Running induces nausea in rats: Kaolin intake generated by voluntary and forced wheel running

Appetite 105 (2016) 85e94

Contents lists available at ScienceDirect

Appetite journal homepage: www.elsevier.com/locate/appet

Running induces nausea in rats: Kaolin intake generated by voluntary and forced wheel running Sadahiko Nakajima Department of Psychological Science, Kwansei Gakuin University, Nishinomiya, 662-8501, Japan

a r t i c l e i n f o

a b s t r a c t

Article history: Received 19 March 2016 Received in revised form 7 May 2016 Accepted 10 May 2016 Available online 15 May 2016

Three experiments were conducted showing rats’ pica behavior (kaolin clay intake) due to running in activity wheels. The amount of kaolin consumed was a positive function of the available time of voluntary running (20, 40, or 60 min), although this relationship was blunted by a descending (i.e., 60 / 40 / 20 min) test series of execution (Experiment 1). Pica was also generated by forced running in a motorized wheel for 60 min as a positive function of the speed of wheel rotations at 98, 185, or 365 m/ h, independent of the order of execution (Experiment 2). Voluntary running generated more pica than did forced running at 80 m/h, although the distance travelled in the former condition was 27% lesser than that in the latter condition (Experiment 3). Because kaolin intake is regarded as a reliable measure of nausea in rats, these results show that wheel running, either voluntary or forced, induces nausea in rats. © 2016 Elsevier Ltd. All rights reserved.

Keywords: Pica Kaolin Wheel running Rats

1. Introduction Rats engage in aberrant pica behavior (kaolin clay ingestion) after various nausea-inducing treatments, including irradiation (Yamamoto, Asano, Matsukawa, Imaizumi, & Yamatodani, 2011; Yamamoto, Takeda, & Yamatodani, 2002), motion sickness (McCaffrey, 1985; Mitchell, Krusemark, & Hafner, 1977; Mitchell, Laycock, & Stephens, 1977; Morita, Takeda, Kubo, & Matsunaga, 1988; Morita, Takeda, Kubo, Yamatodani, et al., 1988; Takeda et al., 1995a), and administration of emetogenic drugs such as lithium chloride (LiCl: Mitchell et al., 1976; Watson & Leitner, 1988; Yamamoto, Ngan, Takeda, Yamatodani, & Rudd, 2004), cyclophosphamide (Mitchell et al., 1976; Tohei, Kojima, Ikeda, Hokao, & Shinoda, 2011; Yamamoto, Nakai, Nohara, & Yamatodani, 2007; Yamamoto et al., 2011), cisplatin (De Jonghe & Horn, 2008; De Jonghe, Lawler, Horn, & Tordoff, 2009; Han et al., 2014; Horn, De Jonghe, Matyas, & Norgren, 2009; Liu, Malik, Sanger, Friedman, & Andrews, 2005; Malik, Liu, Cole, Sanger, & Andrews, 2007; Rudd, Yamamoto, Yamatodani, & Takeda, 2002; Saeki et al., 2001; Sharma, Gupta, Kochupillai, Seth, & Gupta, 1997; Takeda et al., 1995b; Takeda, Hasegawa, Morita, & Matsunaga, 1993; Yamamoto et al., 2007, 2011, 2014), morphine (Aung, Mehendale, Xie, Moss, & Yuan, 2004), apomorphine (De Jonghe & Horn, 2008; Takeda et al., 1995a, 1993), nicotine (Yamamoto et al., 2004), copper

E-mail address: [email protected]. http://dx.doi.org/10.1016/j.appet.2016.05.009 0195-6663/© 2016 Elsevier Ltd. All rights reserved.

