Oral irritant properties of menthol: sensitizing and desensitizing effects of repeated application and cross-desensitization to nicotine

Physiology & Behavior 73 (2001) 25 ± 36

Oral irritant properties of menthol Sensitizing and desensitizing effects of repeated application and cross-desensitization to nicotine J.-M. Dessiriera,b, M. O'Mahonyb, E. Carstensa,* a

Section of Neurobiology, Physiology and Behavior, University of California-Davis, One Shields Avenue, Davis, CA 95616, USA b Department of Food Science and Technology, University of California-Davis, Davis, CA, USA Received 29 June 2000; received in revised form 4 January 2001; accepted 10 January 2001

Abstract The irritant properties of menthol and its interactions with nicotine were investigated psychophysically in human subjects. In the first experiment, 0.3% L-menthol was applied successively to one side of the tongue 10 times at a 1-min interval (30-s interstimulus interval, ISI), and subjects rated the intensity of the perceived irritation. The intensity of irritation progressively decreased across trials, consistent with desensitization. To test for cross-desensitization of nicotine-evoked irritation by menthol, nicotine (0.6%) was applied to both sides of the tongue simultaneously, 5 min after the conclusion of menthol application. Using both a two-alternative forced choice (2-AFC) paradigm, and also by obtaining independent ratings of the irritant intensity on each side of the tongue, it was found that nicotine-evoked irritation was significantly weaker on the menthol-pretreated side. To control for a possible confounding effect of cooling, nicotine was applied bilaterally only after the cooling sensation of menthol had subsided. Nicotine-induced irritation was still significantly weaker on the menthol-pretreated side, consistent with cross-desensitization of nicotine-evoked irritation by menthol. In a final experiment, menthol was repeatedly applied to one side of the tongue at a shorter (20 s) interval (5-s ISI), and elicited a rapid increase in irritant sensation over the initial trials, consistent with sensitization, followed in subsequent trials by a progressive reduction in irritation (desensitization). After a 5-min rest period, selfdesensitization was confirmed. Repeated application of menthol at the same short ISI was then resumed, and resulted in a significant mean increase in irritant intensity consistent with stimulus-induced recovery (SIR). D 2001 Elsevier Science Inc. All rights reserved. Keywords: Menthol; Nicotine; Oral irritation; Trigeminal; Desensitization; Sensitization; Stimulus-induced recovery; Psychophysics; 2-AFC

1. Introduction Menthol is a common additive in chewing gum, mouthwashes and other products because of the cooling and ``fresh'' sensations it elicits. Menthol is also a constituent of some topical analgesic preparations and may reduce pain via cooling or counter-irritant actions, although this has not been extensively documented scientifically (see Ref. [1]). Menthol and nicotine are commonly found together in tobacco products such as chewing tobacco or mentholated cigarettes that are popular, particularly among African American smokers [2]. Surprisingly, very little is known about the sensory interactions between nicotine and menthol. Nicotine's main sensory effect is a burning irritant or even frankly painful * Corresponding author. Tel.: +1-530-752-6640; fax: +1-530-752-5582. E-mail address: [email protected] (E. Carstens).

sensation when applied to oral [3± 6], nasal [7 ± 9] or ocular [10] mucosae or skin [11]. This is thought to be mediated via binding of nicotine to neuronal nicotinic acetylcholine receptors (nAChRs, see Refs. [12,13] for recent reviews) expressed in C- and Ad-nociceptors innervating these tissues [14 ± 17]. The nociceptor afferent fibers project centrally to excite neurons in the spinal cord [18] or trigeminal complex [19 ± 21]. Menthol is well known to activate cold sensitive fibers [17,22 ±24], to induce the sensation of cooling [25,26] and to modulate sensations of warmth and cold [27 ± 29]. Menthol in higher concentrations also induces a sensation of irritation in the oral cavity [25,26,30], but it is currently unknown if menthol at such high concentrations activates nociceptors. Sekizawa et al. [24] reported that menthol increased activity in whole-nerve recordings from the ethmoid nerve, and that the menthol-evoked response was reduced following capsaicin treatment. These authors also described single fibers that gave

0031-9384/01/$ ± see front matter D 2001 Elsevier Science Inc. All rights reserved. PII: S 0 0 3 1 - 9 3 8 4 ( 0 1 ) 0 0 4 3 1 - 0

26

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

excitatory responses to menthol as well as capsaicin. Petersen et al. [31] also reported that menthol activated capsaicinsensitive dorsal root ganglion (DRG) neurons. Finally, Farley and Silver [32] showed that menthol cross-desensitized ethmoid nerve responses to amyl acetate, D- and L-carvone. This cumulative evidence suggests that menthol is capable of activating nociceptive pathways. Additional support for a nociceptive action of menthol is that it, like most other irritants studied to date, induces selfdesensitization, that is, a prolonged ( > 5 min) reduction in subsequent menthol-evoked irritation. Finally, it was recently reported that menthol induced a short-lasting cross-desensitization of capsaicin-evoked irritation [1], suggesting that menthol shares common afferent pathways with capsaicin. Menthol thus joins the list of other irritant chemicals including capsaicin (e.g., Refs. [3,33 ± 35], piperine [4,36] and citric acid [37], all of which induce crossdesensitization, that is, a reduction in irritation evoked by compounds other than itself. The former study [1] is one of very few documenting the utility of menthol as an analgesic agent in over-the-counter preparations (see Refs. [1,38]). A reduction in nicotine-evoked irritation by a crossdesensitizing or cooling action of menthol could thus be one factor contributing to the popularity of mentholated tobacco products. We therefore investigated whether menthol could reduce nicotine-evoked irritation on the tongue (cross-desensitization), using methods similar to those employed in our earlier studies [3,4]. In a second experiment, we sought to differentiate between possible irritant and cooling actions of menthol in inducing crossdesensitization of nicotine irritation. In a final experiment, the ability of menthol to produce sensitization and stimulusinduced recovery (SIR) was investigated (see Section 4). 2. Experiment I 2.1. Materials and methods 2.1.1. Subjects Twenty-two healthy individuals (6 males, 16 females, aged 18 ±43 years), who were students and staff at the University of California at Davis, volunteered to participate in the study. All refrained from eating or drinking at least 1 h prior to each experimental session. In addition, subjects were required not to consume or use any mentholated products for at least 3 h prior to any session. Finally, subjects were also asked not to eat spicy food for 2 days prior to testing. To verify that subjects had complied with these restrictions, they were asked to fill out a questionnaire at the beginning of each session. This study was approved by the Human Subjects Review Committee of the University of California at Davis. 2.1.2. Chemical stimuli and application procedure All chemicals were purchased from Sigma (St. Louis, MO). A 6% (w/v) L-menthol stock solution was prepared by

