The TRPA1 Ion Channel Contributes to Sensory-Guided Avoidance of Menthol in Mice

Body weights for individual mice (n = 99) on each brief-access test day expressed as a percentage of their initial, pre-water restriction weight (test day weight in grams ÷ initial weight in grams × 100). Data include mice from both the menthol and quinine studies. The cross to the right of each distribution gives its mean (horizontal bar) ± SD (vertical bar), as follows: test day 1, 90.8% ± 5.6%; test day 2, 88.8% ± 5.4%; test day 3, 88.5% ± 5.7%; test day 4, 89% ± 6.6%; test day 5, 89.2% ± 6.1%; test day 6, 89.2% ± 6.1%; test day 7, 89.5% ± 5.9%.

Distributions of latencies to first lick for individual mice of each line across menthol concentrations. The cross to the right of each distribution gives its median (horizontal bar) and 95% CI* (vertical bar). Sample sizes are as follows: 0 mM, n = 75; 0.3, 0.5, 1, 1.5, and 2.3 mM, n = 74; 0.7 mM, n = 73. Holm-adjusted p levels for Freidman’s ANOVAs conducted on the influence of menthol concentration on first lick latency for each line follow: B6, 0.45; B6129, 0.16; TRPM8^KO, 0.19; TRPA1^KO, 0.77; TRPA1^HET, 1.

Licking performance of mice in menthol studies. A, Mean standardized cumulative lick functions for water (0 mM menthol) for each mouse line (legend). Each curve tracks mean cumulative licks observed per water trial divided by the mean total licks the line completed at the end of the water access session (i.e., trial 20). B, Distributions of integrals (areas under curves) for water cumulative lick functions that contributed the average plots in panel A. The cross to the right of each distribution gives its mean (horizontal bar) and 95% CI* (vertical bar). One-way ANOVA identified that the mean area under the standardized cumulative lick function to water did not differ across B6 (n = 20), B6129 (n = 16), TRPM8^KO (n = 10), TRPA1^KO (n = 15), and TRPA1^HET (n = 12) mice (F_(4,68) = 1.6, p = 0.2). C, Mean standardized cumulative lick functions to the menthol concentration series for each mouse line. The curve for each concentration tracks mean cumulative licks per trial divided by the mean total licks mice completed to this concentration on trial 20. The legend in the B6 plot relating line shading/thickness to menthol concentration applies to all panels in C. D, Distributions of areas under the standardized cumulative lick curves for each concentration of the menthol series. Points represent data for the individual mice that composed the average plots for each line in panel C. The cross to the right of each distribution gives its mean (horizontal bar) and 95% CI* (vertical bar). Sample sizes were as stated for panel B. Repeated measures ANOVAs revealed that the mean area under the standardized cumulative lick function to menthol significantly changed with concentration for B6 (F_(6,114) = 48.3, p < 0.001), B6129 (F_(2.5,53.2) = 9.8, p < 0.001), TRPM8^KO (F_(6,54) = 3.6, p = 0.004), and TRPA1^HET (F_(6,66) = 3.1, p = 0.009) mice. In contrast, the area under the standardized cumulative lick function to menthol did not vary with concentration for TRPA1^KO mice (F_(6,84) = 0.5, p = 0.8). For panels B, D, distributions of points show uncorrected data, albeit statistical analyses were performed following 5% Winsorization to accommodate outliers.

Distributions of menthol:water lick ratios for individual B6 (n = 20) and B6129 (n = 17 except for 0.7 mM, where n = 16) mice across menthol concentrations. The cross to the right of each distribution gives its median (horizontal bar) and 95% CI* (vertical bar). Lick ratios did not differ between B6 and B6129 mice at any menthol concentration (p > 0.9). Collapsed across mouse line, lick ratios showed an initial significant decrease when stepping from 0.7 to 1 mM (p = 0.02) and further decreased from 1.5 to 2.3 mM (p < 0.001), as denoted by asterisks.