sulfate (Takeda et al., 1993; Yamamoto et al., 2004), ritonavir (Aung et al., 2005; Yuan et al., 2009), 2-deoxy-D-glucose (Watson & Leitner, 1988; Watson et al., 1987), cholecystokinin octapeptide (McCutcheon, Ballard, & McCaffrey, 1992), actinomycin D (Yamamoto et al., 2007), 5-fluorouracil (Yamamoto et al., 2007), and intragastric ethanol ingestion (Constancio, Pereira-Derderian, Menani, & De Luca, 2011). Pica generated by the above treatments can be attenuated by administering anti-emetic drugs (Takeda et al., 1995a, 1995b, 1993; Yamamoto et al., 2011, 2014, 2002; Aung et al., 2005, 2004; Han et al., 2014; Malik et al., 2007; Morita, Takeda, Kubo, Yamatodani, et al., 1988; Rudd et al., 2002; Saeki et al., 2001; Sharma et al., 1997; Tohei et al., 2011; Yuan et al., 2009). These studies strongly suggest that pica behavior is a good index of nausea (or gastrointestinal discomfort) in rats that cannot vomit because of anatomical and/or neural reasons (Horn et al., 2013). Recent research from my laboratory (Nakajima & Katayama, 2014) demonstrated that pica behavior is also generated by voluntary running in a closed activity wheel, implying that running induces nausea in rats. Notably, other sources of information support this reasoning. First, running in an activity wheel establishes Pavlovian conditioned taste aversion (CTA) in rats to a tastant consumed shortly before running (e.g. Heth, Inglis, Russell, & Pierce, 2001; Lett & Grant, 1996; Lett, Grant, & Gaborko, 1998; Nakajima, Hayashi, & Kato, 2000; see Boakes & Nakajima, 2009; for a review), and this running-based CTA is prevented by administration of the anti-emetic granisetron (Eccles, Kim, & O’Hare,

86

S. Nakajima / Appetite 105 (2016) 85e94

2005), supporting the hypothesis that nausea is the underlying physiological factor for establishing running-based CTA. Second, running-based CTA is alleviated not only by preexposure to running but also by prior injection of LiCl (Nakajima, Urata, & Ogawa, 2006), implying that a common process (presumably nausea) is physiologically habituated by preexposure. Third, a unique reduction in taste palatability, measured by the microstructure of rats’ licking, accompanies running- and LiCl-based CTAs, indicating that running- and LiCl-based CTAs are commonly caused by nausea (Dwyer, Boakes, & Hayward, 2008). Although these findings converge with the report of runningbased pica behavior (Nakajima & Katayama, 2014), demonstration of increased kaolin consumption by voluntary running is novel, and calls for replication, which is provided in Experiment 1 of the present research. The major aim of Experiment 2 was to test the generality of finding, by using forced, rather than voluntary, running as an agent to evoke pica behavior. In these explorations, I manipulated the available time for voluntary running (Experiment 1) and the speed of forced running (Experiment 2) as critical independent variables. Finally, Experiment 3 directly compared the sizes of pica generated by voluntary and forced running treatments. In carrying out the research, Experiments 1 and 2 were conducted concurrently, while Experiment 3 was conducted after these experiments with half of the subjects of these experiments. 2. Experiment 1 In our previous research, a 60-min confinement in an unlocked wheel was employed for 4 successive days to demonstrate rats’ pica behavior in their home cages (Nakajima & Katayama, 2014). It has been shown that the amount of pica positively depends on the dose of emetic treatments (Mitchell et al., 1976; Mitchell, Krusemark, et al., 1977; Yamamoto et al., 2002, 2004, 2007) and that the amount of running-based CTA is a positive function of length of available time for running (Hayashi, Nakajima, Urushihara, & Imada, 2002; Masaki & Nakajima, 2006). Therefore, I expected that the pica behavior would increase with the length of available time for running within a reasonable range. In this experiment, the available time was 20, 40, or 60 min, tested in an ascending or descending series of time length. 2.1. Method 2.1.1. Subjects and apparatus Eight experimentally naïve male rats (Slc: Wistar/ST) were housed in individual wire home cages (20 cm wide, 25 cm long, and 18.7 cm high) in a vivarium on a 16:8-h light-dark cycle (lights on at 0800 h) at 23
C and 55% humidity. The animals were 9 weeks old on the first day of this experiment, and they were maintained with food pellets, tap water, and kaolin pellets available ad libitum throughout the experiment. The food pellets (MF diet; Oriental Yeast Co., Tokyo, Japan) were placed in a stainless container (7.5 cm wide, 4.5 cm long, 15 cm deep) positioned inwards with its end apertures 3.5 cm above the cage floor. The tap water was accessible from a stainless needle-pin nozzle protruding through a hole in the center of the back wall of each cage. The kaolin pellets were made of kaolin powder (Shin Nihon Zokei Co., Tokyo, Japan) and gum arabic (Holbein Works, Ltd., Osaka, Japan) at a 99:1 (w/w) ratio; they were mixed with tap water to form cylindrical pellets and were completely dried at room temperature. Each day, three or four kaolin pellets (about 20e25 g in total) were presented to each rat in a stainless steel bowl (8 cm in diameter and 3.5 cm deep) clipped to the cage wall at floor level with an iron hoop holder. A plastic tray (22.5 cm wide, 32 cm long, and 5.5 cm deep) with paper bedding was positioned 10 cm below

