INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Increasing recognition of the importance of pain in infancy and childhood has focused attention on the basic neurobiology of developing pain pathways. Cutaneous reflex function in the newborn rat and human is exaggerated compared with the adult (Andrews and Fitzgerald 1994; Fitzgerald and Gibson 1984) with lower mechanical and thermal thresholds and more synchronized and long-lasting muscle contractions (Falcon et al. 1996; Fitzgerald et al. 1988c; Hu et al. 1997; Marsh et al. 1999a). Repeated low-intensity skin stimulation leads to sensitization of the reflex with lower thresholds and generalized movements of all limbs (Andrews and Fitzgerald 1994; Fitzgerald et al. 1988b,c). The developmental regulation of the behavioral response to persistent noxious stimulation is less clear. In rat pups the drop in mechanical threshold following carageenan inflammation (Marsh et al. 1999b) and the enhanced nociceptive response following mustard oil application (Jiang and Gebhart 1998) is smaller in amplitude than in older animals. In contrast the response to formalin has a 10-fold higher sensitivity in neonatal rats than weanlings (Teng and Abbott 1998). These previous investigations of inflammatory pain in neonatal rat pups have relied on reflex measurements which require stable motor responses. Here we examine the postnatal development of sensory responses to carageenan inflammation directly, using "in vivo" electrophysiological recordings of dorsal horn neurons in young rat pups.

Carageenan inflammation is a useful model of inflammatory pain in adult rats resulting in behavioral hypersensitivity to both thermal and mechanical stimuli which is prominent 2-4 h following intraplantar injection (Hargreaves et al.1988; Meller et al. 1994). Electrophysiological studies of dorsal horn cells show that inflammation in adults generated by a variety of agents causes an increase in cutaneous receptive field size (Hylden et al. 1989; Ren et al. 1992; Woolf and King 1990), spontaneous activity (Hylden et al. 1989; Pertovaara et al. 1998), and afterdischarge (Neumann et al. 1996; Woolf and King 1990). Dubner (1991) proposed that expanded receptive fields will lead to a greater number of neurons activated by a given stimulus. This may result in a given stimulus being perceived as more painful and/or an increased chance of evoking a reflex response, thereby lowering the threshold. Spontaneous activity has been proposed to correlate with clinical observations of spontaneous flashes of pain (Menetrey and Besson 1982) and an increase in afterdischarge may well correlate with a given stimulus producing more prolonged pain.

The aim here was to examine the response of the neonatal nervous system to such an inflammatory insult. Since the developmental regulation of transmitter/receptor systems and maturation of connectivity result in a background sensory processing that differs from the adult (Alvares and Fitzgerald 1999; Baba et al. 2000; Bardoni et al. 1998; Bennett et al. 1996; Fitzgerald and Jennings 1999; Nakatsuka et al. 2000), the response to inflammation is also likely to differ from adults.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Sprague-Dawley rat pups of both sexes at postnatal day (P) 3, 10, and 21 were anesthetized with 2-2.5 g kg¹ urethan i.p. (Sigma). This dose of urethan causes anesthesia for 8 h, and our experiments never lasted more than 6 h. Animals were pretreated with an injection of 1% lambda carageenan (1 µl/g body wt, Sigma) into the plantar surface of the hindpaw (under brief halothane anesthesia) and electrophysiological recordings were made 2- to 5-h postinjection. The injection volume was adjusted according to body weight to control for changes in plantar surface area of the hindpaw with age. Control animals received no treatment prior to the induction of urethan anesthesia. The trachea was cannulated and intermittent positive pressure ventilation was achieved using a T-piece system in conjunction with a small animal lung ventilator pump (Harvard Apparatus). The pups were set up in a small animal Kopf stereotaxic frame, with the head and pelvis firmly held and a small clamp at L1 to stabilize the cord. The lumbar cord was exposed by laminectomy; the dura mater (and arachnoid) removed, and the surface of the cord bathed in mineral oil. When the pup was deeply anesthetized, as shown by areflexia, it was paralyzed with 0.1 ml Flaxedil (May and Baker). Finally the hind limbs were supported with a suture under the Achilles tendon. The pup was kept warm with a heated blanket, and the heart rate was monitored throughout the experiment and maintained within the range of 350-500 beats min¹. Animals were killed with an overdose of Lethobarb (pentobarbitone sodium BP) at the end of the experiment.

