Cellular properties of lateral spinal nucleus neurons in the rat L6–S1 spinal cord. Conventional intracellular recordings were made from 26 lateral spinal nucleus (LSN) neurons in slices of L6–S1 spinal cord from 10- to 15-day-old rats. At rest, LSN neurons did not fire spontaneous action potentials. With injection of a positive current pulse, action potentials had an amplitude of 72 ± 7 (SD) mV and duration at half-peak height of 0.75 ± 0.22 ms. Action potentials were followed by an afterpotential. Most LSN neurons (13/17) exhibited only an afterhyperpolarization (AHP); four neurons exhibited both a fast and a slow AHP separated by an afterdepolarization (ADP). For LSN neurons that exhibited only an AHP, a slow ADP could be identified during bath application of apamin (100 nM). Four of 11 LSN neurons showed a postinhibitory rebound (PIR). Two types of PIR were noted, one with high threshold and low amplitude and the other with low threshold and high amplitude. The PIR with high amplitude was partially blocked in 0 mM Ca2+/high Mg2+ (10 mM) recording solution. Repetitive firing properties were examined in 17 LSN neurons. On the basis of the ratio of the slopes between initial instantaneous firing and steady-state firing frequencies, neurons with low spike frequency adaptation (SFA, 8/17) and high SFA (4/17) were identified. In addition, 2/17 LSN neurons exhibited biphasic repetitive firing patterns, which were composed of a fast SFA, delayed excitation, and low SFA; another two neurons showed only delayed excitation. Plateau potentials also were found in two LSN neurons. Dorsal root stimulation revealed that most LSN neurons (12/13) had polysynaptic postsynaptic potentials (PSP); only one neuron exhibited a monosynaptic PSP. Electrical stimulation of the dorsal root evoked prolonged discharges in low SFA neurons and a short discharge in high SFA neurons. Intrinsic properties were modulated by bath application of substance P (SP). Membrane potentials were depolarized in all eight LSN neurons tested, and membrane resistance was either increased (n = 3) or decreased (n = 2). Both instantaneous firing and steady-state firing were facilitated by SP. In addition, oscillation of membrane potentials were induced in three LSN neurons. These results demonstrate that LSN neurons exhibit a variety of intrinsic properties, which may significantly contribute to sensory processing, including nociceptive processing.
Located ventrolateral to the superficial spinal dorsal horn, the lateral spinal nucleus (LSN) first was described byGwyn and Waldron (1968, 1969). Morphological studies found ascending projections of LSN neurons to a wide variety of supraspinal sites, including the thalamus and lateral periaqueductal gray (PAG) (Battaglia and Rustioni 1992; Harmann et al. 1988), the hypothalamus and telencephalon (Burstein et al. 1987), the amygdala and orbital cortex (Burstein and Potrebic 1993), the tractus solitarius nucleus (Esteves et al. 1993), and the parabrachial nucleus (Ding et al. 1995; Feil and Herbert 1995). LSN neurons also receive descending projections from the raphe nuclei, brain stem reticular formation nuclei, dorsal column nuclei, and PAG (Carlton et al. 1985; Masson et al. 1991). In addition, LSN neurons can be activated by peripheral mechanical stimulation (Menetrey et al. 1980).
LSN neurons project to neurons in spinal lamina I, II, V, and VII (Jansen and Loewy 1997) and receive peptidergic input from local spinal cord neurons (Cliffer et al. 1988). Using immunochemical techniques, LSN neurons were found to contain a variety of peptide receptors, including substance P (Battaglia and Rustioni 1992; Ding et al. 1995; Li et al. 1997; Marshall et al. 1996), neuropeptide Y (Zhang et al. 1995), and kappa-opioid receptors (Schafer et al. 1994). LSN neurons also were found to contain peptides, such as calcitonin gene-related peptide (Conrath et al. 1989) and vasoactive intestinal polypeptide (Fuji et al. 1983; Leah et al. 1988; Sasek et al. 1991). A significant property of LSN neurons (Leah et al. 1988) is that their axons that project to supraspinal sites contain the highest percentage of neuropeptides. The foregoing suggests that the LSN may be involved in the transmission and modulation of afferent input, including nociception. In support of this, Herdegen et al. (1994)reported expression of nitric oxide synthase in LSN neurons after noxious stimulation of the rat hindpaw with formalin.