dissolving menthol crystals in 95% ethanol and polyoxyethylene-sorbitan monooleate (Tween 80). A 0.3% (19.2 mM) solution was then made up by diluting the stock solution with dH2O. The final solution contained 4% ethanol and 1% Tween. To avoid recrystallization of menthol, the dilution was made daily. Fifteen microliters of this solution were pipetted onto small (78.5 mm2) and 40 ml onto large (176.7 mm2) filter paper disks (Whatmann, Maidstone, UK). Nicotine was dissolved in dH2O to a final concentration of 0.6% (37 mM) and 15 ml were pipetted onto the small filter papers. A suction device (Saliva Ejector, 6 in. clear, Sullivan Dental Products, Sacramento, CA) was used throughout the experiment to remove saliva and avoid the spread of the chemical. Subjects were instructed to use the suction device freely, except for the 15-s period prior to giving intensity ratings or two-alternative forced choice (2-AFC) responses (see below). This was especially enforced during the monitoring of the cooling sensation and nicotine irritation, since air drawn into the mouth might itself have a cooling action that could affect nicotine irritation or potentiate any cooling effect of menthol alone. 2.1.3. Experimental design The subjects participated in a single experimental session that consisted of sequential application of menthol to one side of the tongue, followed, after a 5-min period, by bilateral stimulation with nicotine to check for cross-desensitization. Thus, the larger size filter paper disks containing 40 ml of 0.3% menthol were applied with forceps onto one side of the anterior dorsal surface of the tongue at 1-min intervals as in previous studies [3,4]. A filter paper was applied for 30 s and then removed. Thirty seconds later, a fresh filter paper containing the same menthol stimulus was applied in the same manner at the same location (i.e., 30-s interstimulus interval, ISI); this was repeated 10 times. For each application, subjects rated the perceived intensity of irritation 30 s after stimulus onset. Ratings were made using a bipolar category scale used in a previous study [4] (see below). Subjects were asked to focus on, and rate, only the irritant (e.g., biting/pricking/stinging/burning) quality and to ignore gustatory (e.g., bitter) or thermal (e.g., cooling) sensory qualities. The side of the tongue receiving menthol was counterbalanced across subjects. After the 10 unilateral sequential stimuli, a 5-min period ensued during which the cooling sensation elicited by the repeated menthol stimulations was monitored. Subjects were asked to compare the intensity of the cooling sensation on the two sides of the tongue using a 2-AFC test and intensity ratings as described below. These comparisons were made four times, once every minute. After this 5min period, two small-sized filter papers containing 0.6% nicotine were applied simultaneously with two forceps at corresponding sites on the side that previously received menthol, as well as on the opposite side. Subjects then performed the 2-AFC test and intensity ratings again, this

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

time comparing the intensity of nicotine-evoked irritation on each side of the tongue every minute for a total of 4 min. 2.1.4. Rating procedures 2.1.4.1. Bipolar scale. Subjects rated the intensity of successively evoked sensations using a bipolar category scale that was specifically designed to accurately measure changes in perceived intensity across trials (see Fig. 1 of Ref. [4]). The method was virtually identical to that used in our previous studies [4,37] and is only briefly described. On a single sheet of paper is printed a series of 10 identical scales, each having the category `1' in the center, and categories ranging from 2 to 15 toward the right and left of the center. The sensation elicited by the first stimulus is obligatorily assigned a rating of `1.' The sensation elicited by the second stimulus is assigned a higher number toward the right if the sensation is stronger, or a higher number toward the left if the sensation is weaker, than the initial rating. Subsequent ratings are made in this manner for all 10 stimuli. We believe that this procedure has advantages over other commonly used rating scales [4]. First, this procedure allowed subjects to compare each evoked sensation with those elicited earlier in the series, rather than assigning numerical values to each sensation separately. Second, because the scale generates ratings of a more relative nature, it allowed us to provide extra categories at either end to prevent end-effects. Third, the fact that subjects had access to their previous ratings should limit memory errors (see Ref. [39]). 2.1.4.2. Two-alternative forced choice. This method was used both for monitoring the cooling sensation during the rest period, as well as comparing the nicotine-evoked irritation on each side of the tongue [40]. In the former instance, subjects were asked to indicate which side of the tongue gave rise to the stronger cooling sensation. For nicotine, subjects were asked to choose which side yielded a stronger irritant sensation. The use of the 2-AFC ensured sufficient sensitivity to detect small and confusable differences between the two sides of the tongue, and also enabled d0 values to be computed as a measure of difference magnitude. However, this method is inappropriate for perfectly discriminable differences as d0 tends toward infinity. Accordingly, subjects also gave ratings of the intensity of perceived cooling or irritant sensation on each side of the tongue separately using a unipolar category scale with 0 (no sensation) and 10 (intense sensation) as anchors. This unipolar scale was used instead of the bipolar scale because the goal was to compare intensities on the two sides of the tongue over time, rather than to record relative increases or decreases in intensity during repeated stimulation on one side (see Ref. [4]). Furthermore, use of the bipolar scale would be inappropriate for this task because it requires initial ratings of `1' for each side of the tongue even if there were a bilateral difference in intensity.

27

It is possible that nonirritant qualities might have been included in the subject's overall rating of nicotine irritation [41,42]. Any concomitant bitter taste of nicotine would be expected to be equal on both sides and therefore should not influence the 2-AFC or the bilateral intensity ratings. Although none of the subjects reported any taste sensation with nicotine, it is conceivable that menthol might desensitize the taste of nicotine; such an effect might contribute to menthol cross-desensitization of the overall nicotine irritant rating. The same reasoning applies to a possible cooling quality that might be included in menthol irritancy ratings (see Section 4). The 2-AFC and rating tasks were performed every minute, starting at 1 min after the end of the sequential stimulation and for the entire 5-min rest period when monitoring the cooling sensation, and every 30 s starting 15-s post-stimulus onset for a total duration of 1.5 min when evaluating nicotine-evoked irritation. 2.1.5. Statistical analysis The data obtained using the bipolar rating scale were analyzed using analysis of variance (ANOVA) and Fisher's Least Significant Difference (LSD) post hoc tests. To perform the ANOVA, the ratings were first transformed by allocating a positive sign to category ratings on the right side of the scale (stronger sensations) and a negative sign to those on the left side of the scale (weaker sensations). Then, the scale was renumbered by restoring zero to the center of the scale [4]. The self- and cross-desensitization experiments were analyzed using a d0 analysis [43,44] with an additional binomial analysis for the 2-AFC results and Student's t test for the intensity ratings. For all significant differences reported, P < .05. 2.2. Results and discussion 2.2.1. Desensitization by sequential menthol stimulation at 30-s ISI The irritation induced by 0.3% menthol significantly decreased across trials at 30-s ISI (ANOVA, P < .001; Fig. 1), confirming earlier results [25,26]. This is consistent with self-desensitization. Unlike the results of Cliff and Green, however, we did not observe large interindividual differences. Thus, 19 of the 22 subjects exhibited significant reductions in sensation across the 10 trials. Two subjects gave ratings that decreased across trials but fell short of significance, while ratings of one subject ended slightly higher than the initial rating. The rate of decline in ratings varied across subjects, as might be expected since subjects use scales in different ways [45]. 2.2.2. Cooling sensation induced by menthol Menthol elicited a significant cooling sensation. Thus, 1 min after the end of the sequential stimulation, a significant majority of subjects reported the menthol-treated side as