Extracellular recordings were made from cells in the dorsal horn of the L4-L5 lumbar cord using glass-coated tungsten microelectrodes. Recordings were made throughout the dorsal horn and the depth noted from the surface of the cord. Single cells with receptive fields on the hindpaw were mapped using natural mechanical stimuli, i.e., light brush, touch, and pinch. The receptive fields of the cells used in this study were located on the plantar surface of the hindpaw and were all cutaneous mechanoreceptive fields of the slowly adapting or rapidly adapting type. Low threshold (LT, responding to brush only), wide dynamic range (WDR, responding to brush and pinch), and high-threshold (HT, responding to pinch only) cells were assessed at all postnatal ages. Age/treatment did not significantly alter the proportion of cells that were LT, WDR, or HT. Receptive field size was assessed with innocuous mechanical stimulation except in cells which responded to pinch stimulation only, where noxious mechanical stimulation was applied. The receptive field size was calculated as a percentage of the total plantar hindpaw area. This method has the advantage that the size of the hindpaw (which changes considerably over this period) does not have to be taken into account. Mechanical thresholds were determined by applying von Frey hairs to the center of the receptive field. The series of von Frey hairs used in the present study were 0.03, 0.048, 0.07, 0.09. 0.44, 0.8, 1.13, 1.52, 3.12, 3.8, 4.72, 7.48, 9.4, and 13.36 g. Each von Frey hair was applied three times and the mechanical threshold was defined as the lowest von Frey hair required to evoke spike activity in all three trials. The magnitude of response to threshold and suprathreshold (3 von Frey hairs above threshold) mechanical stimulation (1 s) was also assessed. (Response to threshold stimulation can be readily quantified because the von Frey hair scale is not continuous.) Response amplitude was recorded over the 12-s period following stimulus application. This time window was chosen to ensure that more prolonged discharges that may occur at different ages/treatments were included. As a result any spontaneous activity present may also be included in this measure. This is therefore an overall measure of the response to mechanical stimulation, in the presence or absence of ongoing activity, and is not specific to the period of stimulation (1 s) itself. Electrical stimulation of the skin was applied through subcutaneous pin electrodes in the center of the receptive field at stimulus intensities of 100 µA to 10 mA, 100-500 µs. The A fiber threshold was defined as the minimum electrical stimulus needed to produce a short-latency response from the dorsal horn cell. The latency of response to A fiber stimulation was defined as the latency to the first spike after a single stimulus at twice the threshold level. Evoked response amplitude was measured in the 200-ms period following stimulation at P3 and the 70-ms period following stimulation at P10/21. Evoked response amplitude was measured over a longer period in the youngest animals because latencies were long and variable at this age. All cells were also tested at higher stimulus intensities to test for a longer latency C fiber input. Repetitive stimuli were applied with a train of 16 stimuli at a frequency of 0.5 Hz at either ×2 the A fiber threshold or ×3 the C fiber threshold. A fiber sensitization (increased A fiber afterdischarge) was assessed by measuring spike activity during a 200- to 2000-ms window between stimuli in a train of 16 stimuli (0.5 Hz) at twice the A fiber threshold. Background or spontaneous activity was measured for 1 min prior to electrical stimulation. Spike recordings were captured and analyzed by computer using a Maclab interface and software.

Statistical analysis was carried out using two-way analysis of variance (ANOVA) followed by Tukey post tests. Data were transformed (log, inverse, or square root) to meet the assumptions of the two-way ANOVA (normality of errors and homogeneity of variance). This analysis includes a test of interaction, i.e., a dependence of carageenan effect on postnatal age. If there appeared to be such a dependence, the effect of carageenan was assessed separately by Tukey post tests at each postnatal age. The three postnatal age groups were similarly compared for each treatment group separately. If there was no interaction, the two treatment groups were compared by a single post test. Similarly the three age groups were compared by Tukey post tests which pool the two treatment groups.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

One hundred and twenty-six single-unit recordings were made from the lumbar dorsal horn of the following experimental groups: P3: control (n = 19), carageenan inflamed (n = 18); P10: control (n = 19), carageenan inflamed (n = 33); P21: control (n = 17), carageenan inflamed (n = 20). Recording tracks were made in the medial third of the dorsal horn in the tibial terminal zone of L4-L5. Cells were recorded from both superficial and deep laminae (Fig. 1).

View larger version (23K): [in this window] [in a new window]   Fig. 1. Recording depths. Scatter plots showing depth (µm) of cells from the surface of the cord from control () or carageenan () treated pups. Superficial/deep boundary values were estimated from Nissl stained lumbar cord sections.