The behavior and properties of LSN neurons, however, have received little attention. For example, limited information is available about their receptive fields (Menetrey et al. 1980) and synaptic input and intrinsic properties of LSN neurons, which are important factors in determining neuron firing behavior (Lopez-Garcia and King 1994; McCormick 1990; McCormick et al. 1992; Thomson et al. 1989), have not been described. In spinal dorsal horn neurons, synaptic responses can be divided into monosynaptic and polysynaptic excitatory and inhibitory postsynaptic potentials (EPSPs and IPSPs) (Inokuchi et al. 1992; Jiang et al. 1995; King et al. 1988; Thomson et al. 1989; Yajiri et al. 1997; Yoshimura and Jessell 1989b). Dorsal horn neurons also express a variety of intrinsic properties, such as repetitive firing patterns, delayed excitation, plateau potentials, and oscillation (Hochman et al. 1994; Jiang et al. 1995; Lopez-Garcia and King 1994; Murase and Randic 1983;Yoshimura and Jessell 1989a). A good correlation between firing patterns evoked by synaptic input and intrinsic properties was found in superficial and deep spinal dorsal horn neurons (Lopez-Garcia and King 1994; Thomson et al. 1989). In addition, intrinsic properties can be modulated by activation of excitatory amino acid or peptide receptors (Hochman et al. 1994; Morisset and Nagy 1996; Russo et al. 1997), which then change neuron excitability. Accordingly, the goals of these experiments were to examine the intrinsic properties of LSN neurons and their synaptic response to afferent input and to study modulation of these properties by substance P (SP). Parts of this work have been presented in abstract form (Jiang et al. 1997).
Rat pups of both sexes aged postnatal days 10–15 were used. Under ether anesthesia, body temperature was reduced by immersing the pup below the cervical level into an ice-water pool. The dissection began when skin temperature fell to 20–22°C (∼5–10 min) and respiration became shallow. A laminectomy was performed to expose the lower-thoracic and lumbosacral spinal cord. A block of lumbosacral spinal cord (∼8–10 mm) with attached dorsal roots was excised quickly and placed in oxygenated (95% O2-5% CO2) Ringer solution, which contained (in mM): 124 NaCl, 5 KCl, 1.2 KH2PO4, 2.4 CaCl2, 1.3 MgSo4, 26 NaHCO3, and 10 glucose, pH 7.4. The pia matter on the spinal cord surface was removed using surgical forceps and the spinal cord was trimmed manually to a 4-to 5-mm lumbosacral block. The block was placed in a Plexiglass cutting chamber in an Oxford vibratome and fixed to an agar block with cyanoacrylic glue. The chamber was filled with aerated Ringer solution and maintained at 24°C. The spinal cord block was sectioned to yield several transverse slices of 300–400 μm with short (3–4 mm) dorsal rootlets; two to four slices from the L6–S1region (middle of the block) were collected and removed to an incubation chamber. The procedure from removal of the spinal cord block to sectioning of slices for recording was performed in 4–5 min. The slices were incubated in Ringer solution at 30–31°C for ∼1 h before recording.
Slices were transferred to a recording chamber, placed on a nylon mesh, and perfused with oxygenated Ringer solution (95% O2-5% CO2) at rate of 2.5–3 ml/min. Using an inverted microscope, the LSN was identified as a semitransparent triangle lateral and ventral to the substantia gelatinosa. Intracellular recordings were performed using conventional glass electrodes with impedances of 100–120 MΩ when filled with 2 M potassium acetate and 2% neurobiotin. The electrode was driven by a micromanipulator (Newport, Irvine, CA), which was controlled by a pulse generator, in steps of 2–4 μm. To assist penetration into a neuron, a positive current pulse (20 nA, 100 ms in duration) was passed through the electrode or excessive capacitance compensation was applied for 5 ms. Immediately after a neuron was impaled, hyperpolarizing current was applied to eliminate spontaneous firing. The hyperpolarizing current was gradually withdrawn during 1–5 min, and stable intracellular recordings could be maintained for 0.5–4 h.
