Thermal behaviors in mice using a Thermal Gradient Ring system
We analyzed thermal behaviors in mice using a Thermal Gradient Ring system in which the ring floor was divided into 24 zones from cold to hot, with the same temperatures occurring in two zones, allowing mice to freely move based on their temperature preference (Fig. 1A). We first set temperatures from 10 ℃ to 45 ℃ in a room with an ambient temperature of 23 ℃ to 24 ℃. Each floor zone had a Δ2.9 ℃-gradient from 11.5 ℃ to 43.6 ℃. We placed mice into this system for 60 min and continuously observed mouse movements. When there was no thermal gradient (i.e., when the floor temperature was the same as room temperature), WT mice moved equally between zones. On the other hand, when there was a temperature gradient, WT mice were more likely to stay at specific temperature zones, indicating a preference (Fig. 1B). Interestingly, WT mice did not show obvious temperature preferences in the first 20 min of the experiment, likely because the mice were first exploring their preferred temperatures. Eventually, the mice chose specific temperatures to stay at for the last 20 min of the experiment, and the preferred temperature was around 35 ℃ (Fig. 1C).
Mice are nocturnal animals and move more at night, at which time they also have higher body temperatures. Therefore, to examine the effects of body temperature on thermal behaviors, we put a small logger to measure the mouse’s body temperature [25]. The loggers were placed subcutaneously in the back or in the intraperitoneal space (core body temperature). We observed circadian changes in both the subcutaneous and core temperatures, where temperatures were high at night and lower during the day, as previously reported (Additional file 1: Fig. S1A). In addition, subcutaneous temperatures were approximately 2 ℃ lower than core body temperatures. To examine the relationship between body temperature and behaviors in the thermal gradient system, WT mice were first placed on the Thermal Gradient Ring without a thermal gradient. Subcutaneous and core temperatures were measured under this condition for 30 min. Then, the thermal gradient system was turned on (10–45 ℃), and was stabilized within 30 min, at which point we measured the mouse’s movements. Both the subcutaneous and core temperatures became elevated when mice were placed on the ring apparatus, likely because they began exploring, and they were reduced across time (Additional file 1: Fig. S1B). During the daytime (13:00 to 18:00) measurements, mice were placed on a ring, where the temperatures were already set. Both subcutaneous and core temperatures were similarly elevated and reduced (Fig. 1D). Changes in both the subcutaneous and core temperatures were essentially the same at nighttime (21:00-) as in the daytime, except that both temperatures were slightly higher before the behavior assay (Additional file 1: Fig. S1C).
Because both the subcutaneous and core temperatures changed similarly and because the temperature logger measurement in the intraperitoneal space was not affected by the ring floor temperature, we compared only the subcutaneous body temperatures between WT mice and mice lacking TRP channel genes. Subcutaneous body temperatures did not differ significantly between mice examined in the daytime (Fig. 1E) and nighttime (Additional file 1: Fig. S1D). Based on these results, we concluded that initial body temperatures do not affect body temperature in the thermal behavior assay, and thus decided to conduct all of the following experiments in the daytime (13:00 to 18:00).
Mice lacking thermosensitive TRP channels showed various temperature preference and avoidance behaviors
There are 11 thermosensitive TRP channels, each with a distinct temperature threshold for activation. [5] However, most thermal behavior data have been reported using the two temperature plate choice assay or the linear thermal gradient assay [10, 11], which might not be suitable for the analysis of temperature preference or avoidance. Therefore, in this study, we analyzed thermal behaviors across a 60-min duration in mice lacking Trpv1, Trpv3, Trpv4, Trpm2, Trpm8, or TrpA1 using the Thermal Gradient Ring system with different behavioral parameters. When we analyzed "Spent time" on the floor with no thermal gradient, all genotypes visited all zones equally and showed no spatial preference or avoidance (Additional Fig. 2A). When we analyzed "Spent time" on the floor with a temperature gradient from 11.5 ℃ to 43.6 ℃, WT mice stayed longer in the 34.8 ℃ zone, as did TRPA1−/− or TRPM2 −/− mice (Figs. 1 and 2A). Although TRPV4−/− mice have been reported to stay longer in higher temperature zones compared with WT mice over a 2 h period in a linear thermal gradient assay [13], TRPV4−/− mice in our assay stayed primarily in the lower temperature zones (31.9 ℃) compared to WT mice. TRPM8−/− and TRPV3−/− mice stayed longer in the lower temperature zones, as previously reported [10, 11, 14]. Only TRPM8−/− mice showed obvious differences in the time spent at the 11.5 ℃ to 17.3 ℃ zone (Fig. 2B, Additional file 2: Fig. S2B, C), which was expected as TRPM8−/− mice do not express the TRPM8 cold sensor, and, therefore, readily enter the cold zones. We did not observe any difference in temperature preference between WT and TRPA1−/− mice.
We next analyzed the transition of peak temperatures across the 60-min experiment, assessing which temperatures the mice stayed at every 20 min (Fig. 2B, Additional file 2: Fig. S2B). The mice were divided into two groups: one group (TRPM2−/−, TRPA1−/−) gradually moved to the warm zone over time, similar to the WT mice that stayed longer in the zone of around 34 ℃, and the other group (TRPV1−/−, TRPV3−/−, TRPV4−/−, TRPM8−/−) continuously stayed at about 30 ℃. To examine the difference among WT, TRPA1−/−, and TRPM2−/− mice, we extracted data during the 40–60 min window, when mice tended to settle at specific temperatures (Fig. 2C). TRPM2−/− mice stayed in wider temperature zones and did not have a specific peak. TRPA1−/− mice showed the same temperature preference as WT mice in this time period. Therefore, we increased the temperature resolution by changing the floor temperature range to 19.7 ℃ to 35.3 ℃ (this was the 19–36 ℃ temperature setting) (Fig. 2D, Additional file 3: Fig. S3). Under this setting, TRPA1−/− mice showed clearer temperature preferences, with a peak of 33.9 ℃ compared with WT mice, and this temperature preference of TRPA1−/− mice was observed as early as in the 40–60 min period (Additional file 3: Fig. S3C).
