type of current based on its activation voltage, rapid activation, and slow relaxation kinetics. Although both agents reduced IK, it remains to be determined why ceramide shifted the half-activation voltage to more hyperpolarized potentials (≈10 mV) whereas NGF had no significant effect (Table 3). One possibility is that NGF may have other parallel effects that are not dependent on the actions of ceramide. This notion is, however, not consistent with the observation that GSH completely suppressed the capacity of NGF to augment the AP firing. This is a question that must be resolved in future studies examining the actions of NGF on IK. Unlike the sodium current, more is known about the actions of ceramide on IK. Ceramide inhibited a calcium-dependent IK in smooth muscle cells from coronary artery (Li et al. 1999). In T lymphocytes, ceramide inhibited the Kv1.3-type potassium channel through activation of a Src-like tyrosine kinase (Gulbins et al. 1997) and in oligodendrocytes ceramide blocks an inward rectifier type IK by a ras and raf-1 pathway (Hida et al. 1998). Thus, ceramide appears to act on a variety of potassium currents in a number of cell types through different pathways, and thus modulation of different potassium channels could give rise to enhanced neuronal excitability.

In behavioural models of nociception, treatment with NGF increases the sensitivity to both noxious mechanical and thermal stimulation (Lewin et al. 1993; Woolf et al. 1994; Rueff & Mendell, 1996; Shu & Mendell, 1999a). The time course for sensitization of the thermal response was much shorter (tens of minutes) than the mechanical response (hours). Like the in vivo thermal response, sensitization of isolated sensory neurons by NGF was rapid, reaching maximal effects in ≈6 min. Because NGF acted rapidly on capsaicin-sensitive neurons, we would predict that the neurons we recorded from might correspond, in part, to those neurons whose thermal sensitivity was enhanced by NGF (Lewin et al. 1993). Thus, our findings in isolated sensory neurons are consistent with those made in in vivo models and additionally show that NGF acts directly on sensory neurons rather than through some other immunocompetent cell type to alter neuronal sensitivity.

Based on our finding that the NGF-induced enhancement of excitability was blocked by inhibition of SMase, we speculate that NGF could be acting via p75NTR. This notion is consistent with earlier work where NGF elevated ceramide levels through activation of p75NTR in T9 glioma cells (Dobrowsky et al. 1994). Based on this notion and the observation that p75NTR is co-expressed in nearly all of trk-expressing sensory neurons (≈76 % of the total DRG neurons express trk mRNA; Wright & Snider, 1995) our results would predict that NGF should produce thermal or mechanical hyperalgesia, in part, by sensitizing sensory neurons. Recent results indicated that NGF's capacity to elicit hyperalgesia was through activation of TrkANTR only because injection of NGF in p75NTR knockout mice (Lee et al. 1992) produced both mechanical and thermal hyperalgesia (Bergmann et al. 1998). However, these results must be interpreted with caution in the light of new work. von Schack et al. (2001) showed that the p75NTR‘knockout’ generated by Lee et al. (1992) removed only three of the four cysteine-rich binding domains, which may permit expression of a functional protein. Deletion of all four domains resulted in reduced numbers of Schwann cells as well as a greater loss of neurons in L5 DRG compared to the three of four domain deletion (54 % decrease vs. 39 %, respectively). It will be quite interesting to determine if NGF produces hyperalgesia in the four of four domain deletion p75 knockout mice.

Shu & Mendell (1999b) have shown that a 10 min exposure to 100 ng ml−1 of NGF or NT-4/5 augmented by about two-fold the capsaicin-evoked current (ICAP) in rat sensory neurons. This is thought to result from activation of either TrkANTR or TrkBNTR because the enhancement was blocked by K-252a, a presumed selective inhibitor of TrkNTR tyrosine kinase activity. NT-3 had no effect on ICAP. If the actions of NGF were mediated entirely by the p75NTR, then one might expect NT-3 to augment ICAP. The reasons for the lack of effect by NT-3 are not apparent. The role of NT-3 in sensitization of sensory neurons is complicated by the observation that NT-3 suppressed the electrically evoked release of substance P in an isolated spinal cord preparation (Malcangio et al. 1997). These results suggest that NT-3 might be anti-nociceptive. Recently, Shu & Mendell (2001) showed that the NGF-induced enhancement of ICAP was suppressed partially by inhibitors of PKA. Interestingly, K-252a, the tyrosine kinase inhibitor, can block other kinases like PKA and PKC (Kase et al. 1987). Because of the lack of selectivity of K-252a, it is difficult to conclude that NGF sensitizes ICAP solely through activation of the TrkANTR. Thus, NGF may sensitize sensory neurons through multiple signalling pathways whose interactions are, at present, poorly understood.