Effect of inflammation on neonatal dorsal horn cell receptive field size

Figure 2A shows the effect of carageenan on receptive field size at three different postnatal ages. In normal animals, mean receptive field size decreases with age as previously reported (P < 0.001) (Fitzgerald 1985; Fitzgerald and Jennings 1999). In addition, spread of sizes decreases so that receptive field size ranges from 6-100% at P3, 6-28% at P10, and 1-19% at P21. Hence in normal animals, small receptive field sizes are present in all age groups but larger receptive field sizes are restricted to the youngest age group. Following carageenan inflammation, however, the pattern changes. Figure 2 shows that at P21 average receptive field sizes are increased 3.2-fold by carageenan (P < 0.001) and spread of receptive field sizes increases. Receptive field sizes range from 5-100% in carageenan-treated animals compared with 1-19% in control animals. It is of interest that the spread of receptive field sizes following inflammation at P21 is strikingly similar to the spread at P3 in control animals. In P10 animals carageenan also results in an increase in the spread of receptive field sizes from 6-28% in control animals to 4-53% in the carageenan-treated group. In terms of average receptive field size there appears to be a 15% increase in the carageenan group but the means of the transformed data were not significantly different. In the P3 group carageenan had no significant effect on receptive field size. Receptive field size ranges from small to large in both the control (6-100%) and the carageenan (5-100%) groups (Fig. 2).

View larger version (15K): [in this window] [in a new window]   Fig. 2. Receptive field size distribution. A: scatter plots showing dorsal horn neuron receptive field sizes (as a percentage of the area of the plantar surface of the hindpaw) from control () or carageenan () treated pups at the postnatal ages investigated. B: diagrammatic representation of receptive fields sizes at P3 and P21 in control and inflamed animals

Effect of inflammation on neonatal dorsal horn cell responses to mechanical stimulation

Figure 3 shows the effects of carageenan on neonatal dorsal horn cell mechanical thresholds. These thresholds normally increase with postnatal age (P < 0.001) but they were not altered by carageenan inflammation (Fig. 3A). Response amplitude at threshold was also unaltered (data not shown). Carageenan did, however, produce a small increase (41%) in the magnitude of response to suprathreshold stimulation when data were pooled across all ages (P = 0.02, Fig. 3B). This effect is not dependent on postnatal age.

View larger version (12K): [in this window] [in a new window]   Fig. 3. Mechanical stimulation. A: scatter plots showing the mechanical thresholds (as determined using von Frey hairs) of dorsal horn neurons from control () or carageenan () treated pups at the postnatal ages investigated. B: scatter plots showing the response of dorsal horn neurons to suprathreshold mechanical stimulation from control () or carageenan () treated pups at the postnatal ages investigated.

Effect of inflammation on neonatal dorsal horn cell A fiber afferent input

Figure 4 shows the effects of carageenan inflammation on the electrically evoked A fiber responses of neonatal dorsal horn cells. Latencies of response to A fiber stimulation were long and varied widely in the youngest animals but decreased, in both mean and range, with postnatal age (P < 0.001, Fig. 4A). Carageenan inflammation did not significantly alter the latency of response to A fiber stimulation at any age. Figure 4B demonstrates that the evoked response to A fiber stimulation, which increases with postnatal age (P < 0.001), is also unaffected by carageenan inflammation.

View larger version (10K): [in this window] [in a new window]   Fig. 4. Electrical stimulation. A: scatter plots showing the latency of A fiber evoked responses of dorsal horn neurons from control () or carageenan () treated pups at the postnatal ages investigated. B: scatter plots showing the evoked response of dorsal horn neurons to A fiber stimulation from control () or carageenan () treated pups at the postnatal ages investigated.

Effect of inflammation on sensitization to repetitive A fiber stimulation and spontaneous activity

Figure 5A shows that A fiber sensitization was significantly increased over threefold by carageenan treatment (P < 0.001). Although A fiber induced sensitization itself significantly declined with postnatal age (P = 0.04), this effect of carageenan was not significantly different across the ages tested. However the range of average spike activity generated increased from 0-5 spikes (control) to 0-6 spikes (carageenan) at P3, 0-4 spikes (control) to 0-6 spikes (carageenan) at P10, and from 0-2 spikes (control) to 0-16 spikes at P21 (carageenan).