each cage to collect excreta, food shatters, and kaolin splinters. Crushed kaolin and food in the tray were collected with a spoon and chopsticks, dried for a day, segregated, and weighed to obtain correct amounts of kaolin and food intake. The rats were transferred, by a carrying cart having individual compartments, to a conventionally illuminated experimental room nearby, which had 8 hand-made activity wheels hung on a wire net arranged in a 4  2 fashion. The top and bottom rows were 140 and 90 cm above the room floor, respectively, and a long acrylic plate was fixed just below each row to catch excretions. Each wheel had an internal width of 15 cm and a diameter of 30 cm. The sides of the wheel were perforated metal sheets, and the running surface was made of 0.2-cm metal rods spaced 1 cm apart. The wheels could be tuned in both directions. The minimum torque to initiate the movement when the forepaws of animal were 10 cm from the lowest point of the unlocked wheel was around 25 cN measured by a Correx tension gauge (Haag-Streit A.G., Koeniz, Switzerland). A full turn of each wheel was counted automatically by a handcrafted system consisting of a small magnet on the outer rim of the wheel, a reed switch, and an electric pedometer fixed on the wire net. Each wheel could be locked by two plastic tied laundry pinches. 2.1.2. Procedure At 1030 h of each day, all rats were weighed with an electric balance (KS-251, Dretec Co., Koshigaya, Japan) to the nearest 1 g and then moved to the individual compartments of the cart. On the initial 5 baseline days, the cart was kept in the vivarium for 60 min. On the next 4 days, animals were transferred to the experimental room, where half of the rats (Group 1) were allowed to run in the unlocked wheels for 60 min. For the other half of the rats (Group 2), the wheels were locked by the pinches 20 min after the onset of the running session and they were detained in the stationary wheels for the remaining 40 min. This procedure equated the durations in the experimental room between the two groups, but it inevitably resulted in the situation that the groups differed in the time elapsed between the end of running and returning to the home cages (i.e., the opportunity for kaolin consumption). Notably, this confounding factor seems unimportant, since in our observation the rats did not rush to consume kaolin after returning to the home cages. After the second baseline treatment for 4 days, the running treatment was reinstated for 4 days with the available running time being 40 min for all rats; the animals were detained in the remaining 20 min in the locked wheels. After the third baseline treatment of 4 days, the final running phase was executed for 4 days: the length of available running time was 20 min (i.e., 40-min post-running detention) for Group 1 or 60 min for Group 2. Accordingly, the three running conditions (20, 40, and 60 min) were implemented in the descending series for Group 1, while they were in the ascending series for Group 2. The experiment ended with 2 baseline days. 2.1.3. Measurement The amounts of food and kaolin consumed in the home cages (i.e., 23-h intakes) were recorded every day by removing the food and kaolin containers, immediately after the rats were moved to the individual compartments of the carrying cart. The containers were weighed with an electric balance (BJ-1500, Sartorius, K.K., Tokyo, Japan) to the nearest 0.1 g, refilled, and replaced at 1130 h, immediately before the rats were returned to the home cages. The experimental protocol was administered by laboratory assistants who were unaware of the purpose of the research. 2.1.4. Analysis A paired-t test or an analysis of variance (ANOVA) was applied to each data set of interest. Pooled error term was used in subsequent

S. Nakajima / Appetite 105 (2016) 85e94

simple main effect analyses of any significant interaction or Ryan’s multiple comparisons with the t-statistics. For simplicity, test statistics will be reported only when they are statistically significant with an alpha (Type 1 error) set a priori at 0.05. Average scores will be reported as arithmetic means with standard errors across subjects.