The observation that NGF acutely enhanced the excitability of these sensory neurons, which have been grown in the presence of NGF, raises interesting questions regarding the apparent lack of desensitization of this particular receptor/pathway. NGF remaining from the culture medium was washed away with superfusion of normal Ringer solution so that these neurons were free of NGF for approximately 30 min prior to re-exposure. Our results suggest that this NGF-free period is sufficient to establish suitable conditions for NGF induction of sensitization (if such a process is necessary). Consistent with this notion is the finding that the half-time for dissociation of NGF from its low-affinity binding site (site II) is approximately 3 s (Sutter et al. 1979). The issue of desensitization is complicated by our finding that after a 10-30 min washout, neither the ceramide-induced enhancement of AP firing nor the NGF-induced suppression of IK was diminished. In contrast, the NGF-induced enhancement of ICAP returned to control values ≈10 min after removal of NGF, although if external calcium was removed, the enhancement by NGF did not recover (> 1 h; Shu & Mendell, 2001). Desensitization of ICAP can be prevented by the removal of external calcium (Cholewinski et al. 1993; Docherty et al. 1996; Koplas et al. 1997). Therefore, it is difficult to ascertain directly whether recovery of the NGF-induced sensitization of ICAP in normal calcium was due to a process specific to the NGF pathway or whether the recovery was part of the tachyphylaxis of the capsaicin-gated ion channel. It is possible that reversal of the sensitization of the capsaicin-gated current and the sodium and potassium currents exhibit different time courses due to differences in signalling pathways. Additional studies are required to understand the pathways activated by NGF and how these pathways modulate the activity of different ion channels.

Our findings raise an interesting question as to the effector system(s) whereby ceramide modulates the activity of neuronal ion channels. Ceramide can stimulate directly a ceramide-activated protein kinase (CAPK, a serine/ threonine type; Joseph et al. 1993; Mathias et al. 1993; Liu et al. 1994) and a ceramide-activated protein phosphatase (CAPP, a member of the serine/threonine type 2A group; Dobrowsky & Hannun, 1992; Dobrowsky et al. 1993). It seems unlikely that ceramide sensitizes sensory neurons through activation of a phosphatase. Okadaic acid, an inhibitor of phosphatases 1 and 2A, augmented the evoked-release of substance P and calcitonin gene-related peptide from sensory neurons (Hingtgen & Vasko, 1994). Facilitation of peptide release is not consistent with a ceramide-induced increase in phosphatase activity producing the increased AP firing in sensory neurons. Although the roles of CAPK and CAPP in regulating neuronal excitability have not been explored, our observations suggest that modulation of these currents results from kinase(s), rather than phosphatase, activity.

The inflammatory prostaglandins, notably PGI2 and PGE2, produce hyperalgesia in behavioural models (Kress & Reeh, 1996) and also enhance the excitability of sensory neurons (Hingtgen et al. 1995; Nicol et al. 1997). Interestingly, both the TTX-R INa and IK currents are modulated by ceramide in a manner similar to that produced by PGE2. We have shown previously that PGE2 increased the number of evoked APs via cyclic AMP-dependent activation of protein kinase A (PKA; Cui & Nicol, 1995; Nicol et al. 1997). In sensory neurons, PKA plays a critical role in augmenting the amplitude and rate of inactivation of the TTX-R INa (England et al. 1996; Gold et al. 1996). In addition, activation of PKA results in a suppression of a delayed rectifier-like IK (Evans et al. 1999). Therefore, modulation of these currents could result from activation of multiple protein kinases. In support of this notion, we found that PGE2 could further suppress IK by ≈35 % after exposure to ceramide. This suggests that these agents act on different channel subtypes or that their different effector actions may be additive at the same target. During inflammation it is possible that there are multiple parallel signalling pathways that either together or separately converge on the TTX-R INa and IK currents. Therefore, modulation of these two currents, and possibly others, by inflammatory mediators can result in the heightened sensitivity to noxious stimulation observed in behavioural models. Future studies exploring the roles of the sphingomyelin signalling pathway may prove significant in our understanding of how neurotrophins and novel signalling lipids contribute to the regulation of neuronal activity.

