INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Itching (pruritus) is normally evoked by insect bites or contact with nettles, but chronic itch is associated with dermatitis and kidney or liver disease and often is only poorly controlled. There is a general lack of knowledge about neural itch mechanisms. We recently reported that neurons in both superficial and deep laminae of the lumbar dorsal horn responded to histamine, which is pruritic in humans (Simone et al. 1987, 1991), as well as to algesic chemical (capsaicin, mustard oil, nicotine) and physical stimuli (Carstens 1997; Jinks and Carstens 1999, 2000a). The present study sought to further investigate neural itch mechanisms.

The association of itch with the desire to scratch provides a rationale to use scratching behavior to assess itch in animals. Many pruritogens, including histamine, serotonin (5-HT), substance P, leukotriene B4, and platelet activating factor, elicit dose-related scratching (Andoh and Kuraishi 1998; Andoh et al. 1998; Berendsen and Broekkamp 1991; Kitagawa et al. 1997; Kuraishi et al. 1995; Woodward et al. 1995; Yamaguchi et al. 1999). Recent studies indicate that 5-HT is a much more potent inducer of scratching in mice compared with histamine (Kuraishi et al. 1995; Yamaguchi et al. 1999). Algesic chemicals such as capsaicin elicited little if any scratching in mice (Kuraishi et al. 1995) or rats (Frenk et al. 1988), suggesting that scratching behavior differentiates between pruritic versus algesic chemical stimuli.

The present study had two main aims. First, we investigated if intracutaneous (ic) 5-HT excites superficial dorsal horn neurons. Second, in correlative behavioral experiments we have investigated whether pruritic (5-HT, histamine) or algesic chemicals (capsaicin, Formalin) induce scratching behavior in the rat. We hypothesized that superficial dorsal horn neurons involved in signaling itch should be excited by pruritic but not algesic chemcals over a time course matching that of scratching behavior. An abstract of this work has appeared (Jinks and Carstens 2000b).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

All experiments were approved by the UC Davis Animal Use and Care Advisory Committee. Adult male Sprague-Dawley rats, weighing 450-580 g, were used. Rats had continuous access to food and water and were kept on a 12-h light-dark cycle with lights on at 8 AM.

Electrophysiological experiments

Experiments were conducted using 14 rats, 4 of which had been previously tested for scratching behavior (see following text). Rats were anesthetized with pentobarbital sodium (induction: 65 mg/kg ip, and maintenance: 10-20 mg · kg¹ · h¹ iv), and the lumbar spinal cord was exposed by laminectomy to record extracellular single-unit activity. Chemicals were delivered ic in 1-µl volumes to skin in the receptive field of the dorsal horn unit via a 30.5-gauge needle (Carstens 1997; Jinks and Carstens 2000a).

An ic histamine search strategy (Jinks and Carstens 2000a) was used to identify superficial dorsal horn units. Immediately after ic histamine, the recording microelectrode searched the superficial dorsal horn (down to 300 µm) for an actively firing unit. All presently recorded units were isolated <2 min after delivering the first histamine search stimulus (n = 17 of 24) or after a second search stimulus delivered at the same (n = 3) or different (n = 4) skin site. Once a unit was isolated, we waited for the firing rate to decline to a steady baseline (usually 10-12 min) and then reinjected histamine at the same site. Figure 2, A and D, shows examples of this procedure.

When unit activity once again returned to a steady baseline, responsiveness to low-threshold mechanical stimulation was tested, and receptive fields were provisionally mapped with a weak von Frey filament (1.2-g bending force). Unit responses to other stimuli were then tested in the following order: 0.9% NaCl ic, 5-HT ic (1%), heat delivered by a 1-cm² Peltier thermode (52°C, 5 s), noxious mechanical stimulation, ic capsaicin (0.03%), and topical application of mustard oil (10%; Fluka). This ensured that unit responses to the potential pruritogens, histamine and 5-HT, were not depressed by prior noxious stimuli. Only after applying 5-HT was the unit's mechanical responsiveness more thoroughly assessed. Units were classified as wide dynamic range (WDR) if they responded to innocuous levels of mechanical stimuli as well as to noxious mechanical (pressure, pinch) and heat stimuli. They were classified as nociceptive-specific (NS) if they did not respond to innocuous mechanical stimuli but did respond to noxious levels of mechanical pressure-pinch, and to noxious thermal stimuli. One unit did not respond to noxious mechanical stimuli and was classified as mechanically insensitive. A response was considered positive if the stimulus elicited a >200% increase in firing rate above the prestimulus level.