28

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

Fig. 1. Desensitization by menthol. (A) Graph plots mean irritant intensity reported 25 s following stimulus onset as a function of trials of menthol (19.2 mM) application at 30-s ISI. Error bars: S.E. Asterisk indicates first significant decrease from initial rating ( P < .05, ANOVA with LSD post hoc test).

having a stronger cooling sensation (21/22, binomial, P < .001). This corresponded to a significant group d0 value of 2.4 ( P =.001). Mean intensity ratings were also significantly different (t test, P < .001; Fig. 2). Despite a significant decrease in the intensity of the cooling sensation on the treated side (t test, P < .001), the sensation was still significantly stronger on that side 4 min after the end of the sequential stimulation (t test, P =.004). A minute later, nicotine was applied to both sides simultaneously. 2.2.3. Menthol cross-desensitization of nicotine-evoked irritation The irritant sensation elicited by nicotine was significantly reduced on the side that had received menthol. Thus, a significant majority of subjects chose the nontreated side as stronger (21/22, binomial, P < .001; d0 = 2.4, P =.001). In addition, the mean intensity rating was significantly higher on the nontreated side (t test, P < .001; Fig. 3). This

Fig. 3. Effect of menthol on nicotine-evoked irritation after a 5-min rest period postmenthol. Menthol-induced cooling sensation was still present. Graph plots the mean irritant intensity for the menthol pretreated (T, filled circles) and nontreated (NT, open circles) side of the tongue vs. time following bilateral application of nicotine. Note the significant depressant effect of prior menthol treatment. Error bars: S.E. Asterisks indicate significant difference between T and NT ( P < .05, t test).

suggests that menthol cross-desensitized irritation elicited by nicotine. However, because the cooling sensation induced by menthol had not completely vanished at the time that nicotine was applied, it could not be ruled out that residual cooling might have at least partly accounted for the reduction in nicotine irritation on the menthol-treated side. In support of this, the perceived intensity of capsaicin burn in the oral cavity has been observed to decrease linearly with temperature down to 17°C [46]. Topical application of menthol decreased histamine-induced itch sensation on the forearm as effectively as a drop in temperature from 32.8°C to 29.7°C [47], confirming menthol's capacity to suppress chemogenic sensations. However, menthol's antipruritic effect remains controversial because other studies failed to observe it [38,48]. Nevertheless, it is conceivable that the residual cooling sensation from menthol stimulation in the present study was responsible for the decrease in nicotine irritation. For this reason, and to determine whether menthol induces cross-desensitization of nicotine-evoked irritation independent of any cooling effect, a second experiment was performed in which nicotine was applied only after the cooling sensation of menthol had vanished. 3. Experiment II 3.1. Materials and methods

Fig. 2. Cooling sensation elicited by menthol. Graphs plots the mean intensity of cooling sensation for the treated (T, filled circles) and nontreated (NT, open circles) side of the tongue vs. time following unilateral menthol stimulation. Error bars: S.E. Asterisks indicate significant difference between T and NT ( P < .05, t test).

3.1.1. Subjects A total of 27 subjects (5 males, 22 females, aged 18 ± 50 years), who were students and staff at the University of California at Davis, volunteered to participate in the study. None had participated in Experiment I. All were nonsmokers and were asked to comply with the same restrictions as in Experiment I.

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

3.1.2. Chemicals, application procedures and experimental design Chemicals, application procedures and experimental design were identical to those of experiment I except for the differences outlined below. The intention of this experiment was to apply nicotine only after the cooling sensation on the menthol-treated side had vanished, or was equivalent to that on the untreated side. However, since potential cross-desensitization by menthol might be short-lived (see Ref. [1]), it was critical to hold the rest period to a minimum. In 15 female subjects (aged 18 ±22 years), menthol was applied 10 times at a 1min interval (30-s ISI), after which subjects rested for 5 min with the mouth closed. At the fifth minute, they were asked to indicate which side of the tongue had a stronger cooling sensation (2-AFC) and also to give intensity ratings on each side separately using the same 0 ±10 category scale as in Experiment I. This procedure was repeated every minute until either of the following criteria was reached: (1) both intensity ratings on the unipolar scale were equal, or (2) both were less than `1,' whichever occurred first. A bilateral rating of `1' was accepted as a criterion for absence of cooling, because bilateral ratings of `0' might not be given for some time even after the sensation had actually vanished. This reasoning is based on prior evidence that the judgment of absence of sensation is prone to response bias (criterion shift, see Ref. [49]). Nicotine was then applied 1 min after either criterion for absence of cooling sensation was reached. Nicotine was applied, and 2-AFC tests and bilateral intensity ratings were made, as in Experiment I. Because in both Experiments I and II the menthol vehicle contained ethanol, which itself has irritant and self-desensitizing properties (e.g., Ref. [50,51]), it was necessary to control that application of the vehicle did not affect nicotineevoked irritation. For this, 12 other subjects (5 males, 7 females, aged 19 ± 50 years) participated in a separate control session that was identical except that vehicle (4% ethanol and 1% Tween), rather than menthol, was applied 10 times at 1-min intervals and subjects rated the perceived intensity of irritation using the bipolar scale. Then, after a 5min rest period, 0.6% nicotine was applied bilaterally and subjects performed the 2-AFC and rating tests.

29

after the last stimulus application. Thus, 5 min after the end of the sequential menthol stimulation, a significant majority of subjects (14/15; binomial, P =.001) indicated that the menthol-stimulated side yielded a stronger cooling sensation, which corresponded to a significant group d0 of 2.12 ( P =.0026). The mean rating of cooling on that side was also significantly higher (3.5 vs. 0.9; t test, P < .001). The time at which the cooling sensation had vanished (rating = 0), or was equal or less to than that on the nontreated side, varied from 5 to 16 min after the last presentation of menthol (mean value: 10.73 min ‹ 3.53 S.D.). After menthol-evoked cooling sensation has disappeared, nicotine was applied bilaterally. The side not previously receiving menthol was chosen as stronger by a majority of subjects (12/15, binomial, P =.035), which corresponded to a group d0 value of 1.19 ( P =.019). The mean ratings were also significantly different (t test, P =.008; Fig. 4). It is therefore concluded that menthol cross-desensitized nicotine-induced irritation independent of its cooling effect. However, the smaller d0 value obtained in this second experiment indicates that the cross-desensitizing effect of menthol was weaker that that observed in Experiment I. This finding has two possible explanations. First, the residual cooling sensation in Experiment I might have partly contributed to the suppression of nicotine-evoked irritation (see Section 5). Second, the tongue may have already started to recover from a short-lived cross-desensitizing effect of menthol by the time nicotine was applied in Experiment II. The latter possibility is supported by the observation that menthol cross-desensitization of capsaicin-evoked irritation wore off after approximately 5 min [1]. 3.2.1. Vehicle control Vehicle did not elicit any irritation upon repeated application, nor did it affect subsequent nicotine-evoked irritation. During repeated vehicle application, the inten-

3.2. Results and discussion Sequential application of menthol at 1-min intervals (30s ISI) again resulted in a significant decrease in the intensity of irritation (ANOVA, P < .001), confirming our results from Experiment I, as well as those of Cliff and Green [25,26]. In this experiment, however, interindividual variability was greater than that observed in Experiment I, since 5 of the 15 subjects rated the last stimulus as equal to or stronger than the initial stimulus, two significantly so. Also consistent with results in Experiment I, menthol elicited a significant cooling sensation that lasted >5 min

Fig. 4. Effect of menthol on nicotine-evoked irritation after the cooling sensation from menthol stimulation had vanished. Graph as in Fig. 3, plotting the mean irritant intensity for the menthol pretreated (T, filled circles) and nontreated (NT, open circles) side of the tongue vs. time following bilateral application of nicotine. Note the significant crossdesensitization despite absence of lingering cooling.