References

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements
  • Aguayo, L. G. & White, G. (1992). Effects of nerve growth factor on TTX- and capsaicin-sensitivity in adult rat sensory neurons. Brain Research 570, 6167.
  • Akopian, A. N., Sivilotti, L. & Wood, J. N. (1996). A tetrodotoxin-resistant voltage-gated sodium channel expressed in sensory neurons. Nature 379, 257262.
  • Ballou, L. R., Laulederkind, S. J. F., Rosloneic, E. F. & Raghow, R. (1996). Ceramide signalling and the immune response. Biochimica et Biophysica Acta 1301, 273287.
  • Bergmann, I., Reiter, R., Toyka, K. V. & Koltzenburg, M. (1998). Nerve growth factor evokes hyperalgesia in mice lacking the low-affinity neurotrophin receptor p75. Neuroscience Letters 255, 8790.
  • Cholewinski, A., Burgess, G. M. & Bevan, S. (1993). The role of calcium in capsaicin-induced desensitization in rat cultured dorsal root ganglion neurons. Neuroscience 55, 10151023.
  • Cui, M. L. & Nicol, G. D. (1995). Cyclic AMP mediates the prostaglandin E2-induced potentiation of bradykinin excitation in rat sensory neurons. Neuroscience 66, 459466.
  • Cummins, T. R. & Waxman, S. G. (1997). Down-regulation of TTX-resistant sodium currents and upregulation of a rapidly repriming TTX-sensitive sodium current in small spinal sensory neurons after nerve injury. Journal of Neuroscience 17, 35033514.
  • Cunha, F. Q., Poole, S., Lorenzetti, B. B. & Ferriera, S. H. (1992). The pivotal role of tumour necrosis factor α in the development of inflammatory hyperalgesia. British Journal of Pharmacology 107, 660664.
  • Dib-Hajj, S., Black, J. A., Cummins, T. R., Kenney, A. M., Kocsis, J. D. & Waxman, S. G. (1998a). Rescue of α-SNS sodium channel expression in small dorsal root ganglion neurons after axotomy by nerve growth factor in vivo. Journal of Neurophysiology 79, 26682676.
  • Dib-Hajj, S., Black, J. A., Felts, P. & Waxman, S. G. (1996). Down-regulation of transcripts for Na channel α-SNS in spinal sensory neurons following axotomy. Proceedings of the National Academy of Sciences of the USA 93, 1495014954.
  • Dib-Hajj, S., Tyrrell, L., Black, J. A. & Waxman, S. G. (1998b). NaN, a novel voltage-gated Na channel is expressed preferentially in peripheral sensory neurons and down-regulated after axotomy. Proceedings of the National Academy of Sciences of the USA 95, 89638968.
  • Dobrowsky, R. T. & Hannun, Y. A. (1992). Ceramide stimulates a cytosolic protein phosphatase. Journal of Biological Chemistry 267, 50485051.
  • Dobrowsky, R. T., Kamibayashi, C., Mumby, M. C. & Hannun, Y. A. (1993). Ceramide activates heterotrimeric protein phosphatase 2A. Journal of Biological Chemistry 268, 1552315530.
  • Dobrowsky, R. T., Werner, M. H., Castellino, A. M. Chao, M. V. & Hannun, Y. A. (1994). Activation of the sphingomyelin cycle through the low-affinity neurotrophin receptor. Science 265, 15961599.
  • Docherty, R. J., Yeats, J. C., Bevan, S. & Boddeke, H. W. (1996). Inhibition of calcineurin inhibits the desensitization of capsaicin-evoked currents in cultured dorsal root ganglion neurones from adult rats. Pflügers Archiv 431, 828837.
  • England, S., Bevan, S. & Docherty, R. J. (1996). PGE2 modulates the tetrodotoxin-resistant sodium current in neonatal rat dorsal root ganglion neurones via a cyclic AMP-protein kinase A cascade. Journal of Physiology 495, 429440.