In six rats, only one unit was recorded. In eight rats, two units were recorded [simultaneously in 2; 1 on each side of the cord in 6; on the same side but at spatially separate (heel vs. toe) locations in 2]. In the two dual-unit recordings, action potential waveforms were discriminated using a template-matching procedure (Forster and Handwerker 1990).

At the conclusion of the experiment, the recording site was lesioned electrolytically, and spinal cords were removed and fixed in 10% Formalin. Later, 50-µM sections of spinal cord, counter-stained with neutral red, were examined by light microscopy to identify lesion sites.

Peak firing rates evoked by ic chemical or heat stimuli were calculated by constructing peristimulus time histograms (PSTHs, binwidth: 1 s) and selecting the maximum value. Latency to peak firing rate was measured from the time of injection. Tachyphylaxis was assessed by comparing the mean total number of impulses recorded during the 15-min period following the initial and a second ic injection of 5-HT (made 30 min later) using a paired t-test, with significance occurring at a P value <0.05. For each chemical tested, significant increases in firing rate were assessed at different time intervals postinjection by averaging the firing rate over this time interval for all units, and comparing mean firing rate with mean preinjection spontaneous firing rate using a two-factor ANOVA, followed by post hoc, paired t-tests, with significance occurring at a P value <0.05. Averaged PSTHs of responses of WDR and NS units are displayed separately (Fig. 1). However, since there were not obvious differences between these groups, they were pooled for purposes of data analysis.

View larger version (33K): [in this window] [in a new window]   Fig. 1. Averaged responses of superficial dorsal horn units to chemical and thermal stimuli. A: wide dynamic range (WDR) units. Averaged peristimulus time histograms (PSTHs) show, from left to right, responses to histamine (3%), serotonin (5-HT) (1%), capsaicin (0.03%), mustard oil (10%), noxious heat (52°C, 5 s), and vehicle (saline) control. Gaps in PSTHs due to sampling interruptions. Numbers of units are indicated in parentheses. Inset shows unit recording sites compiled on an L4 spinal cord section; numbers indicate dorsal horn laminae (from Paxinos and Watson 1998). B: nociceptive-specific (NS) units (format as in A).

Scratching behavior

These experiments quantified the number of bouts of hindlimb scratching directed at a site of irritant chemical injection into the nape of the neck. All rats were habituated to a loose cylindrical Plexiglas restrainer in four 30-min sessions the week prior to conducting experiments. At least 3 days prior to receiving injections, rats were briefly anesthetized with 3% halothane, and the fur over the rostral portion of their back was removed. The following irritant chemicals were tested: 5-HT (0.5, 1, or 2%; Sigma Chemical, St. Louis, MO; n = 8 rats tested at all 3 concentrations); Formalin (2 and 5%; n = 7 rats tested at both concentrations), capsaicin (0.01, 0.03, or 0.1%; Sigma; n = 5 rats tested at all 3 concentrations), and histamine (3 or 10%; Sigma; n = 4 rats tested at both concentrations), and were dissolved in 0.9% sodium chloride, except for capsaicin, which was dissolved in 40% ethanol at a 0.1% concentration, and then diluted with 0.9% NaCl as necessary. Vehicles (0.9% NaCl; n = 4 rats previously tested with 5-HT; and 40% ethanol, n = 5 rats previously tested with capsaicin) were also tested.

Rats were placed in the recording chamber for 30 min before receiving the chemical injection. Rats received consecutive injections at least 3 days apart in mixed order. All ic injections were given ic into the skin on the nape of the neck in a 10-µl volume. Microinjections were made by placing the rat into the restrainer for 5 min, and then gently grasping skin of the rostral back with rubber-coated forceps. A 30.5-gauge needle, connected to a 50-µl hamilton microsyringe via PE-50 tubing, was inserted into the superficial skin and left in place for 5 min, at which time the chemical was injected over a 20-s time period. The needle was left in place for 10 s postinjection, removed, and each rat was then immediately placed into a separate 30 cm × 30 cm × 40 cm opaque, open-top, plastic chamber, and videotaped from above for 40 min.

To test for tachyphylaxis, a separate group of rats (n = 7) received a microinjection of 2% 5-HT as described, videotaped for 40 min, and then immediately reinjected at the same site (marked by felt-tip pen) with the same dose of 5-HT and videotaped for another 40 min.