30

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

sity of evoked irritation did not change significantly from the initial intensity (ANOVA, P =.4). In addition, after the 5-min rest period and bilateral application of nicotine, 7 of the 12 subjects reported the nontreated side as yielding stronger irritation (binomial, P =.77) and the mean ratings on each side were not significantly different (5.3 vs. 5.0; t test, P =.31). We conclude that the vehicle did not contribute to cross-desensitization of nicotine irritation by menthol. 4. Experiment III Experiments I and II confirm that menthol induces a progressive self-desensitization when applied at a 1-min interval (30-s ISI). The same phenomenon has been observed for other irritant chemicals, such as zingerone [52,53], nicotine [3,4], ethanol and cinnamic aldehyde [51]. However, additional irritant compounds exhibit sensitization (i.e., an increase in perceived irritation), rather than desensitization, when applied to the tongue at 30-s ISI. This is the case for capsaicin (e.g., Refs. [3,33,34]), piperine [4], citric acid [37] and monovalent salts (NaCl, KCl) at concentrations eliciting irritation [54,55]. It thus appears that some chemicals induce sensitization, and others desensitization, when applied at short ISI, although sensitization can be highly variable among individuals and sometimes even across replicate sessions (e.g., Ref. [4]). Following the original work of Green [33,54], most subsequent studies employing sequential stimulus application have adopted the 1-min interval application with 30-s ISI. However, a recent study showed that capsaicin-induced sensitization is more pronounced and consistent at shorter ISIs of 15 or 5 s [56], compared to the more variable effects seen at 30-s ISI. It is conceivable that other irritant chemicals might also induce consistent sensitization at shorter ( < 30 s) ISI. The sequential application of menthol at 30-s ISI induces an overall desensitization, although there was variability across subjects with some exhibiting sensitization instead (Ref. [26] and present results). We therefore investigated if sequential application of menthol at a short (5 s) ISI might more consistently induce sensitization, rather than desensitization. Irritant chemicals that generally induce sensitization, such as capsaicin and piperine, also exhibit SIR [36]. Thus, following sequential application of the irritant and a rest period of >5 min, reapplication of the irritant at first elicits little or no irritation due to self-desensitization; however, repeated sequential reapplication eventually induces an increase in irritant sensation that is called SIR. The second part of Experiment III was therefore undertaken to determine if menthol exhibits SIR. Thus, following sequential application of menthol at a short ISI and a subsequent rest period, the menthol was recurrently applied at the same ISI of 5 s to determine if the irritant sensation again grew in intensity.

4.1. Materials and methods 4.1.1. Subjects Nineteen healthy individuals (4 males, 15 females, aged 18 ± 45 years), who were students and staff at the University of California at Davis, volunteered to participate in the study. A total of 17 had participated in Experiment I. All were nonsmokers and complied with the same restrictions as in Experiments I and II. 4.1.2. Chemicals and application procedures The same 0.3% menthol solution used in Experiment I and II was prepared and 15 ml were pipetted onto small and 40 ml onto large filter paper disks. 4.1.3. Experimental design The experimental design was the same as that used in Experiments I and II except for the differences noted below. Menthol was sequentially applied on to one side of the subject's tongue 10 times at 20-s intervals with an ISI of 5 s. A large filter paper soaked with 40 ml of the 0.3% menthol solution was placed on to one side of the anterior dorsal surface of the tongue, left on for 15 s, removed and replaced 5 s later with a new menthol-soaked filter paper. With each application, the subject gave an intensity rating 10 s after stimulus onset using the bipolar scale. The side of the tongue receiving menthol was counterbalanced across subjects. Following the stimulus series, subjects rested for 5 min with the mouth closed. Menthol was then applied bilaterally using small filter papers as for nicotine in Experiments I and II, and subjects performed the same 2AFC and rating tests 30 s, 1 min and 1.5 min after stimulus onset, to test for self-desensitization. The following procedure was undertaken to test for SIR. Thirty seconds after the last rating was obtained, menthol was recurrently applied unilaterally at the same 5-s ISI to the same side of the tongue that had previously received sequential unilateral menthol. 4.2. Results and discussion 4.2.1. Sensitization followed by desensitization at 5-s ISI When menthol was applied at 20-s intervals (5-s ISI), the perceived irritation first increased, albeit not significantly (ANOVA with LSD post hoc test, P =.16), and then decreased progressively to levels significantly lower than the initial intensity (ANOVA, P =.01; Fig. 5A; thick line). However, examination of the individual rating curves reveals different patterns (Fig. 5A). One group (N = 3 subjects) exhibited consistent desensitization as in Experiment I. The majority of subjects (N = 15) exhibited an initial increase in irritant intensity, followed by a progressive decrease to a level that was in some, but not all, cases significantly lower than the initial rating. One subject exhibited a clear sensitization. This interindividual variability is reminiscent of that observed for sensitization by

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

31

These results indicate that sequential application of menthol at a shorter ISI induced an initial sensitization and diminished the magnitude of desensitization across trials. The brief sensitization with menthol is different from the pattern of progressive sensitization observed with capsaicin, piperine, citric acid or NaCl. It may be that the rapid presentations of menthol lead to an enhancement of irritation via temporal and/or spatial summation, but at the same time trigger a desensitization process that cannot be neutralized even with continued replenishment of menthol. Similarly, dual processes for capsaicin sensitization and desensitization were previously suggested, whereby maintenance of a constant capsaicin concentration in the vicinity of the lingual receptors is sufficient to prevent desensitization from occurring [56]. The cellular mechanisms underlying sensitization and desensitization at the level of nociceptor endings in the lingual epithelium are currently not understood. The present results suggest that there are fundamental differences between menthol and capsaicin in their ability to initiate cellular processes of desensitization when their intraoral concentration is maintained (see Section 5).