  • Evans, A. R., Vasko, M. R. & Nicol, G. D. (1999). The cAMP transduction cascade mediates the PGE2-induced inhibition of potassium currents in rat sensory neurones. Journal of Physiology 516, 163178.
  • Ferreira, S. H., Lorenzetti, B. B., Bristow, A. F. & Poole, S. (1988). Interleukin-1β as a potent hyperalgesic agent antagonized by a tripeptide analogue. Nature 334, 698700.
  • Fjell, J., Cummins, T. R., Davis, B. M., Albers, K. M., Fried, K., Waxman, S. G. & Black, J. A. (1999). Sodium channel expression in NGF-overexpressing transgenic mice. Journal of Neuroscience Research 57, 3947.
  • Gold, M. S., Reichling, D. B., Shuster, M. J. & Levine, J. D. (1996). Hyperalgesic agents increase a tetrodotoxin-resistant Na+ current in nociceptors. Proceedings of the National Academy of Sciences of the USA 93, 11081112.
  • Gulbins, E., Szabo, I., Baltzer, K. & Lang, F. (1997). Ceramide-induced inhibition of T lymphocyte voltage-gated potassium channel is mediated by tyrosine kinases. Proceedings of the National Academy of Sciences of the USA 94, 76617666.
  • Hamill, O. P., Marty, A., Neher, E., Sakmann, B. & Sigworth, F. J. (1981). Improved patch-clamp techniques for high resolution current recording from cells and cell-free membrane patches. Pflügers Archiv 391, 85100.
  • Hida, H., Takeda, M. & Soliven, B. (1998). Ceramide inhibits inwardly rectifying K+ currents via a ras- and raf-1-dependent pathway in cultured oligodendrocytes. Journal of Neuroscience 18, 87128719.
  • Hingtgen, C. M. & Vasko, M. R. (1994). The phosphatase inhibitor, okadaic acid, increases peptide release from rat sensory neurons in culture. Neuroscience Letters 178, 135138.
  • Hingtgen, C. M., Waite, K. J. & Vasko, M. R. (1995). Prostaglandins facilitate peptide release from rat sensory neurons by activating the adenosine 3′,5′-monophosphate transduction cascade. Journal of Neuroscience 15, 54115419.
  • Holzer, P. (1991). Capsaicin: cellular targets, mechanisms of action, and selectivity for thin sensory neurons. Pharmacological Reviews 43, 143201.
  • Joseph, C., Byun, H.-S., Bittman, R. & Kolesnick, R. N. (1993). Substrate recognition by ceramide-activated protein kinase. Evidence that kinase activity is proline-directed. Journal of Biological Chemistry 268, 2000220006.
  • Kase, H., Iwashi, K., Nakanishi, S., Matsuda, Y., Yamada, K., Takahashi, M., Murakata, C., Sato, A. & Kaneko, M. (1987). K-252 compounds, novel and potent inhibitors of protein kinase C and cyclic nucleotide-dependent kinases. Biochemical and Biophysical Research Communications 142, 436440.
  • Koplas, P. A., Rosenberg, R. L. & Oxford, G. S. (1997). The role of calcium in the desensitization of capsaicin responses in rat dorsal root ganglion neurons. Journal of Neuroscience 17, 35253537.
  • Kress, M. & Reeh, P. W. (1996). Chemical excitation and sensitization in nociceptors. In Neurobiology of Nociceptors, ed. Belmonte, C. & Cervero, F., pp. 258297. Oxford University Press, Oxford
  • Lee, K.-F., Li, E., Huber, J., Landis, S., Sharpe, A. H., Chao, M. V. & Jaenisch, R. (1992). Targeted mutation of the gene encoding the low affinity NGF receptor p75 leads to deficits in the peripheral sensory nervous system. Cell 69, 737749.
  • Levade, T. & Jaffrezou, J.-P. (1999). Signalling sphingomyelinases: which, where, how and why Biochimica et Biophysica Acta 1438, 117.