Four rats were injected and videotaped at a time. The experimenter left the room shortly after the injections were made and rats placed in the chamber. Videotapes were later analyzed for the number of hindlimb scratching bouts, which were collected in 2-min time bins over the 40-min videotaping period. A series of one or more scratching movements directed at the injection site was defined as a scratching bout, which ended when the rat either licked its hind paw or placed its hind paw back on the floor. The number of episodes of head or whole-body ("wet dog") shakes/40 min was also counted. The investigators scoring videotapes were blinded as to which treatment the rats had received.

Comparisons of numbers of scratching bouts and head/whole-body shakes between each treatment group and vehicle controls, as well as between different doses of a given chemical, were made using a one-factor ANOVA, followed by post hoc t-tests, with significance occurring at a P value <0.05.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Electrophysiology

UNIT SAMPLE. Eighteen WDR and six NS units (1 of which was mechanically insensitive) were recorded. The mean depth of recording was 142 ± 113 (SD) µM, and histologically recovered recording sites were located in superficial laminae of the dorsal horn (Fig. 1). For WDR units, the low-threshold mechanosensitive receptive fields spanned more than (n = 3) or less than (n = 15) 50% of the plantar surface, in the region of the toes (n = 11), mid-paw (n = 4), or heel (n = 3).

View larger version (43K): [in this window] [in a new window]   Fig. 2. Individual examples of superficial dorsal horn neuronal responses to irritants. A: PSTHs show, from left to right, a mechanically insensitive NS unit's responses to histamine, 5-HT, capsaicin, mustard oil, and noxious heat. Insets show, from left to right, the cutaneous injection site (arrow), spinal recording site, and example of action potential waveform. B-D: additional examples (format as in A).

VEHICLE CONTROL. The effect of vehicle injection (0.9% NaCl) was assessed in 15 units (12 WDR and 3 NS; Fig. 1, A and B, far-right PSTHs). For all units pooled, NaCl elicited a small but significant increase in firing rate during the first minute postinjection (P < 0.02, paired t-test), which was no longer significant at 1.5-2.5 min postinjection.

RESPONSE TO 5-HT AND TACHYPHYLAXIS. Immediately preceding the first 5-HT injection, the mean spontaneous firing rate was 2.9 ± 3.6 Hz, which was not significantly different from the prehistamine level (P = 0.59; paired t-test). 5-HT 1% ic elicited prolonged responses in 21/24 (88%) units (WDR: 15/18; NS: 6/6; Table 1) lasting at least 15 min, and in 10 units at least 25 min. 5-HT usually elicited a bi- or multiphasic response pattern; the averaged responses of WDR and NS units are shown in Fig. 1, A and B, respectively, and individual examples are shown in Fig. 2. The initial phasic component seen in all but one unit (Fig. 2A) had a peak firing rate (41 ± 24 impulses/s) that occurred at a mean latency of 14.2 ± 11.2 s, and then declined rapidly. This was followed by a gradual increase in firing rate that peaked at 8.2 ± 3.2 min. The mean peak firing rate of this second phase (28 ± 15 impulses/s) was significantly lower than the initial phasic response, but was significantly higher than the preinjection spontaneous firing rate (P < 0.0001, paired t-test). The mean firing rate was still significantly greater than mean spontaneous firing at 15-17 min post-5-HT (P < 0.001, paired t-test). Unit responses and scratching behavior elicited by ic injection of 5-HT 1% both exhibited prolonged time courses (compare Fig. 5A with Fig. 1A and B, 2nd PSTHs from left).

View larger version (26K): [in this window] [in a new window]   Fig. 3. Tachyphylaxis of neuronal responses to repeated intracutaneous (ic) 5-HT. A: averaged PSTHs of 8 units to 2 injections of 5-HT at 30-min interstimulus interval (ISI). B: individual unit example. PSTHs show, from left to right, responses of WDR unit to 2 repeated injections of 5-HT (at site indicated by arrow in figuring of hind paw receptive field), ic capsaicin, topical mustard oil, and noxious heat. Insets show, from left to right, receptive field, superficial recording site, and action potential waveform.

RESPONSES TO ALGESIC STIMULI. Units were additionally tested for their response to noxious heat, ic capsaicin, and topical mustard oil. Table 1 summarizes the responses. Of the NS units, 100% responded to capsaicin and mustard oil, and 67% to noxious heat. Of the WDR units, 88% responded to capsaicin, 56% to mustard oil, and 73% to heat. Examples of responses to different stimuli are shown for a mechanically insensitive NS unit in Fig. 2A, for WDR units in Figs. 2, B and D, and 3B, and for an NS unit in Fig. 2C.