3. Results and discussion 3.1. Kaolin intakes Wheel running generated pica behavior, as summarized in the leftmost pair of bars in Fig. 1. The kaolin intake of 12 running days (4 days  3 running phases) was significantly greater than that of 3 immediately preceding baseline days (1 day each for the forthcoming phases), paired t (7) ¼ 4.44, p ¼ 0.003. Furthermore, the top panel of Fig. 2 shows that the amount of pica is a positive function of the length of availability of running in activity wheels. Because the order effect (i.e., ascending or descending series concerning the length of running availability) is obvious as depicted in this panel, this group factor was also included in the statistical analyses. A 2 (group)  3 (running time)  4 (day) ANOVA, applied to the data set shown in this panel, yielded a significant main effects of running time, F (2, 12) ¼ 7.15, p ¼ 0.009, supporting the aforementioned positive relationship between the available running time and the amount of kaolin intake. The main effect of day, F (3, 18) ¼ 8.82, p < 0.001, and the time  day interaction, F (6, 36) ¼ 3.50, p ¼ 0.008, were also significant. Subsequent analyses of the time  day interaction revealed that the effect of time was significant on the second, third, and fourth days of the running phase: F (2, 48) ¼ 3.64, p ¼ 0.034; F (2, 48) ¼ 6.92, p ¼ 0.002; F (2, 48) ¼ 10.41, p < 0.001, respectively. Multiple comparison analyses of the time factor of the second day revealed that 60-min running generated significantly greater kaolin intake than the 40-min running, Ryan’s t (48) ¼ 2.65, p ¼ 0.011; the difference between the other two comparisons were not significant. On the third day, 60-min running generated significantly greater kaolin intake than the 20- and 40-min running, ts (48) > 2.81, ps < 0.007, which did not differ from each other. On the fourth day, 20-min running generated significantly less kaolin intake than the 40- and 60-min running, Ryan’s ts (48) > 3.16, ps < 0.003, which did not differ from each other. The original 2  3  4 ANOVA also yielded a significant group  time

87

interaction, F (2, 12) ¼ 13.04, p ¼ 0.001. Simple main effect analyses of this interaction revealed that the effect of running time was significant only in the ascending group, F (2, 12) ¼ 19.61, p < 0.001. Multiple comparison analyses of the time factor of the ascending group revealed that 60-min running generated significantly greater kaolin intake than the 20- and 40-min running, Ryan’s ts (12) > 4.38, ps < 0.001. Wheel running had a carryover effect on the post-running baseline for 2 days: a 2 (group)  3 (running time of the preceding phase)  2 (day) ANOVA, yielded a significant main effect of running time, F (2, 12) ¼ 7.39, p ¼ 0.008, and the rats consumed significantly more kaolin after 40-min (3.19 ± 0.56 g) or 60-min (2.78 ± 0.36 g) running phase than after 20-min running phase (1.59 ± 0.36 g), Ryan’s ts (12) > 2.74, ps < 0.018. 3.2. Food intakes In contrast to our previous report (Nakajima & Katayama, 2014), the food intake was not affected by wheel running: the average collapsed over 3 pre-running baseline days was 19.54 ± 0.27 g, and the average collapsed over 12 running days (4 days  3 phases) was 19.10 ± 0.29 g. 3.3. Wheel turns The bottom panel of Fig. 2 shows the mean number of wheel turns, separately shown for the two order-different groups of rats. Although the ascending group of rats clearly increased the number of turns as a function of days of available running time, this trend was small in the descending group of rats. A 2 (group)  3 (running time)  4 (day) ANOVA, applied to the data set shown in this panel, yielded a significant main effect of time, F (2, 12) ¼ 14.83, p < 0.001 and its interaction with group, F (2, 12) ¼ 5.04, p ¼ 0.026. Subsequent analyses of this interaction revealed that the effect of group was significant only at the 60-min running, F (1, 18) ¼ 7.32, p ¼ 0.015. In addition, the effect of time was significant only in the ascending group F (2, 12) ¼ 18.16, p < 0.001, and all comparisons among the 20-, 40-, and 60-min conditions of this group were significant, Ryan’s ts (12) > 2.82, ps < 0.016. Fig. 3 depicts the relationship between the number of wheel turns and the kaolin intake for each rat: each data point represents the score averaged over 4 running days. In the ascending group of rats (A1e4), greater the running, the larger the kaolin intake. However, the trend was mixed in the descending rats (D1e4). Accordingly, the overall relationship between the amount of running and the kaolin in take was modest. 3.4. Body weights The rats’ body weights were unaffected by wheel running. The average weight increased throughout the experimental period from 300.3 ± 3.0 to 389.1 ± 7.5 g, reflecting the rats’ growth. 4. Experiment 2