View larger version (11K): [in this window] [in a new window]   Fig. 5. Afterdischarge/spontaneous activity. A: scatter plots showing the mean interstimulus discharge generated during a train of 16 electrical stimuli at ×2 the A fiber threshold. Recordings were made from dorsal horn neurons in control () or carageenan () treated pups at the postnatal ages investigated. B: scatter plots of background activity measured in a 1-min window prior to electrical stimulation. Recordings were made from dorsal horn neurons in control () or carageenan () treated pups at the postnatal ages investigated.

Carageenan significantly increased spontaneous activity by 2.8-fold (P = 0.02, Fig. 5B). The increase was not dependent on age.

Effect of inflammation on neonatal dorsal horn cell C fiber afferent input

Inflammation did not significantly alter the number of cells displaying C fiber responses or the percentage of those cells showing wind up (data not shown).

No long-latency responses were evoked in response to C fiber stimulation at P3 confirming earlier reports (Fitzgerald 1988a; Fitzgerald and Jennings 1999; Nakatsuka et al. 2000). At P10 and P21, 26 and 32%, respectively, of dorsal horn cells had C fiber responses. The number of evoked C fiber spikes ranged from 1-3 spikes in control and 1-2 spikes in carageenan-treated animals at P10. At P21 the number of evoked spikes increased from 1-3 spikes in the control group to 1-7 in the inflamed group.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The present study demonstrates that the characteristic electrophysiological changes that occur in the dorsal horn of the adult spinal cord following an inflammatory insult alter over the postnatal period. Inflammation did not alter receptive field size at P3 or P10. However at P21 receptive field size was significantly increased by 3.2-fold. The mechanical thresholds of individual dorsal horn sensory neurons, which increased postnatally, were unaffected by inflammation. However inflammation resulted in an increased response to suprathreshold stimuli, increased afterdischarge, and an increased spontaneous activity at all ages. Although increased sensitization to repeated stimulation was present across all the ages, the increase in range of spike activity evoked at P21 following inflammation suggests that this effect may mature postnatally.

Normal postnatal regulation of dorsal horn cell properties

The effects of inflammation on neonatal dorsal horn cells have to be viewed in the context of the normal postnatal development of sensory processing. Dorsal horn properties differ from those in adults, undergoing considerable postnatal regulation, and this was confirmed in this study. Receptive fields are larger and cells are sensitized by repeated A fiber stimulation (Fitzgerald 1985; Fitzgerald and Jennings 1999; Jennings and Fitzgerald 1998). This has been attributed to a lack of inhibitory control in the neonate with spinal and descending spinal controls maturing postnatally (Beggs et al. 1999). Neonatal capsaicin treatment, which destroys afferent C fibers, results in mature animals retaining large receptive fields in the dorsal horn and cortex (Cervero and Plenderleith 1985), perhaps the result of inadequate interneuron function or descending inhibition (Fitzgerald and Koltzenburg 1986). Moreover the neonatal dorsal horn has few nociceptive neurons and is dominated by low-threshold inputs (Fitzgerald and Jennings 1999), supported by both terminal labeling (Fitzgerald et al. 1994; Mirnics and Koerber 1995) and electrophysiological recording (Nakatsuka et al. 2000). While responses to A fiber input are enhanced in immature spinal cord, long-latency C fiber evoked spike responses are not evoked in dorsal horn cells before the end of the second postnatal week (Fitzgerald 1988a; Jennings and Fitzgerald 1998; Nakatsuka et al. 2000; Park et al. 1999).

Inflammatory hypersensitivity in the neonate and adult

In the adult, inflammation results in the behavioral phenomena of hypersensitivity involving an increased pain response to noxious stimulation and a fall in sensory thresholds. In terms of dorsal horn cell electrophysiology, an increase in afterdischarge and an increase in receptive field size may contribute to hypersensitivity. In addition, the recruitment of previously ineffective low-threshold innocuous A fiber input to nociceptive specific neurons could contribute (Woolf and King 1990). Finally spontaneous activity has been proposed to correlate with clinical observations of spontaneous flashes of pain (Menetrey and Besson 1982).

The present data show that, despite their differing baseline properties, inflammation is clearly capable of increasing the excitability of dorsal horn sensory neurons in the neonate but the effects are developmentally regulated. Increased response to suprathreshold stimuli, increased afterdischarge, and an increase in spontaneous activity are all present from P3 onward, showing that neonatal sensory neurons are able to display properties consistent with behavioral hypersensitivity and possibly background pain. However the increase in afterdischarge and the expansion of receptive field size at P21 suggest that the ability to generate a hypersensitive state does increase postnatally as suggested from earlier reflex studies (Jiang and Gebhart 1998; Marsh et al. 1999b).