  • Lewin, G. R., Ritter, A. M. & Mendell, L. M. (1993). Nerve growth factor-induced hyperalgesia in the neonatal and adult rat. Journal of Neuroscience 13, 21362148.
  • Li, P.-L., Zhang, D. X., Zou, A.-P. & Campbell, W. B. (1999). Effect of ceramide on KCa channel activity and vascular tone in coronary arteries. Hypertension 33, 14411446.
  • Lindsay, R. M. (1988). Nerve growth factors (NGF, BDNF) enhance axonal regeneration but are not required for survival of adult sensory neurons. Journal of Neuroscience 8, 23942405.
  • Liu, B., Andrieu-Abadie, N., Levade, T., Zhang, P., Obeid, L. M. & Hannun, Y. A. (1998). Glutathione regulation of natural sphingomyelinase in tumor necrosis factor-α-induced cell death. Journal of Biological Chemistry 273, 1131311320.
  • Liu, B. & Hannun, Y. A. (1997). Inhibition of the neutral magnesium-dependent sphingomyelinase by glutathione. Journal of Biological Chemistry 272, 1628116287.
  • Liu, J., Mathias, S., Yang, Z. & Kolesnick, R. N. (1994). Renaturation and tumor necrosis factor-α stimulation of a 97-kDa ceramide-activated protein kinase. Journal of Biological Chemistry 269, 30473052.
  • Malcangio, M., Garrett, N. E., Cruwys, S. & Tomlinson, D. R. (1997). Nerve growth factor- and neurotrophin-3-induced changes in nociceptive threshold and the release of substance P from the rat isolated spinal cord. Journal of Neuroscience 17, 84598467.
  • Mandel, G., Cooperman, S. S., Maue, R. A., Goodman, R. H. & Brehm, P. (1988). Selective induction of brain type II Na+ channels by nerve growth factor. Proceedings of the National Academy of Sciences of the USA 85, 924928.
  • Mathias, S., Pena, L. A. & Kolesnick, R. N. (1998). Signal transduction of stress via ceramide. Biochemical Journal 335, 465480.
  • Mathias, S., Younes, A., Kan, C.-C., Orlow, I., Joseph, C. & Kolesnick, R. N. (1993). Activation of the sphingomyelin signaling pathway in intact EL4 cells and in a cell-free system by IL-1β. Science 259, 519522.
  • Nicol, G. D., Vasko, M. R. & Evans, A. R. (1997). Prostaglandins suppress an outward potassium current in embryonic rat sensory neurons. Journal of Neurophysiology 77, 167176.
  • Nicol, G. D., Vasko, M. R. & Zhang, Y.-H. (2001). Nerve growth factor enhances the excitability of capsaicin-sensitive rat sensory neurons. Society for Neuroscience Abstracts 27, 56.5
  • Novakovic, S. D., Tzoumaka, E., McGivern, J. G., Haraguchi, M., Sangaameswaran, L., Gogas, K. R., Eglen, R. M. & Hunter, J. C. (1998). Distribution of the tetrodotoxin-resistant sodium channel PN3 in rat sensory neurons in normal and neuropathic conditions. Journal of Neuroscience 18, 21742187.
  • Omri, G. & Meiri, H. (1990). Characterization of sodium currents in mammalian sensory neurons cultured in serum-free defined medium with and without nerve growth factor. Journal of Membrane Biology 115, 1329.
  • Porreca, F., Lai, J., Bian, D., Wegert, S., Ossipov, M. H., Eglen, R. M., Kassotakis, L., Novakovic, S., Rabert, D. K., Sangaameswaran, L. & Hunter, J. C. (1999). A comparison of the potential role of the tetrodotoxin-insensitive sodium channels PN3/SNS and NaN/SNS2, in rat models of chronic pain. Proceedings of the National Academy of Sciences of the USA 96, 76407644.