Fig. 5. Response to repeated stimulation of menthol. (A) Individual results. Graph plots mean (thick line) and individual irritant intensity ratings (thin lines) reported 25 s following stimulus onset as a function of trials of menthol (19.2 mM) application at 30-s ISI. Arrows denote the three subjects who desensitized and one subject who sensitized. The remaining subjects followed a common trend of increase followed by decrease in intensity with varying time courses. (B) Averaged response for the 15 subjects exhibiting a similar pattern (see above). Error bars: S.E. Asterisk indicates first significant increase/decrease from initial rating ( P < .05, ANOVA with LSD post hoc test).

irritants, such as capsaicin or piperine (e.g., Refs. [3,4]). When data from the 15 subjects exhibiting an initial sensitization followed by a desensitization were analyzed separately (eliminating the three desensitizing and the one sensitizing subject who did not follow the majority trend; Fig. 5B), the initial sensitization was significant ( P < .001) and the mean intensity of irritation at the end of the stimulus series was lower compared to the initial rating of `1,' although this did not achieve significance ( P =.057). Because 17 subjects participated in both Experiments I and III, it was possible to directly compare their ratings across conditions. The mean rating obtained on the second trial was significantly higher for the 5-s than the 30-s ISI condition, indicating a shift from desensitization to sensitization with shorter ISI. The mean rating on the10th trial was significantly higher for the 5-s than the 30-s ISI, indicating that irritation declined after the initial increase at the 5-s ISI but not as rapidly as at the 30-s ISI. Conceivably, desensitization approaching that observed in Experiment I might have been reached with more trials.

4.2.2. Menthol self-desensitization Following sequential application of menthol at 20-s intervals (5-s ISI; see above) and a subsequent 5-min rest period, menthol was applied bilaterally in the same subjects. All 19 subjects chose the side of the tongue that had not previously received menthol as having a stronger irritation (binomial, P < .001), and the mean ratings on the two sides were significantly different (t test, P < .001; Fig. 6), in confirmation of earlier studies [25,26]. Because discrimination was perfect for all subjects in the 2-AFC, the group d 0 value tends toward infinity. In this case, various methods can be used to estimate d 0. The method of Bock and Jones [57] yields an approximate group d 0 of 2.83 ( P < .001). This value can be compared with the d0 value of 2.4 obtained for menthol cross-desen-

Fig. 6. Menthol self-desensitization. Graph as in Figs. 3 and 4, plotting the mean irritant intensity for the menthol pretreated (T, filled circles) and nontreated (NT, open circles) side of the tongue vs. time following bilateral application of menthol. Note the significant self-desensitization. Error bars: S.E. Asterisks indicate significant difference between T and NT ( P < .05, t test).

32

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

Fig. 7. SIR during repeated stimulation with menthol following selfdesensitization. Graph plots mean irritant intensity reported 25 s following stimulus onset as a function of trials of menthol (19.2 mM) application at 30-s ISI. Error bars: S.E. Asterisk indicates the first significant increase from the initial intensity rating.

sitization to nicotine. That self-desensitization was greater than cross-desensitization is consistent with previous results (e.g., Refs. [3,4,36,58]) and suggests that while menthol and nicotine stimulate a common population of nociceptors, it is likely that menthol also activates nicotine-insensitive nociceptive pathways. 4.2.3. Stimulus-induced recovery Following the experiment described in the preceding paragraph, menthol was then recurrently applied to the side of the tongue that had previously received menthol in the same subjects at the same 20-s interval (5-s ISI). The mean intensity of irritation rose significantly across trials

(ANOVA, P =.028; Fig. 7), consistent with SIR as previously established for capsaicin [36]. Inspection of individual rating curves revealed that only 13 of the 22 subjects reported a significant increase in intensity by the 10th trial, with considerable intersubject variability in ratings across trials. Interestingly, a significant proportion of subjects (17/ 22; binomial, P =.02) reported a significant increase in irritation on at least one trial, with considerable interindividual variability in the time at which the peak intensity was reported. It is of further interest to compare SIR with the individual ratings obtained during the initial stimulus series. The general pattern was one of sensitization followed by desensitization, but there was also considerable variability in the time at which the peak intensity was reported (Fig. 5A). Among the subjects exhibiting SIR, distinct patterns could be identified. One group (N = 5) showed a pattern similar to that observed during the initial stimulus series, i.e., initial sensitization followed by desensitization (Fig. 8A; only three subjects represented for clarity). Another group (N = 3) exhibited a progressive increase in irritant intensity across trials (Fig. 8B). A third group (N = 3) exhibited a biphasic pattern starting with further desensitization, and followed later by an increase in irritant intensity (Fig. 8C). A fourth group (N = 6) exhibited a triphasic pattern characterized first by an increase in irritant intensity, followed by a return to or below baseline and then a second late increase (Fig. 8D). The categories described here are somewhat arbitrary; for example, it might be argued that the curves shown in Fig. 8A and D constitute one group with differences in the degree of late desensitization. Similarly, some subjects in Fig. 8C that exhibit a limited initial desensitiza-

Fig. 8. Variability of menthol-induced SIR across subjects. Graphs plot the irritant intensity reported 25 s following stimulus onset as a function of trials of menthol (19.2 mM) application at 30-s ISI for each subject. Each panel gives three examples of subjects having a similar pattern of response. (A) Increase followed by decrease, (B) increase only, (C) decrease followed by increase, (D) increase followed by decrease then increase.

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

tion could be considered to belong to the group in Fig. 8B. Finally, desensitization might have been achieved with more stimulus trials in Fig. 8B and C. The variable patterns shown in Fig. 8 might therefore merely represent the variability in time (and/or stimulus trials) needed to achieve maximal irritant intensity. Despite differences in the time course, recovery of the irritant sensation of menthol does occur at some time in a majority of subjects. This argues in favor of an SIR-like phenomenon for menthol. 5. General discussion The present experiments demonstrate three new properties of menthol as an oral irritant chemical: (a) exposure to menthol cross-desensitized irritation elicited by nicotine; (b) when applied at a short (5-s) ISI, a significant proportion of subjects perceived the menthol irritation to increase briefly before desensitization appeared; and (c) when menthol was reapplied following a rest period, most subjects appeared to exhibit recovery from desensitization, i.e., SIR. 5.1. Molecular mechanisms of menthol irritation It is currently not certain whether menthol depolarizes trigeminal nerve endings by binding to a specific molecular receptor, or via a nonspecific membrane effect. The concept of a specific menthol receptor is supported by the following observations. Most importantly, the D-menthol isomer was much weaker than L-menthol for both cooling [59] and irritant sensations [30]. The latter study used time-intensity to show that L-menthol induced significantly stronger and longer-lasting intraoral sensations of burning and cooling compared to D-menthol. Second, intraoral application of menthol activated the rat lingual nerve in a manner that was not proportional to the octane/water partition coefficient, in contrast to other alcohols [60]. Finally, menthol also modulates voltage-sensitive Ca2 + currents in helix [61] and in mammalian DRG neurons [31,62], and this is not reproduced by exposure to cyclohexanol, the highly lipidsoluble alcohol from which menthol is derived. However, the type of sensory stimuli transduced by the recorded neurons in the later studies cannot be determined (i.e., thermoreceptors vs. nociceptors). Thus, while the above findings suggest that menthol's sensory effects are mediated via a specific molecular receptor, nonspecific actions of menthol might also play some role in activating nociceptors, particularly since higher concentrations of menthol are needed to elicit irritation compared to cooling. 5.2. Menthol desensitization, self-desensitization and cross-desensitization We presently showed that menthol induces either selfdesensitization or an initial sensitization, depending on the application paradigm. This emphasizes the general impor-