Fig. 1. Effect of wheel running on the kaolin consumption. Each pair of bars illustrates the average of all running days and the corresponding score of immediately preceding baseline days. For Experiments 1 and 2, the running data are based on 12 days (4 days  3 phases), while the baseline data are on 3 days (1 day  3 phases). For Experiment 3, the running data are based on 4 days, while the baseline data are on 1 day. The error bars indicate the standard errors across subjects. V ¼ voluntary running, F ¼ forced running.

Taste aversion can be established not only by voluntary running in an unlocked wheel, but also by forced running in a motorized wheel (Eccles et al., 2005; Forristall, Hookey, & Grant, 2007; Masaki & Nakajima, 2006). According to Forristall et al., running-based CTA has two underlying physiological processes: (1) motion sickness (nausea) induced by the back-and-forth “rocking” movements of the wheel, and (2) mesolimbic dopamine activation induced by the action of running itself. They have argued that forced running produces taste avoidance but not taste disgust, because it lacks the rocking movements of the wheel, which induce nausea. This

88

S. Nakajima / Appetite 105 (2016) 85e94

Fig. 2. Mean amount of kaolin intake (top panel) and mean number of wheel turns (bottom panel) in Experiment 1, separately shown for the ascending (20 / 40 / 60 min) and descending (60 / 40 / 20 min) running groups of rats (each n ¼ 4), as a function of days of available time for voluntary running. The error bars indicate the standard errors across subjects (Some error bars are smaller than the symbols). The dotted horizontal line in the top panel represents the baseline kaolin intake shown in Fig. 1.

argument has been corroborated by Grant et al. (2012), who carefully manipulated the rocking movements of an unlocked wheel and found that conditioned disgust reactions, indicated by rejective reactions of “gaping”, were reduced by stopping the rocking movements. Their argument, however, is incongruent with the study demonstrating alleviation of running-based CTA by an antiemetic drug (Eccles et al., 2005), because the CTA of this study was established by forced running in a motorized wheel. Experiment 2 was designed to test whether forced running in a motorized wheel yields pica behavior, a sign of nausea. The speed of the motorized wheel was changed across conditions in an ascending or descending order. 4.1. Method

Fig. 3. Relationship between the number of wheel turns (i.e., the amount of voluntary running) and the kaolin intake in Experiment 1. The scores are averages over the 4 running days of each phase: the data points represent the individual rats.

4.1.1. Subjects and apparatus A new batch of 8 experimentally naïve male rats of the same strain and age were maintained as in Experiment 1. They were transferred by a carrying cart having individual compartments to a conventionally illuminated experimental room nearby, which had 4 commercial wheels with side cages (FRW-30, Melquest, Toyama, Japan) on a table. Access to the side cage of each wheel was blocked by a plate, so that only the wheel compartment was used. Each wheel had an internal width of 9 cm and a diameter of 32 cm. The sides of the wheel were clear acrylic plates, and the running surface