Electrical activity of LSN neurons was amplified using an Axoclamp-2A (Axon Instruments, Burlingame, CA) in bridge mode and recorded on video tape. Data were collected both on-line and off-line; off-line analysis employed VTX (Keithley, Taunton, MA) and Visual Basic programming software. To generate synaptic input, a co-axial stainless steel stimulating electrode was positioned on the dorsal root. Intrinsic properties of LSN neurons were studied by intracellular injection of either positive or negative current pulses with various durations and intensities.
Recording electrodes contained 2% neurobiotin (Vector, Burlingame, CA). At the end of experiments, a 1-nA positive current was injected through the electrode (200-ms pulse duration at 3 Hz for 10 min). Slices were fixed in 4% paraformaldehyde and 0.2% picric acid in a 0.1 M sodium phosphate buffer solution (PBS, pH 7.4) overnight at 4°C. The tissue was sectioned into 40 μm slices and transferred to PBS containing 0.5% Triton-X on a shaker for 2 h. After two PBS washes, the slices were treated with avidin conjugate (avidin-biotin-horseradish peroxidase complex, diluted 1,000 times in PBS) for 2 h on a shaker. Afterward, the slices were washed and treated with 0.05% diaminobenzidine and 0.003% H2O2 in PBS. The sections were mounted on gelatin-coated slides, dried, defatted, and coverslipped.
Under an inverted microscope, the LSN appears as a semitransparent region ventrolateral to the substantia gelatinosa. Figure1 shows the location of the LSN and a neuron labeled with neurobiotin in the upper part of the LSN. LSN neurons having a resting membrane potential more negative than −55 mV, action potentials of more than 60 mV amplitude, and overshoot were considered healthy and subjected to the experimental protocols. All reported values are mean ± 1 SD.
Action potentials and afterpotentials
The average resting membrane potential of the 26 LSN neurons studied was 71 ± 6.2 mV. At rest, LSN neurons did not fire action potentials. Thus a positive pulse (10 ms) was injected into LSN neurons to generate action potentials. Characteristics of these LSN neurons are summarized in Table 1. Most LSN neurons (13/17) exhibited a monophasic afterhyperpolarization (AHP; Fig.2 A). The remaining four LSN neurons exhibited compound afterpotentials: a fast and a slow AHP and a afterdepolarization (ADP) (Fig. 2 A). To examine the duration and amplitude of afterpotentials, these 17 LSN neurons were injected with 500-ms positive pulses or a positive DC current adjusted manually to produce action potentials; we did not attempt to separate afterpotentials in the four neurons with compound afterpotentials. The amplitude and duration of the AHP are summarized in Table 1. It is known that fast and slow AHPs are mediated through different channels (Sah 1996). The absence of an ADP in most LSN neurons made it impossible to distinguish AHP components. Thus we used apamin (100 nM), a calcium-dependent potassium channel blocker that is reported to block the slow AHP (Sah 1996), to examine the composition of AHPs. The results (Fig. 2 B) showed that the duration of AHP in four LSN neurons studied was reduced significantly (control: 93.4 ± 26.3 ms; apamin: 30 ± 10 ms) after bath application of apamin, during which an ADP became apparent (peak at 42.5 ± 6.3 ms) (Fig. 2 B), suggesting that the apparent monophasic AHP in these neurons is conducted by multiple channels.
Current-voltage (I-V) relationships were examined using hyperpolarizing pulses (300 ms) in 11 LSN neurons. The majority of LSN neurons (7/11) exhibited a predominately linear I-Vrelation, in which the average of four steady-state membrane potentials in response to the same intensity of intracellular stimulation was taken and plotted against each current step (Fig.3 A). In four LSN neurons, a time-dependent inward rectification was observed (Fig. 3 B), in which the initial hyperpolarization of membrane potentials in response to negative pulses decayed to steady-state levels 200 ms after onset of the current pulses. The I-V relationships (average of 4 trials) were plotted as peak and steady-state membrane potentials against each current step. Whereas the steady-state I-V plot is characterized by an upward bend (Fig. 3 B, ■), the peakI-V relationship is linear.