Scratching and shaking behavior

5-HT. Vehicle (NaCl 0.9% or ethanol 40%) evoked almost no scratching (Fig. 4A). 5-HT reliably induced significant hindlimb scratching in each of eight rats (P < 0.01, 1-factor ANOVA) in a concentration-dependent manner (P < 0.01, paired t-test; Figs. 4A and 5A). 5-HT also elicited a significantly greater number of "wet dog" shakes above vehicle control (P < 0.02, 1-factor ANOVA; Fig. 4A). Indeed, there was a significant correlation (r = 0.72; P < 0.001) between the total number of scratching bouts and shakes recorded per animal per session under all treatment conditions (Fig. 4D). Mean 5-HT-evoked scratching behavior began after a 2- to 4-min latency and exhibited a bimodal distribution, with the first peak at 10-12 min and a later peak at 28-30 min (Fig. 5A).

View larger version (25K): [in this window] [in a new window]   Fig. 4. Scratching behavior. A: bar graph plotting averaged number of scratching bouts (filled bars) and shakes (open bars)/40 min for vehicle controls and different concentrations of 5-HT. Error bars: SE. * Significantly different (P < 0.01, ANOVA). B: graph as in A for capsaicin. C: graph as in A for Formalin and histamine. D: scatter plot of number of shakes vs. number of scratching bouts recorded for each animal in each treatment session. The correlation coefficient (r = 0.72) was significant (P < 0.001).

View larger version (18K): [in this window] [in a new window]   Fig. 5. 5-HT- and Formalin-induced scratching. A: bar graphs plot mean number of scratching bouts per 2-min collection period vs. time after ic 5-HT injection at time 0. Each graph plots data for the indicated concentration of 5-HT. B: bar graphs as in A for scratching induced by Formalin at the doses indicated.

View larger version (17K): [in this window] [in a new window]   Fig. 6. Tachyphylaxis in scratching behavior elicited by repeated injection of 5-HT. Graph plots mean number of scratching bouts (filled bars) and shakes (open bars) counted over 40-min collection periods following an initial injection of 2% 5-HT (Trial 1) and after a second identical 5-HT injection made at the same skin site 40 min after the 1st injection (Trial 2). Error bars: SE. **, * Significantly different from Trial 1 (P < 0.001 and 0.05, respectively, paired t-test).

CAPSAICIN. Capsaicin elicited low to moderate scratching and shaking behavior in some rats (44% of trials), but overall this was not significantly different from controls, nor was it dose-related (Fig. 4B), and was usually less vigorous and of shorter duration compared with 5-HT and Formalin-induced scratching.

FORMALIN. Formalin evoked low to moderate scratching that was consistent (7/7 rats) and significantly greater than vehicle (P < 0.02, 1-factor ANOVA), but was not concentration-related at the concentrations tested (Fig. 4C, left filled bars). Mean Formalin-evoked scratching behavior began at a latency of 8-10 min and peaked at 14-16 min (Fig. 5B). Formalin also evoked a significantly greater number of "wet dog" shakes compared with vehicle control (P < 0.02, 1-factor ANOVA; Fig. 4C, left open bars).

HISTAMINE. Histamine at the concentrations tested did not cause significant scratching or shaking behavior (Fig. 4C, right bars).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

We presently identified a population of histamine-sensitive neurons in the superficial dorsal horn that additionally responded to ic 5-HT over a time course that matched scratching behavior. Tachyphylaxis was observed in the behavioral and neuronal responses to repeated injection of 5-HT. These correlations suggest that the 5-HT-responsive neurons may be involved in signaling itch, as discussed further below.

Neurobehavioral correlates and methodological considerations

We presently used a search strategy designed to identify neurons responsive to ic histamine. This strategy could identify potential histamine-selective neurons, but has the drawback that the overall proportion of histamine-responsive WDR and NS neurons in the superficial dorsal horn cannot be specified. Using a mechanical search strategy, we found that a large majority of WDR neurons in the deep dorsal horn responded to ic histamine and other irritants (Carstens 1997; Jinks and Carstens 1999). Using antidromic stimulation to identify spinothalamic tract neurons in cat, Andrew and Craig (2001) recently identified a small subpopulation of mechanically insensitive, histamine-responsive lamina I neurons, some of which also responded to mustard oil. Thus while our results indicate that a majority of WDR and NS neurons in superficial dorsal horn respond nonselectively to different irritants, a subset may have more restricted chemoselectivity and thus be well-suited to selectively signal itch.