33

tance of temporal factors in the perception of irritant stimuli, which are important to consider in real-life applications such as in the industry. For example, menthol is a common additive to toothpaste, and the tingling irritant sensation associated with brushing the teeth would be better evaluated in a sensory test that utilized either continuous or intermittent application of menthol at a short ISI. By analogy, it would be interesting to determine if other irritants, such as nicotine [3,4] or zingerone [52,53], which induce reliable self-desensitization when applied at 15 ± 30-s ISI, might exhibit sensitization if applied at shorter ISI, as was presently the case for menthol. Menthol self-desensitization has been well established [25,26] and was confirmed in Experiment III of the present study (Fig. 6). All other irritant chemicals studied to date appear to possess this property (e.g., Refs. [4,33,36,37,51± 53,55]). On the other hand, not all irritant chemicals are capable of inducing cross-desensitization, suggesting that the cellular processes underlying self- and cross-desensitization may be different. Capsaicin cross-desensitizes irritation elicited by a variety of irritant chemicals [3,35,58,63]. Other irritants inducing cross-desensitization include piperine (e.g., Refs. [4,36]), citric acid [37] and nicotine at higher concentrations [64]. We presently observed cross-desensitization of nicotine-evoked irritation by menthol, and it was recently reported that menthol cross-desensitized irritation elicited by capsaicin [1]. In both cases, the cross-desensitization appeared to be independent of menthol's cooling effect (see below). Although the cellular mechanism underlying menthol cross-desensitization is not yet known, one might speculate that it is linked to menthol activation of nociceptor endings and the initiation of Ca2 + -mediated intracellular events that ultimately reduce cellular excitability, as suggested for capsaicin (e.g., Refs. [65,66]). Curiously, menthol does not appear to affect pain sensation elicited by noxious heat [27,29]. This also argues against the possibility that menthol cross-desensitization to capsaicin and nicotine is the result of a nonspecific anesthetic effect. Until recently, irritant chemicals shown to induce crossdesensitization also tended to induce sensitization upon repeated application: capsaicin (e.g., Refs. [3,33]), piperine (e.g., Refs. [4,36]) and citric acid [37]. The present results with menthol, and our recent study with high concentrations of nicotine [64], indicate that nonsensitizing irritants are also capable of inducing cross-desensitization. However, not all sensitizing irritants induce cross-desensitization, as we recently demonstrated for 5 M NaCl, which induces robust sensitization at 30-s ISI but cross-sensitizes, rather than cross-desensitizes, capsaicin-evoked irritation [55]. The duration of menthol-induced cross-desensitization appears to be much briefer compared to that induced by capsaicin (Ref. [1] and present results). Menthol crossdesensitization of nicotine was weaker after a longer (10 min) compared to shorter (5 min) rest period, possibly indicating that the effect was wearing off. In contrast, capsaicin cross-desensitization is still apparent after much

34

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

longer rest periods [36,67]. In the present study, crossdesensitization of nicotine was still present 5 ± 16 min postmenthol, while in a previous study [1], cross-desensitizing effects to capsaicin had ceased after approximately 5 min. This might be explained by the higher concentration of menthol used here (19.2 vs. 3.3 ± 10 mM) or by a difference in the effect of menthol on sensation elicited by nicotine vs. capsaicin. However, another interpretation of the decrease in the magnitude of menthol cross-desensitization between Experiments I and II is that cooling might have contributed to the reduction of nicotine-evoked irritation in Experiment I. Central inhibitory effects of cooling on pain have been described (e.g., Ref. [68]). Cooling also reduces oral irritation elicited by capsaicin [46]. Furthermore, as noted earlier, cooling the arm by physically reducing the temperature, or by topical application of menthol, both reduced histamineevoked itch sensation [47]. Physical cooling of skin also reduced responses of spinal dorsal horn neurons to intracutaneous histamine [69]. It is uncertain whether the antinociceptive and antipruritic effects of cooling and menthol are due to a peripheral action on stimulus transduction, or to a central action. Cooling and menthol excite peripheral cold receptors (e.g., Ref. [23]), which might participate in central inhibition of nociceptive transmission (see Ref. [70]). However, the inability of menthol to increase the heat pain threshold [27,29] is not consistent with the idea that menthol has a general antinociceptive action. 5.3. Stimulus-induced recovery Following menthol-induced self-desensitization, reapplication of menthol induced a recovery of irritant sensation in a significant majority of subjects that appears to be consistent with SIR. SIR was first documented for capsaicin and piperine [36], both of which induce sensitization. The SIR observed in the present study with menthol was, however, not as consistent or marked as that observed for capsaicin or piperine, and exhibited considerable intersubject variability in its time course (Fig. 8). The rather weak and variable SIR observed with menthol might reflect the limited degree of sensitization that was obtained in the first stimulus series at short 5-s ISI (Fig. 5). That is, the degree of SIR may be related to the degree of sensitization that is elicited in the first stimulus series. The cellular mechanisms underlying SIR with capsaicin or menthol are not known. For capsaicin, two possible explanations have been proposed [56]. The first one proposes that SIR and sensitization are mediated by a common excitatory cellular mechanism that is opposed by a separate cellular process of desensitization. For capsaicin, the excitatory process activated by repetitive stimulation either outweighs desensitization or prevents desensitization from beginning, resulting in sensitization. During the rest period, the process of desensitization either begins or continues, while that of sensitization diminishes, such that

desensitization is expressed. Recurrent stimulation reinitiates the sensitization mechanism that eventually overcomes desensitization to result in SIR. The second explanation proposes that SIR requires the recruitment of low-affinity capsaicin (VR-1) receptors that were not desensitized by prior exposure. The initial repeated application of menthol at a short 5-s ISI resulted in a rapid sensitization, followed by desensitization (Fig. 5). If opposing cellular mechanisms mediate sensitization and desensitization as suggested above for capsaicin, then with menthol the sensitization process manifests itself earlier, but is quickly overcome by the development of self-desensitization. After the rest period, recurrent application of menthol at a 5-s ISI resulted in a more slowly developing and maintained increase in average irritant intensity (Fig. 7), consistent with SIR. 5.4. Relation to usage of mentholated tobacco products The present results have shown a significant crossdesensitization of nicotine-evoked irritation for up to 16 min following menthol pretreatment (Figs. 3 and 4). When menthol and nicotine are delivered simultaneously, as during the consumption of mentholated tobacco products, both compounds will activate sensory nerve endings in the oral, nasal and pulmonary mucosae. Thus, the ensuing menthol cross-desensitization and a potential antinociceptive effect of cooling would reduce the degree of nicotine-induced irritation. We did not currently test if nicotine reciprocally cross-desensitizes menthol-induced irritation. However, this is a tenable possibility based on our recent finding that high (300 mM) but not lower (7.4 or 74 mM) nicotine concentrations cross-desensitized capsaicin-induced irritation [64]. One might speculate that cross-desensitization of the irritant sensations elicited by one or both compounds would enhance the acceptance of mentholated tobacco products by reducing aversive sensations, especially when the products are tried for the first time. With chronic tobacco usage, cross-desensitization could be detrimental since the tissue-protective signal provided by the irritant sensation is reduced. Alternatively, the irritant impact or ``kick'' of nicotine, and possibly also irritant/cooling sensations elicited by menthol, may be positive hedonic factors contributing to the liking of tobacco products, analogous to the preference that many people have for spicy foods (see Ref. [71]). When one starts smoking or ingesting mentholated tobacco, the summed irritant sensations of nicotine and menthol will be maximal, and will decline as cross-desensitization and cooling effects set in. This, in turn, might lead to increased consumption to obtain the desired irritant sensation. In the absence of any evidence that menthol has major central physiological or psychological effects (e.g., Ref. [72]), we conclude that the preference many tobacco users show for mentholated products is due to some positive affective factor, such as increased impact, antinociception