Postinhibitory rebound (PIR) is characterized by a transient depolarization of the membrane potential at the end of current injection (Fig. 3 B). Although PIR is relatively rare in LSN neurons (4/11), two types of PIR were noted as shown in Fig.4. One type of PIR (2/4 LSN neurons) required a higher stimulating intensity to be activated (Fig.4 A). In contrast, a lower stimulating intensity was required to activate PIR in two other LSN neurons (Fig. 4 B). In addition, burst-like firing could be produced in these two LSN neurons by higher intensity current. Because Ca2+ current has been reported to underlie burst firing (Huguenard 1996), we applied a low Ca2+ (0 Ca2+ and 10 mM Mg2+) bath solution to test whether Ca2+current participates in PIR; PIR was blocked partially by the low Ca2+ bath solution (Fig. 4 C).
Repetitive firing properties of LSN neurons were studied by injection of positive current pulses (3-s duration). The majority of LSN neurons studied (12/16) showed a quick onset of action potentials in response to current steps. Although firing frequency of these neurons increased with increase in current intensity, it adapted (spike frequency adaptation, SFA) during each current step. The LSN neurons studied can be divided into low SFA (n = 8) and high SFA (n = 4) neurons (Fig.5 Ab) according to the ratio of frequency slopes (Fig. 5 Aa). High SFA neurons fired action potentials that adapted quickly (Fig. 5 Ab), whereas low SFA neurons continued firing action potentials during current pulses with little adaptation (Fig. 5 Aa).
Other types of repetitive firing also were observed in LSN neurons. Two LSN neurons apparently showed a delayed onset of action potentials (delayed excitation, DE) at the beginning of current injection (Fig.6 A). Unlike LSN neurons that adapted during stimulation, the firing rate of these neurons increased during current injection. In addition, another two LSN neurons exhibited a mix of SFA and DE (Fig. 6 B). In these two neurons, current pulses with low intensities produced a gap in action potentials between the initial and steady-state firing. As the current intensity increased, the gap was filled with firing that had the lowest rate during the current pulses. Prolonged repetitive firing was observed after termination of the current pulse in another two LSN neurons (Fig. 6 C), which was apparently due to a prolonged depolarization of membrane potential after the positive pulses (plateau potential). One of these neurons also showed a DE (Fig.6 Cb).
Synaptic responses were examined in 13 LSN neurons by electrical stimulation of an attached dorsal root. For subthreshold postsynaptic responses, most LSN neurons (12/13) gave polysynaptic potentials; one neuron exhibited an apparent monosynaptic postsynaptic potential (Fig.7 A). For high-intensity stimulation, either short (n = 5) or prolonged firing (n = 7) was produced in these LSN neurons. A correlation between intrinsic firing property and synaptic input-induced firing was examined in eight LSN neurons. It was found that LSN neurons with low SFA (n = 3) had prolonged firing (Fig. 7 B), whereas LSN neurons with high SFA (n = 3) had short firing (Fig. 7 C) in response to electrical stimulation. However, the two LSN neurons with mixed repetitive firing properties (initial SFA and DE) responded to electrical stimulation with short firing (data not shown).
Effects of SP on cellular properties
The effect of SP was examined in eight LSN neurons. Similar to neurons in the spinal dorsal horn (Murase et al. 1989), SP dose-dependently depolarized all eight LSN neurons (for 10−7 M: 3.61 ± 1.12 mV, 4.2 ± 1.7 min; for 10−6 M: 5.79 ± 1.42 mV, 5.3 ± 2.1 min; for 10−5 M: 8.64 ± 2.23 mV, 6.5 ± 1.8 min). Membrane resistance showed an increase (n = 3) from 20 to 33% and decrease (n = 2) from 16 to 24% in response to SP bath application at different concentrations. Figure8 A shows that SP dose-dependently depolarized membrane potential and increased membrane impedance in a LSN neuron. To examine the effect of SP on repetitive firing properties, the membrane potential of LSN neurons was held manually at control levels by injecting a hyperpolarizing current. The results from five LSN neurons studied showed that both the instantaneous and steady-state firing were all facilitated by SP and the effect lasted ∼4 min (Fig. 8 B). In three of eight LSN neurons, rhythmic changes in membrane potentials were produced by SP application (frequency range: 0.08–0.16 Hz, Fig. 8 C). Depending on the resting membrane potential, the neurons either only expressed oscillation of membrane potential (Fig. 8 Ca) or fired action potentials at the peak of depolarization (Fig.8 Cb).