  • Rueff, A. & Mendell, L. M. (1996). Nerve growth factor and NT-5 induce increased thermal sensitivity of cutaneous nociceptors in vitro. Journal of Neurophysiology 76, 35933596.
  • Rush, A. M., Brau, M. E., Elliott, A. A. & Elliott, J. R. (1998). Electrophysiological properties of sodium current subtypes in small cells from adult rat dorsal root ganglia. Journal of Physiology 511, 771789.
  • Sangameswaran, L., Delagado, S. G., Fish, L. M., Koch, B. D., Jakeman, L. B., Stewart, G. R., Sze, P., Hunter, J. C., Eglen, R. M. & Herman, R. C. (1996). Structure and function of a novel voltage-gated tetrodotoxin-resistant sodium channel specific to sensory neurons. Journal of Biological Chemistry 271, 59535956.
  • Schütze, S., Machleidt, T. & Krönke, M. (1994). The role of diacylglycerol and ceramide in tumor necrosis factor and interleukin-1 signal transduction. Journal of Leukocyte Biology 56, 533541.
  • Schweizer, A., Feige, U., Fontana, A., Müller, K. & Dinarello, C. A. (1988). Interleukin-1 enhances pain reflexes. Mediation through increased prostaglandin E2 levels. Agents and Actions 25, 246251.
  • Shu, X.-Q. & Mendell, L. M. (1999a). Neurotrophins and hyperalgesia. Proceedings of the National Academy of Sciences of the USA 96, 76937696.
  • Shu, X.-Q. & Mendell, L. M. (1999b). Nerve growth factor acutely sensitizes the response of adult rat sensory neurons to capsaicin. Neuroscience Letters 274, 159162.
  • Shu, X.-Q. & Mendell, L. M. (2001). Acute sensitization by NGF of the response of small-diameter sensory neurons to capsaicin. Journal of Neurophysiology 86, 29312938.
  • Sutter, A., Riopelle, R. J., Harris-Warrick, R. M. & Shooter, E. M. (1979). Nerve growth factor receptors. Characterization of two distinct classes of binding sites in chick embryo sensory ganglia cells. Journal of Biological Chemistry 254, 59725982.
  • Tate, S., Benn, S., Hick, C., Trezise, D., John, V., Mannion, R. J., Costigan, M., Plumpton, C., Grose, D., Gladwell, Z., Kendall, G., Dale, K., Bountra, C. & Woolf, C. (1998). Two sodium channels contribute to the TTX-R sodium current in primary sensory neurons. Nature Neuroscience 1, 653655.
  • Vasko, M. R., Campbell, W. B. & Waite, K. J. (1994). Prostaglandin E2 enhances bradykinin-stimulated release of neuropeptides from rat sensory neurons in culture. Journal of Neuroscience 14, 49874997.
  • Von Schack, D., Casademunt, E., Schweigreiter, R., Meyer, M., Bibel, M. & Dechant, G. (2001). Complete ablation of the neurotrophin receptor p75NTR causes defects in the nervous and vascular system. Nature Neuroscience 4, 977978.
  • Woolf, C. J., Safieh-Garabedian, B., Ma, Q.-P., Crilly, P. & Winter, J. (1994). Nerve growth factor contributes to the generation of inflammatory sensory hypersensitivity. Neuroscience 62, 327331.
  • Wright, D. E. & Snider, W. D. (1995). Neurotrophin receptor mRNA expression defines distinct populations of neurons in rat dorsal root ganglia. Journal of Comparative Neurology 351, 329338.
  • Zhang, Y.-H., Vasko, M. R. & Nicol, G. D. (2000). Ceramide enhances the excitability of capsaicin-sensitive sensory neurons. Society for Neuroscience Abstracts 26, 1694

Acknowledgements

  1. Top of page
  2. Abstract
  3. Methods
  4. Results
  5. Discussion
  6. References
  7. Acknowledgements

We are grateful to Drs Jim Kenyon and John Schild for discussions and comments. This work was supported by a grant from NINDS (NS37951).