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

or a ``fresh'' oral sensation. Each of these factors could lead to increased usage of mentholated tobacco products, thus, resulting in exposure to higher doses of nicotine with a resultant increase in liability for addiction. Acknowledgments This work was supported by a grant from the California Tobacco-Related Disease Research Program No. 6RT-0231. References [1] Green BG, McAuliffe BL. Menthol desensitization of capsaicin irritation: evidence of a short-term anti-nociceptive effect. Physiol Behav 2000;68:631 ± 9. [2] McCarthy WJ, Caskey NH, Jarvik ME, Gross TM, Rosenblatt MR, Carpenter C. Menthol vs. nonmenthol cigarettes: effects on smoking behavior. Am J Public Health 1995;85:67 ± 72. [3] Dessirier J-M, O'Mahony M, Carstens E. Oral irritant effects of nicotine: psychophysical evidence for decreased sensation following repeated application and lack of cross-desensitization to capsaicin. Chem Senses 1997;22:483 ± 92. [4] Dessirier J-M, Nguyen N, Sieffermann J-M, Carstens E, O'Mahony M. Oral irritant properties of piperine and nicotine: psychophysical evidence for asymmetrical desensitization effects. Chem Senses 1999;24:405 ± 13. [5] Dessirier J-M, O'Mahony M, Sieffermann J-M, Carstens E. Mecamylamine inhibits nicotine but not capsaicin irritation on the tongue: psychophysical evidence that nicotine and capsaicin activate separate molecular receptors. Neurosci Lett 1998;240:65 ± 8. [6] Jarvik ME, Assil KM. Mecamylamine blocks the burning sensation of nicotine on the tongue. Chem Senses 1988;13:213 ± 7. [7] Grieff L, Wollmer P, Andersson M, Pipkorn U, Persson CGA. Effects of nicotine on the nasal mucosa. Thorax 1993;48:651 ± 5. [8] Hummel T, Livermore A, Hummel C, Kobal G. Chemosensory eventrelated potentials in man: relation to olfactory and painful sensations elicited by nicotine. Electroencephalogr Clin Neurophysiol 1992; 84:192 ± 5. [9] Cain WS. Sensory attribute of cigarette smoking. In: Gori GB, Bock FG, editors. Banbury report: 3. A safe cigarette? . Cold Spring Harbor, NY: Cold Spring Harbor Laboratory, 1980. pp. 239 ± 49. [10] Jansco N, Jansco-Gabor A, Takats I. Pain and inflammation induced by nicotine, acetylcholine and structurally realted compounds and their prevention by desensitizing agents. Acta Physiol 1961;19:113 ± 32. [11] Keele CA, Armstrong D. Substances producing pain and itch. Baltimore: Williams and Wilkins, 1964. [12] Holladay MW, Dart MJ, Lynch JK. Neuronal nicotinic acetylcholine receptors as targets for drug discovery. J Med Chem 1997;40: 4169 ± 94. [13] Lindstrom J, Anand R, Gerzanich V, Peng X, Wang F, Wells G. Structure and function of neuronal nicotinic acetylcholine receptors. Prog Brain Res 1996;109:125 ± 37. [14] Alimohammadi H, Silver WL. Evidence for nicotinic acetylcholine receptors on nasal trigeminal nerve endings of the rat. Chem Senses 2000;25:61 ± 6. [15] Tanelian DL. Cholinergic activation of a population of corneal afferent nerves. Exp Brain Res 1991;86:414 ± 20. [16] Steen KH, Reeh PW. Actions of cholinergic agonists and antagonists on sensory nerve endings in rat skin, in vitro. J Neurophysiol 1993; 70:397 ± 405. [17] Wang Y, Erickson RP, Simon SA. Selectivity of lingual nerve fibers to chemical stimuli. J Gen Physiol 1993;101:843 ± 66.

35

[18] Jinks SL, Carstens E. Activation of spinal wide dynamic range neurons by intracutaneous microinjections of nicotine. J Neurophysiol 1999;82:3046 ± 55. [19] Carstens E, Saxe I, Ralph R. Brainstem neurons expressing c-fos immunoreactivity following irritant chemical stimulation of the rat's tongue. Neuroscience 1995;69:939 ± 53. [20] Carstens EE, Kuenzler N, Handwerker HO. Activation of neurons in rat trigeminal subnucleus caudalis by application of different classes of irritant chemicals to the oral and occular mucosa. J Neurophysiol 1998;80:465 ± 92. [21] Liu L, Simon SA. Capsaicin and nicotine both activate a subset of rat trigeminal ganglion neurons. Am J Physiol 1996;270:C1807 ± 14. [22] Hensel H, Zottermann Y. The effect of menthol on thermoreceptors. Acta Physiol Scand 1951;24:27 ± 34. [23] Lundy RF, Contreras RJ. Neural responses of thermal-sensitive lingual fibers to brief menthol stimulation. Brain Res 1994;641:208 ± 16. [24] Sekizawa S, Tsubone H, Kuwahara M, Sugano S. Nasal receptors responding to cold and L-menthol airflow in the guinea pig. Respir Physiol 1996;103:211 ± 9. [25] Cliff MA, Green BG. Sensory irritation and coolness produced by menthol: evidence for selective desensitization of irritation. Physiol Behav 1994;56:1021 ± 9. [26] Cliff MA, Green BG. Sensitization and desensitization to capsaicin and menthol in the oral cavity: interactions and individual differences. Physiol Behav 1996;59:487 ± 94. [27] Green BG. Menthol inhibits the perception of warmth. Physiol Behav 1986;38:833 ± 8. [28] Green BG. Menthol modulates oral sensations of warmth and cold. Physiol Behav 1985;35:427 ± 34. [29] Green BG. The sensory effects of L-menthol on human skin. Somatosens Mot Res 1992;9:235 ± 44. [30] Gwartney E, Heymann H. The temporal perception of menthol. J Sens Stud 1995;10:393 ± 400. [31] Petersen M, Wagner G, Pierau FK. Modulation of calcium-currents by capsaicin in a subpopulation of sensory neurones of guinea pig. Naunyn-Schmiedeberg's Arch Pharmacol 1989;339:184 ± 91. [32] Farley LG, Silver WL. Self- and cross-adaptation to chemical stimulation of the nasal trigeminal nerve in the rat. Chem Senses 1992;17: 507 ± 18. [33] Green BG. Capsaicin sensitization and desensitization on the tongue produced by brief exposures to a low concentration. Neurosci Lett 1989;107:173 ± 8. [34] Green BG. Temporal characteristics of capsaicin sensitization and desensitization on the tongue. Physiol Behav 1991;49:501 ± 5. [35] Green BG. Evidence that removal of capsaicin accelerates desensitization on the tongue. Neurosci Lett 1993;150:44 ± 8. [36] Green BG. Rapid recovery from capsaicin desensitization during recurrent stimulation. Pain 1996;68:245 ± 53. [37] Dessirier J-M, O'Mahony M, Iodi-Carstens M, Carstens E. Sensory properties of citric acid: psychophysical evidence for sensitization, self-desensitization, cross-desensitization and cross-stimulus-induced recovery following capsaicin. Chem Senses 2000;25:769 ± 89. [38] Yosipovitch G, Szolar C, Hui XY, Maibach H. Effect of topically applied menthol on thermal, pain and itch sensations and biophysical properties of the skin. Arch Dermatol Res 1996;288:245 ± 8. [39] Kim K-O, O'Mahony M. A new approach to category scales of intensity: I. Traditional versus rank-rating. J Sens Stud 1998;13:241 ± 9. [40] Green DM, Swets JA. Signal detection theory and psychophysics. New York: Wiley, 1966. [41] Clark CC, Lawless HT. Limiting response alternatives in time ± intensity scaling: an examination of the halo-dumping effect. Chem Senses 1994;19:583 ± 94. [42] Frank RA, Van der Klaauw NJ, Schifferstein HN. Both perceptual and conceptual factors influence taste ± odor and taste ± taste interactions. Percept Psychophys 1993;54:343 ± 54. [43] Ennis DM. The power of sensory discrimination methods. J Sens Stud 1993;8:353 ± 70.