LSN neurons have been found to receive peripheral input and to have reciprocal connections with supraspinal sites (Burstein et al. 1987; Feil and Herbert 1995; Harmann et al. 1988; Masson et al. 1991; Menetrey et al. 1980), suggesting that the LSN may participate in sensory processing, including nociception. To better understand the role of the LSN in sensory processing, the intrinsic cellular properties and the synaptic responses of LSN neurons were examined here. LSN neurons in segments L6–S1 in the rat spinal cord were found to exhibit a variety of cellular properties: inward rectification, PIRs, low and high SFAs, DE, biphasic firing, and SP-induced oscillation. It also was demonstrated that the majority of LSN neurons receive polysynaptic input, consistent with the absence of evidence for direct input via the dorsal roots, and exhibit a strong correlation between SFA properties and responses produced by synaptic input. Thus the cellular properties likely contribute to the mechanisms by which LSN neurons process synaptic input. In addition, we found in the small sample studied that cellular properties of LSN neurons could be modified by SP.
The passive membrane properties, membrane resistances, and time constants reflect whole membrane channels and cell size. Depending on animal age, location, and the experimental preparation, passive membrane properties can vary significantly. The membrane resistance (mean 78.5 MΩ, range 37–167 MΩ) of LSN neurons is similar to membrane resistances of deep dorsal neurons, both in in vitro slice (54.5 MΩ, range 11–138 MΩ) (King et al. 1988) and in vivo adult rat preparations (mean 38 MΩ, range 14–141 MΩ) (Jiang et al. 1995). However, the membrane resistance of LSN neurons differed from those of superficial dorsal horn neurons, both in adult rat slice [257 ± 17.7 MΩ (Yoshimura and Jessell 1989a) and 241 ± 12 MΩ (Yoshimura and Jessell 1989b)] and immature rat slice preparations (48–267 MΩ) (Murase and Randic 1983). The membrane time constant (mean 10.6 ms, range 5.6–21.1 ms) in LSN neurons was also similar to that of deep dorsal horn neurons (mean 9.1 ms, range 1.8–19.7 ms) (Jiang et al. 1995). The time constants reported in superficial dorsal horn neurons differ depending on the animal and the preparation. Whereas the membrane time constants of superficial dorsal horn neurons in an in vivo cat preparation ranged between 0.8 and 2.0 ms (Iggo et al. 1988), they were 21.3 ± 1.9 ms in an in vitro rat slice preparation (Yoshimura and Jessell 1989b). This difference may be due to the different animal species used in these experiments or the different experimental preparation. Nevertheless the similarities in passive membrane properties between LSN neurons and deep dorsal horn neurons suggest that these two groups of neurons may have similar size.
The mean amplitude of action potentials (72.3 ± 6.6 mV) and resting membrane potentials (−71.0 ± 6.2 mV) of LSN neurons were also similar to those of deep dorsal horn neurons (amplitude of action potentials, 77 ± 11.8; resting membrane potentials, −67 ± 8 mV) (King et al. 1988). However, the resting membrane potentials of LSN neurons differ from those found in an in vivo adult rat preparation (−60.9 ± 3.9 mV) (Jiang et al. 1995), which may be due to the different experimental preparations. The mean width of the action potential at half-amplitude of LSN neurons (0.75 ± 0.22 ms) was also different from that determined in the in vivo adult preparation (0.33 ± 0.15), which may be due to age because the animals used in this study were immature. The conflicting results regarding the width of action potentials in the young rat slice preparation [1.4 ± 0.5 ms (King et al. 1988); 0.82 ms (Thomson et al. 1989)] also may arise from different locations of the neurons studied or different experimental conditions.