After initiating the electrophysiological studies, we observed that histamine did not elicit scratching or shaking behavior and thus may not be pruritic in Sprague-Dawley rats. Instead, 5-HT was found to elicit dose-related scratching as well as neuronal firing over a behaviorally relevant time course. An electrophysiological strategy using ic 5-HT might be more appropriate to search for itch-related dorsal horn neurons in rats. However, a drawback is the partial tachyphylaxis to repeated ic 5-HT (Fig. 3). While we showed previously that dorsal horn neuronal responses to histamine do not exhibit significant tachyphylaxis or sensitization (Carstens 1997; Jinks and Carstens 2000a), cross-tachyphylaxis induced by 5-HT might lower the incidence of neurons responsive to other subsequently delivered chemical stimuli. However, such an effect is not absolute, given the present finding that a majority of units responded to capsaicin and mustard oil following application of 5-HT (Table 1).

Recent psychophysical studies show that histamine elicits significantly stronger and more prolonged itch when injected into a skin "bleb" previously injected with local anesthetic (Atanassoff et al. 1999). It was hypothesized that histamine excites both "antipruritic" and itch-signaling primary afferents; the former engage a spinal process to inhibit itch transmission and are more strongly suppressed by the local anesthetic. In this scenario, it cannot be determined whether pruritogen-responsive neurons signal itch or are antipruritic (e.g., inhibitory interneurons). While it is advantageous to identify ascending projections (Andrew and Craig 2001), it is virtually impossible to determine whether a nonprojecting neuron functions as an excitatory or inhibitory interneuron. We presently made no attempt to identify the projections of the recorded neurons, which might represent a functionally heterogeneous population.

A methodological disparity is that different skin areas were injected in the behavioral (rostral back) and electrophysiological (hind paw) experiments. We chose to perform electrophysiological studies with lumbar neurons for comparison with prior studies (Carstens 1997; Jinks and Carstens 1998, 1999, 2000a), while assessing hindlimb scratching directed toward the back because this has been used in most prior behavioral studies (e.g., Kuraishi et al. 1995). It would be ideal to compare behavioral and neural responses to stimulation at the same skin site. However, it is problematic to record from neurons with input from the upper back because their receptive fields are in the surgical field. A fruitful approach would be to assess behavioral reactions to chemical injections into the hind paw. Hindpaw injection of Formalin (Dubuisson and Dennis 1977) or capsaicin (Gilchrist et al. 1996) elicits reactions (lifting, limb-guarding, licking, etc.) suggestive of pain and hyperalgesia. Hind paw injection of 5-HT induces hyperalgesia (Abbott et al. 1996; Sufka et al. 1992; Tokunaga et al. 1998) and was recently reported to elicit relatively more biting and less licking of the paw compared with Formalin (Hagiwara et al. 1999), prompting the authors to suggest that biting may reflect itch. More information on behavioral patterns elicited by pruritic and algesic chemical stimulation of the paw would be useful in developing animal models that distinguish itch from pain sensation. A drawback of our present approach was the absence of obvious signs of pain following ic injection of algogens into the back skin, although we noted informally that some rats vocalized at the time of injection of histamine and capsaicin.

Another methodological disparity was our use of different ic injection volumes in the behavioral (10 µl) and electrophysiological (1 µl) experiments. The 1-µl injection volume was used in the electrophysiological experiments to allow comparison with our prior studies (Carstens 1997; Jinks and Carstens 1999, 2000a) and to reduce edema and to facilitate clearance from the dermis, thereby allowing multiple ic injections at the same site. The 10-µl volume was used in the behavior study because pilot studies revealed that injection of 1 µl of 2% 5-HT in the back did not elicit scratching even though it excited dorsal horn neurons. This difference might reflect the requirement of spatial summation of chemonociceptive neurons to trigger scratching behavior.