36

J.-M. Dessirier et al. / Physiology & Behavior 73 (2001) 25±36

[44] Bi J, Ennis DM, O'Mahony M. How to estimate and use the variance of d 0 from difference tests. J Sens Stud 1995;12:87 ± 104. [45] Kim K-O, O'Mahony M. A new approach to category scales of intensity: II. Use of d0 values. J Sens Stud 1998;13:251 ± 67. [46] Green BG. Sensory interaction between capsaicin and temperature in the oral cavity. Chem Senses 1986;11:371 ± 82. [47] Bromm B, Scharein E, Darsow U, Ring J. Effects of menthol and cold on histamine-induced itch and skin reactions in man. Neurosci Lett 1995;187:157 ± 60. [48] Melton H, Shelly WB. The effect of topical antipruritic therapy on experimentally induced itch. J Invest Dermatol 1950;17:325 ± 32. [49] O'Mahony M. Understanding discrimination tests: a user-friendly treatment of response bias, rating and ranking R-index tests, and their relashionship to signal detection. J Sens Stud 1992;7:1 ± 47. [50] Green BG. Effects of thermal, mechanical, and chemical stimulation on the perception of oral irritation. In: Green BG, Mason JR, Kare MR, editors. Chemical senses: vol. 2. Irritation. New York: Marcel Dekker, 1990. pp. 171 ± 92. [51] Prescott J, Swain-Campbell N. Response to repeated oral irritation by capsaicin, cinnemaldehyde and ethanol in PROP tasters and non-tasters. Chem Senses 2000;25:239 ± 46. [52] Prescott J, Stevenson RJ. Desensitization to oral zingerone irritation: effects of stimulus parameters. Physiol Behav 1996;60:1473 ± 80. [53] Prescott J, Stevenson RJ. Psychophysical responses to single and multiple presentations of the oral irritant zingerone: relationship to frequency of chili consumption. Physiol Behav 1996;60:617 ± 24. [54] Green BG, Gelhard B. Salt as an oral irritant. Chem Senses 1989;14: 259 ± 71. [55] Dessirier J-M, O'Mahony M, Iodi-Carstens M, Yao E, Carstens E. Oral irritation by sodium chloride: sensitization, self-desensitization and cross-sensitization to capsaicin. Physiol Beh 2001;72:317 ± 24. [56] Green BG, Rentmeister-Bryant H. Temporal characteristics of capsaicin desensitization and stimulus-induced recovery in the oral cavity. Physiol Behav 1998;65:141 ± 9. [57] Bock RD, Jones LV. The measurement and prediction of judgement and choice. San Fransico: Holden-Day, 1968. [58] Green BG. Capsaicin cross-desensitization on the tongue: psychophysical evidence that oral chemical irritation is mediated by more than one sensory pathway. Chem Senses 1991;16:675 ± 89. [59] Eccles R, Griffiths DH, Newton CG, Tolley NS. The effects of D and

[60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70]

[71]

[72]

L isomers of menthol upon nasal sensation of airflow. J Laryngol Otol 1988;102:506 ± 8. Simon SA, Sostman AL. Electrophysiological responses to non-electrolytes in lingual nerve of rat and in lingual epithelia of dog. Arch Oral Biol 1991;36:805 ± 13. Swandulla D, SchaÈfer K, Lux HD. Calcium channel current inactivation is selectively modulated by menthol. Neurosci Lett 1986;68: 23 ± 8. Swandulla D, Carbone E, SchaÈfer K, Lux HD. Effect of menthol on two types of Ca currents in cultured sensory neurons of vertebrates. Pfluegers Arch Eur J Physiol 1987;409:52 ± 9. Gilmore MM, Green BG. Sensory irritation and taste produced by NaCl and citric acid: effects of capsaicin desensitization. Chem Senses 1993;18:257 ± 72. Dessirier J-M, Chang H-K, O'Mahony M, Carstens E. Cross-desensitization of capsaicin-evoked oral irritation by high but not low concentrations of nicotine. Neurosci Lett 2000;290:133 ± 6. Cholewinski A, Burgess GM, Bevan S. The role of calcium in capsaicin-induced desensitization in rat cultured dorsal root ganglion neurons. Neuroscience 1993;55:1015 ± 23. Chard PS, Bleakman D, Savidge JR, Miller RJ. Capsaicin-induced neurotoxicity in cultured dorsal root ganglion neurons: involvement of calcium-activated proteases. Neuroscience 1995;65:1099 ± 108. Karrer T, Bartoshuk L. Capsaicin desensitization and recovery on the human tongue. Physiol Behav 1991;49:757 ± 64. Wahren LK, TorebjoÈrk E, JoÈrum E. Central suppression of cold-induced C fibre pain by myelinated fibre input. Pain 1989;38:313 ± 9. Jinks SL, Carstens E. Skin cooling attenuates rat dorsal horn neuronal responses to intracutaneous histamine. NeuroReport 1998;9:4145 ± 9. Beise RD, Carstens E, Kohlloeffel LUE. Psychophysical study of stinging pain evoked by brief freezing of superficial skin and ensuing short-lasting changes in sensations of cool and cold pain. Pain 1998;74:275 ± 86. Rozin P. Getting to like the burn of chili pepper: biological, psychological and cultural perspectives. In: Green BG, Mason JR, Kare MR, editors. Chemical senses: vol. 2. Irritation. New York: Marcel Dekker, 1990. pp. 231 ± 69. Gelal A, Jacob P, Yu L, Benowitz NL. Disposition kinetics and effects of menthol. Clin Pharmacol Ther 1999;66:128 ± 35.