Afterpotentials are important in shaping neuron firing patterns (Fulton and Walton 1986; Gorelova and Reiner 1996; Llinas and Yarom 1981). A variety of afterpotentials in spinal cord neurons have been reported in different animal preparations (Jiang et al. 1995; King et al. 1988; Thomson et al. 1989; Yoshimura and Jessell 1989b), including fast AHP, slow AHP, fast ADP, and slow ADP. The mean duration and amplitude of afterpotentials in this study (duration: 93.4 ± 26.3 ms; amplitude: 8.1 ± 2.4 mV) were similar to those found in deep dorsal horn neurons (duration range, 14–268 ms; amplitude range, 2–12.6 mV) (Jiang et al. 1995). It has been demonstrated that the slow AHP is mediated by an apamin-sensitive Ca2+-dependent K+channel (for review, see Sah 1996). For neurons displaying a mono-phasic AHP in this study, the duration of the AHP was reduced and an ADP was unmasked after bath application of apamin. This indicates that multiple ion channels underlie the monophasic AHP in LSN neurons. It has been shown that inhibition of the slow AHP can either increase or decrease SFA based on the locations of neurons (Gorelova and Reiner 1996; Spanswick et al. 1995). It would be interesting to characterize the contribution of the slow AHP to cellular excitability in LSN neurons.
Membrane rectification has been shown to participate in shaping discharge patterns in spinal dorsal horn neurons (Yoshimura and Jessell 1989a). Fast and time-dependent inward rectification, outward rectification, and linear I-V relations all were observed in both superficial and deep dorsal horn neurons of the spinal cord (Jiang et al. 1995; Yoshimura and Jessell 1989a). It appears that the I-V relationship in LSN neurons is relatively simple. Most LSN neurons (7/11) exhibited a linear I-V relation; some (4/11) showed a time-dependent inward rectification.
PIR transiently increases neuron excitability by quickly depolarizing membrane potential after hyperpolarization, by which action potentials can be triggered depending on the magnitude of PIR. PIR has been observed throughout the CNS (Dekin 1993; Johnson and Getting 1991; Stewart and Wong 1993), including both the superficial and deep dorsal horn in spinal cord (Jiang et al. 1995; Lopez-Garcia and King 1994; Yoshimura and Jessell 1989a). It has been shown that PIR can be blocked partially by Cs+(Yoshimura and Jessell 1989a). In the present study, two types of PIR, low and high threshold, were observed. In low-threshold PIR neurons, burst-like firing also was observed at higher intensity stimulation. It was noted that bath application of low Ca2+solution switched the low-threshold PIR to high-threshold PIR, in which burst-like firing also disappeared. High-threshold PIR in the present study, on the other hand, is similar to that reported byYoshimura and Jessell (1989a). These results suggest that there are at least two membrane currents, K+ and Ca2+, participating in PIR.
The plateau potential is a prolonged membrane depolarization after the release from a depolarizing pulse, which can trigger action potentials. It has been demonstrated that the plateau potential participates in synaptic integration (Russo and Hounsgaard 1996) and contributes to the mechanism of wind-up (Russo and Hounsgaard 1994; Russo et al. 1997). In the current study, a plateau potential was observed in only 2/16 LSN neurons, which is less than reported for deep dorsal horn neurons (Morisset and Nagy 1996). However, because the plateau potential can be induced and enhanced by SP and/orcis-(±)-1-aminocyclopentane-1,3-dicarboxylic acid (Russo et al. 1997), and LSN neurons contain receptors for SP (Battaglia and Rustioni 1992; Ding et al. 1995; Li et al. 1997; Marshall et al. 1996), whether plateau potentials can be induced in LSN neurons awaits further investigation.