ROLE OF SUPERFICIAL DORSAL HORN NEURONS IN ITCH AND PAIN. Several theories have been proposed for the neural encoding of itch and pain sensations. "Specificity" theory holds that itch and pain are signaled separately by distinct pathways. Variants of the specificity theory incorporate the fact that pain inhibits itch by occlusion (Handwerker 1992) or central inhibitory mechanisms (Atanassoff et al. 1999; Brull et al. 1999; McMahon and Koltzenburg 1992). The alternative "intensity" theory states that a common pathway signals both itch and pain, and that these qualities are distinguished based on some neural code such as firing frequency (McMahon and Koltzenburg 1992; von Frey 1922).

Scratching and shaking behavior

Of the irritants studied, only 5-HT elicited significant scratching and shaking in a dose-dependent manner. Formalin elicited significant scratching and shaking, but it was not dose related and was less than that evoked by 5-HT. Capsaicin elicited sporadic scratching that was insignificant at all doses tested. Histamine elicited almost no scratching or shaking at concentrations up to 10%. Our results are reasonably consistent with data from Kuraishi's group showing that 5-HT elicited dose-related scratching (Yamaguchi et al. 1999), while neither histamine, capsaicin, nor Formalin elicited scratching in ddY mice (Kuraishi et al. 1995; Yamaguchi et al. 1999), although histamine elicits scratching in ICR mice (Kitagawa et al. 1997) and hairless guinea pigs (Woodward et al. 1995). Conceivably, histamine may be algesic rather than pruritic in rats, compared with humans in whom histamine is pruritic but can elicit pain when injected into deeper skin layers and/or at higher concentrations (Keele and Armstrong 1964). These data indicate that 5-HT is a more effective pruritogen than histamine in some rodent strains.

Rodent mast cells contain 5-HT (Graziano 1988; Gustafsson 1980; Purcell et al. 1989), and unmyelinated nerve endings express 5-HT₂ receptors at the dermal-epidermal junction (Carlton and Coggeshall 1997) where itch is evoked in humans (Keele and Armstrong 1964; Shelley and Arthur 1957). 5-HT is less pruritic than histamine in humans (Fjellner and Hägermark 1979; Hägermark 1995; Weisshaar et al. 1997), but the reverse may hold in rodents. Another interpretation is that ic 5-HT evokes pain. Intraplantar 5-HT produces inflammation and hyperalgesia (Sufka et al. 1992; Tokunaga et al. 1998). Behavioral responses to the algogen Formalin are potentiated by 5-HT and blocked by 5-HT antagonists (Abbott et al. 1996, 1997; Giordano and Rogers 1989). In mice, 5-HT elicited scratching via a 5-HT₂ receptor (Yamaguchi et al. 1999). The relationship between scratching and 5-HT dose was bell-shaped, and it may be speculated that 5-HT is pruritic at lower doses but becomes painful at higher doses where scratching behavior decreases.

5-HT also elicited "wet dog" shakes reminiscent of opiate withdrawal and possibly related to grooming behavior. Shaking behavior correlated well with scratching (Fig. 4, A and D) and therefore may provide another useful parameter to assess chemogenic sensations.

We observed significant scratching with 10 µl ic Formalin. The Formalin pain test (Dubuisson and Dennis 1977) usually involves larger subcutaneous (sc) hindpaw injections (50 µl). Kuraishi et al. (1995) did not observe scratching in mice following sc Formalin (100 µl) in neck skin, which may have induced pain. In human studies noxious chemicals in superficial layers of skin often evoke itch, while they are more likely to be painful in the dermis or subdermis (Keele and Armstrong 1964; Shelley and Arthur 1957). Furthermore, itch may be suppressed by ic injection volumes >50 µl (Arthur and Shelley 1959; Keele and Armstrong 1964). The Formalin-induced scratching observed presently usually began 8-10 min postinjection, coinciding with the interphase and ending at the beginning of the second phase of activity elicited by intraplantar Formalin in C fibers (McCall et al. 1996) and dorsal horn neurons (Chapman and Dickenson 1995; Dickenson and Sullivan 1987). That scratching occurred at the nadir of Formalin-induced nociception indirectly supports the idea that scratching reflects itch rather than pain.

Capsaicin elicits burning pain and hyperalgesia in humans (LaMotte et al. 1991; Schmelz et al. 2000) by binding to C fiber VR-1 receptors (Caterina et al. 1997, 2000). Topical capsaicin can also elicit itch (Green 1990; Green and Shaffer 1993). Presently, capsaicin elicited scratching in some rats, possibly reflecting itch dependent on the exact ic location of the injection needle. Future behavioral studies assessing itch in animals should take into account factors including concentration, volume, skin site, and method of application, which may all influence the degree of resulting itch or pain.