Analysis of other parameters: “Travel distance”, “Moving speed,” and “Acceleration”
Next, we analyzed other parameters (Travel distance and Moving speed) in each temperature zone using the 11.5–43.6 ℃ gradient setting. Travel distance was shorter at the low temperatures for all mice except for TRPM8−/− mice, which had a high Travel distance even in the 11.5 ℃ zone (Fig. 3A), likely because they do not sense cold temperatures. We also analyzed Travel distance every 20 min (Fig. 3B). The mice could be divided into two groups: one group (TRPM2−/−, TRPA1−/−) showed reduced Travel distance across time, which was similar to WT mice, and the other (TRPV1−/−, TRPV3−/−, TRPV4−/−, TRPM8−/−) traveled further than WT mice and kept travelling throughout the entire 60-min experiment. Interestingly, this grouping was the same as the one we found for the Spent time analysis (Fig. 2B). This feature was consistent even without the presence of a thermal gradient (Additional file 4: Fig. S4A, B), suggesting that this phenomenon was not related to the floor temperature gradient.
We also analyzed Moving speed in each zone, because mice were expected to move faster in the temperature zones that they want to avoid. Because mice showed similar Spent time at temperature zones less than 20.2 ℃ or over 40.7 ℃, it is unclear whether mice felt uncomfortable when they were in these temperatures (Fig. 2A). We found that TRPM8−/− mice had a consistent Moving speed, even at the coldest zone of 11.5 ℃, while the other genotypes, including WT mice, showed a higher Moving speed in the coldest zone (Fig. 3C, Additional file 5: Fig. S5C). This interpretation could be justified by the result that mice moved in all zones at a speed of less than 0.1 m/s (0.068 m/s) when there is no thermal gradient on the floor (Additional file 5: Fig. S5A). This indicated that mice felt comfortable when moving at a speed of less than 0.1 m/s, and did not change over time (Additional file 5: Fig. S5B). WT, TRPA1−/−, and TRPM2−/− mice moved at a slower speed than the other genotypes in the 34.8 ℃ zone, which was consistent with the data that WT, TRPA1−/−, and TRPM2−/− mice stayed longest in the temperature zone (Additional file 2: Fig. S2B).
Moving speed could be a factor of Travel distance, possibly revealing that mice that travelled a lot move faster compared to mice that travel less. Therefore, to exclude any effects by Travel distance we analyzed the mean of Acceleration to find the temperature zone at which mice changed their Moving speed. There was no difference in mean of Acceleration between WT mice and mice lacking TRP channels in the warm temperature zones. All mice showed an increase in Acceleration in the temperature zones of 11.5 ℃ and 14.4 ℃, except for TRPM8−/− mice, which did not accelerate in any zone (Fig. 4A). We also analyzed the real tracking data (black lines) with Acceleration (purple lines) across all temperature zones. Since the Thermal Gradient Ring is a circular device, Acceleration can be calculated as both positive and negative values depending on whether the mouse is running or turning back, respectively. Changes in moving was more clearly recognized when we plotted the temperature at which mice stayed and Acceleration as the y axis and time (0–60 min) as the x-axis (Fig. 4B). WT, TRPA1−/−, and TRPM2−/− mice stopped moving at temperatures around 35 ℃ in the midpoint of the analysis (Fig. 4C), while TRPM8−/−, TRPV1−/−, TRPV3−/−, and TRPV4−/− mice continued to move.
Analysis in the cold and hot temperature zones
TRPA1 was initially reported as a cold sensor under 17 ℃, [17] although there is controversy surrounding this [6, 16]. Therefore, we examined mouse behaviors under even colder conditions (7.8–47.2 ℃, which were achieved using the temperature setting of 5.0–54.9 ℃). However, WT and TRPA1−/− mice did not stay long at temperature zones lower than 11.4 ℃ in the first 20 min, and there was no difference in Spent time and Moving speed under this gradient setting (Fig. 5A, B). WT mice still entered the temperature zones lower than 11.4 ℃, although Spent time was decreased in these zones in the last 20 min (40–60 min). On the other hand, TRPA1−/− mice did not enter these cold zones during the last 20 min. Consistent with the Spent time results, WT mice showed increased Moving speed both in the first and last 20 min, which was similar to TRPA1−/− mice. Recently, TRPM3/TRPV1/TRPA1-triple knockout mice were reported to show an absence of responses to noxious heat stimuli [20]. Therefore, we examined mouse behaviors in conditions over 40 ℃, with the thermal gradient being 14.6 ℃ to 54.9 ℃ (achieved by setting the temperature from 10.0 to 60.0 ℃). TRPV1 is a sensor for noxious heat, and TRPV1−/− mice are known to be insensitive to temperatures over 52 ℃ on a hot plate test [15]. WT mice did not enter temperature zones over 50 ℃, although both TRPA1−/− and TRPV1−/− mice stayed in the high temperature zones in the first 20 min (Fig. 5C). However, TRPA1−/− mice did not enter this zone in the last 20 min, suggesting that TRPV1−/−, but not TRPA1−/− tolerate the noxious high temperatures. These phenotypes were also recognized by the slower Moving speed of TRPV1−/− mice than WT mice in the high temperature zones (Fig. 5D).