SFA has been reported in both superficial and deep dorsal horn neurons (Jiang et al. 1995; Thomson et al. 1989;Yoshimura and Jessell 1989a). In LSN neurons, SFA could be divided clearly into high- and low-adapting groups based on the ratio of the slopes between initial instantaneous firing and steady-state firing. A unique, biphasic SFA was observed in two LSN neurons in the present study. These two neurons appeared to have a combination of high SFA and delayed excitation (but adapted in firing frequency). In the majority of LSN neurons tested, the firing behavior evoked by synaptic input correlated well with SFA (low and high SFA). Similar to neurons in the dorsal horn (Lopez-Garcia and King 1994; Thomson et al. 1989), the low SFA neurons in the present study had prolonged discharges, whereas the high SFA neurons fired briefly to electrical stimulation. In the two biphasic firing neurons, synaptic input only produced a few discharges.
Effect of SP on cellular properties
SP is well documented to participate in nociceptive mechanisms (for reviews, see Henry 1982; Jessell 1981; also Moochhala and Sawynok 1984). It has been shown that SP has multiple effects on membrane currents (Adams et al. 1983; Dun and Minota 1981;Murase et al. 1989; Stanfield et al. 1985) and on intrinsic properties (Russo et al. 1997) of spinal dorsal horn neurons. These include depolarizing the membrane potential, changing membrane resistance, modulating plateau potential, and increasing neuronal excitability. In the present study, similar findings were made on eight LSN neurons after bath application of SP (10−7 M to 10−5 M). The SP effects lasted for 4.5–6.8 min depending on SP concentration. In all eight LSN neurons studied, SP depolarized membrane potential and increased neuron excitability by increasing initial instantaneous and steady-state firing frequencies. Membrane resistance was increased in three or decreased in two LSN neurons by SP, which also has been observed by other investigators (Krnjevic 1977;Murase and Randic 1984; Murase et al. 1982; Russo et al. 1997). The effect of SP was suggested by Murase et al. (1989) to be due to the balance between the voltage-sensitive Ca2+ current and the voltage-insensitive, Ca2+-sensitive cationic conductance.
In addition, we observed that SP could cause rhythmic changes in membrane potential (oscillation) in three LSN neurons. Oscillation of membrane potential has been reported in spinal cord neurons, the occurrence of which can be a consequence of intrinsic mechanisms (Hochman et al. 1994; Jiang et al. 1995), N-methyl-d-aspartate receptor activation (Hochman et al. 1994) or neural network mechanisms (Sandkühler and Eblen-Zajjur 1994).
Extracellular recordings of LSN neurons by Menetrey et al. (1980) documented that a small population of LSN neurons can be activated by peripheral mechanical stimulation and that the receptive fields for these neurons are large. We found in the present study that all LSN neurons studied responded to electrical stimulation of dorsal roots and that most of them (11/12) had polysynaptic input. Once activated, the LSN could contribute to nociceptive modulation due to its high concentration of nociception-related peptides and its synaptic connection with other brain structures.
The current study of rat LSN neurons in the L6–S1 spinal segments in an in vitro slice preparation has revealed a diversity of intrinsic cellular properties and polysynaptic communication in the LSN. The results also suggest that while intrinsic properties could control firing behavior of LSN neurons, which is supported by the correlation between synaptic evoked firing and SFA, synaptically released neuromodulators such as SP are able to modulate these intrinsic properties. This suggests that the LSN is able to participate in sensory processing, including nociception. However, as it has been shown that synaptic activities of LSN neurons are less responsive to natural stimulation (Menetrey et al. 1980) than electrical stimulation of the dorsal root, the role of LSN in sensory processing may be specifically important in certain conditions, such as hyperalgesia.
The authors thank S. Birely for secretarial assistance and M. Burcham for preparation of the figures.
This work was supported by National Institute of Neurologcial Disorders and Stroke Grant NS-19912.
Present address of M. C. Jiang: Dept. of Physiology, Northwestern University Medical School, 303 E. Chicago Ave., Chicago, IL 60611.
Address for reprint requests: G. F. Gebhart, Dept. of Pharmacology, Bowen Science Bldg., University of Iowa, Iowa City, IA 52242.
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 © 1999 The American Physiological Society