Mechanisms of inflammatory hypersensitivity in neonates and adults

One proposed mechanism of hypersensitivity involves the expansion of receptive fields that characteristically occurs following inflammation due to the recruitment of previously ineffective inputs (Woolf and King 1990). Expansion of receptive fields will result in a greater number of dorsal horn neurons activated by a given stimulus. This will result in reduced spatial discrimination, increased input, but also reduced thresholds of tertiary cells, such as motor neurons (see Fig. 6). Expanded receptive fields can therefore contribute to lower behavioral thresholds, in that previously ineffective/subthreshold input at the motor neuron or thalamic level may now be capable of evoking the withdrawal reflex or activating nociceptive neurons.

View larger version (23K): [in this window] [in a new window]   Fig. 6. Schematic representation of how large receptive fields can result in lower thresholds. A given stimulus may well activate neurons in the spinal cord dorsal horn but the resultant input to motor neurons may be subthreshold in terms of evoking the withdrawal reflex. However in the adult inflammation-induced expansion of receptive fields will result in a greater number of dorsal horn neurons being activated by the same stimulus. Therefore the input to motor neurons may now be capable of evoking the withdrawal reflex. This could also explain why neonates, which have large receptive fields, have lower thresholds.

Inflammation-induced expansion of receptive field size only occurred in the P21 group in the present study. Interestingly in the P21 carageenan group the distribution of receptive field size is remarkably similar to the receptive field size distribution in either the P3 control or the P3 carageenan-treated groups (see Fig. 2). It could be argued that inflammation leads to a recapitulation of the receptive field size distribution in the young neonate. In the youngest neonates input to the dorsal horn appears to be at a maximum, as reflected by large receptive fields, possibly as a result of the lack of inhibitory control at this early stage. Therefore despite the fact that inflammation can excite neonatal dorsal horn cells, the lack of inhibition means there is effectively no "ineffective input" to be recruited until P21 when receptive field size has been restricted.

As receptive fields are larger in the naive neonate and decrease postnatally, this may also in part explain the lower mechanical thresholds of the flexion reflex observed in the neonate, which increase postnatally. In addition the thresholds of individual dorsal horn sensory neurons in the neonate are lower and also increase postnatally. The mechanical thresholds of individual sensory neurons in the neonate cover a range of 0.03-4.72 g in the present study, which is considerably lower than mechanical behavioral thresholds (C. Torsney and B. Glickstein, unpublished observations), supporting the concept of summation at the motor side of the reflex.

Carageenan inflammation did not reduce the mechanical thresholds of individual dorsal horn sensory neurons at any postnatal age in the present study. This is consistent with results in adults (Hylden et al. 1989). Dorsal horn cell mechanical thresholds are evidently developmentally regulated but are not reduced or regulated by carageenan inflammation. This further supports the role of receptive field size in determining behavioral sensory thresholds. Differences in neonatal and adult hypersensitivity may result from developmental differences in the transmitter systems involved. In the adult inflammation induced hypersensitivity involves glutamate acting on N-methyl-D-aspartate (NMDA) receptors, brain-derived neurotrophic factor (BDNF) on TrkB receptors, and substance P acting on neurokinin receptors (for review see Woolf and Costigan 1999). Neuropeptide levels (Marti et al. 1987; Reynolds and Fitzgerald 1992), substance P receptors (Charlton and Helke 1986; Kar and Quirion 1995), TrkB receptor (Ernfors et al. 1993), and NMDA receptor distribution and subunit expression are all developmentally regulated (Gonzalez et al. 1993; Watanabe et al. 1994). These postnatal alterations in transmitter systems may underlie the postnatal maturation of the hypersensitive response. Additionally the novel gene expression induced by inflammation in the neonate (Beland and Fitzgerald 2001) is different from the adult (Neumann et al. 1996) in that C fibers are affected to the same extent as A fibers.

In conclusion, these data suggest that inflammation can excite neonatal dorsal horn cells at the postnatal ages examined (P3, P10, and P21). This is exemplified by increased afterdischarge, increased responses to suprathreshold stimulation, and increased spontaneous activity. However expansion of receptive field size is not observed until at least the second postnatal week. Therefore the neurophysiological changes that underlie inflammatory hypersensitivity are developmentally regulated.