Distributions of lick counts to the menthol concentration series (A) and menthol:water lick ratios (B) for individual TRPM8^KO (n = 10) and simultaneously run wild-type (B6 line, n = 10) mice. The color-coded (legend, panel B) cross associated with each distribution gives its median (horizontal bar) and 95% CI* (vertical bar). Lick counts and ratios for each menthol concentration did not differ between TRPM8^KO and control mice (p > 0.16). C, Mean differences in lick ratios between female and male TRPM8^KO (five females, five males) and wild-type (five females, five males) mice in panel B. The plotted difference (circle) for each concentration and mouse line (legend for mouse line is given in panel B) was calculated as the mean lick ratio for males minus the mean lick ratio for females. Vertical bar spanning each mean difference represents its 95% CI* based on 10,000 resamples. The trends in this plot suggested that for each mouse line, there was no major influence of sex on lick ratios to the menthol concentration series. Accordingly, statistical analyses collapsed across mouse line to increase analyzed sample sizes revealed lick ratios to each concentration of menthol did not differ between female and male mice (p > 0.2).

Distributions of lick counts to the menthol concentration series (A) and menthol:water lick ratios (B) for individual TRPA1^KO (n = 15) and simultaneously run wild-type (n = 16; 10 B6 mice and six B6129 mice) mice. Plots and analyses reflect and accommodated missing data for two B6129 mice: one made no licks from 0.3 to 2.3 mM menthol (i.e., licked only to water) and the other made no licks to 0.7 mM menthol. The color-coded (legend, panel B) cross associated with each distribution gives its median (horizontal bar) and 95% CI* (vertical bar). TRPA1^KO mice showed higher lick counts and ratios to 2.3 mM menthol compared to control mice (p < 0.03), as denoted by asterisks.

Distributions of AITC:water lick ratios for individual TRPA1^KO (n = 5) and wild-type (n = 5; B6129 line) mice. Plot reflects missing data for one B6129 mouse that did not lick to 0.3 mM AITC. Ratios were calculated using licks to water measured during menthol testing. Color-coded (legend) boldened horizontal bars give median lick ratios for each line across AITC concentrations. Sample sizes were low for statistical analysis, although a trend for concentration-dependent avoidance of AITC was apparent and absent in the median responses of wild-type and TRPA1^KO mice, respectively.

Distributions of lick counts to the menthol concentration series (A) and menthol:water lick ratios (B) for individual TRPA1^HET (n = 12) and simultaneously run wild-type (n = 12; B6129 line) mice. The color-coded (legend, panel B) cross associated with each distribution gives its median (horizontal bar) and 95% CI* (vertical bar). TRPA1^HET mice made fewer licks to 0.3 mM menthol compared to controls (Holm-adjusted Wilcoxon test, p = 0.046; denoted by asterisk in panel A), albeit no other line differences were detected.

Temporal analysis of menthol licking behavior. Plots track for each mouse line their median lick ratio for the three highest concentrations of menthol as these ratios evolved over the first 10 consecutive trials of test sessions. For each trial of a given menthol concentration, lick ratios for individual mice were calculated as: , where i represents the trial number (1–10). Sample sizes were as follows: B6, n = 20; B6129, n = 17; TRPM8^KO, n = 10; TRPA1^KO, n = 15; TRPA1^HET, n = 12. Although infrequent, a few mice did not respond on the first trial of each menthol concentration, and one B6129 mouse did not respond on the second test trial with these stimuli. These animals were not included in median calculations for trials 1 and 2.

Distributions of lick counts to the menthol concentration series (A) and menthol:water lick ratios (B) for individual TRPM8^KO (n = 10) and TRPA1^KO (n = 15) mice. The color-coded (legend, panel B) cross associated with each distribution gives its median (horizontal bar) and 95% CI* (vertical bar). Data points are re-plotted from Figures 5, 6; note that 95% CI*s do not perfectly match between figures due to re-bootstrapping. As denoted by asterisks in panel A, TRPA1^KO mice made more licks to 0.3, 1, 1.5, and 2.3 mM menthol compared to TRPM8^KO mice (p < 0.03).