S. Nakajima / Appetite 105 (2016) 85e94

89

was made of 0.5-cm metal rods spaced 1.1 cm apart. A full turn of each wheel was counted automatically. Each wheel could be rotated in one direction by an electric motor, and the constant speed of wheel turns (98, 185, or 365 m/h, 1 m ¼ 1 turn) was manually set by the dial knob of the controller. Notably, application of the break-and-run patterns of the rats of Experiment 1 to the animals of this experiment by yoking procedure is impractical, because rats cannot go along with such unpredictable movements of wheels. Consequently, the constant movements were employed in this experiments. It is also worth to remark that even at the highspeed condition (365 m/h) rats walk slowly, but not trot or gallop (Gillis & Biewener, 2001). The term “running” was used here and onward simply for consistency of wording in the study. 4.1.2. Procedure Daily procedures were executed in two squads of 4 rats each: the starting time of the first squad was 1315 h and that of the second squad was 1430 h. As in Experiment 1, rats were weighed and then moved to the individual compartments of the cart. On the initial 5 baseline days, the cart was kept in the vivarium for 60 min. On the next 4 days, rats were transferred to the experimental room, where they were confined in the disengaged activity wheels, and then the dial of the motor was set to initiate rotation. Each session started and ended by snapping the switch of the wheel motor. This phase was followed by two additional cycles of a 4-day baseline and a 4-day running treatment, and the test sequence was ended with a 2-day baseline. The speed of rotation was increased across the 3 running phases (98, 185, and 365 m/h) for the first squad of rats (ascending rats), while it was decreased (365, 185, and 98 m/h) for the second squad (the descending rats). 4.1.3. Measurement and analysis The amounts of food and kaolin consumed in the 23-h period were recorded every day as in Experiment 1 by removing the pellet containers immediately after the rats were moved to the individual compartments of the carrying cart, and replaced them immediately before the rats were returned to the home cages. The statistical procedures were the same as in Experiment 1. One of the rat was dropped from the experiment, because it was injured during wheel running, resulting in 4 rats for the ascending group and 3 rats for the descending group. 4.2. Results and discussion 4.2.1. Kaolin intakes Contrary to the claim of Forristall et al. (2007), forced running appears to induce nausea, because it generated pica behavior in the rats of Experiment 2, as shown in the second pair of bars in Fig. 1. Namely, the kaolin intake was significantly increased by forced running, paired t (6) ¼ 4.34, p ¼ 0.005. The present finding accords with the report that anti-emetic drug alleviates CTA caused by forced running (Eccles et al., 2005). Fig. 4 illustrates the mean kaolin intake, separately shown for the two order-different groups of rats, as a function of days of given running speed. The order effect is negligible, and the high speed running clearly resulted in large kaolin intake. A 2 (group)  3 (running speed)  4 (day) ANOVA yielded significant main effects of speed, F (2, 10) ¼ 9.85, p ¼ 0.004, and day, F (3, 15) ¼ 10.21, p < 0.001, and their interaction, F (6, 30) ¼ 6.17, p < 0.001. Subsequent analyses of the speed  day interaction revealed that the effect of speed was significant on the second, third, and fourth days of the running phase: F (2, 40) ¼ 5.88, p ¼ 0.005; F (2, 40) ¼ 12.23, p < 0.001; F (2, 40) ¼ 15.26, p < 0.001, respectively. Multiple comparison analyses of the speed factor of these days showed that the high-speed running generated significantly greater kaolin

Fig. 4. Mean amount of kaolin intake in Experiment 2, separately shown for the ascending (98 / 185 / 365 m/h) running group of rats (n ¼ 4) and the descending (365 / 185 / 98 m/h) running groups of rats (n ¼ 3), as a function of days of forced running. The error bars indicate the standard errors across subjects (Some error bars are smaller than the symbols). The dotted horizontal line represents the baseline kaolin intake shown in Fig. 1.

intake than the middle-speed running, Ryan’s ts (40) > 2.28, ps < 0.028, or the slow-speed running, Ryan’s ts (40) > 3.39, ps < 0.002. The simple main effect of day was significant only for the high-speed running condition, F (3, 45) ¼ 19.99, p < 0.001. The carryover effect, noted in Experiment 1, was also evident in Experiment 2, when we devote attention to the two post-running baseline days: a 2 (group)  3 (running speed of the preceding phase)  2 (day) ANOVA, yielded a significant main effect of speed, F (2, 10) ¼ 4.93, p ¼ 0.032. The averages collapsed over the days were 0.93 ± 0.25, 2.61 ± 0.63, and 2.15 ± 0.45 g, respectively, after the low-, middle-, high-speed running phases: the difference between the baselines after the low- and middle-speed was statistically significant, Ryan’s t (10) ¼ 3.13, p ¼ 0.011. 4.2.2. Food intakes As in the previous report (Nakajima & Katayama, 2014), wheel running slightly, but statistically significantly, decreased rats’ food intake, paired t (6) ¼ 5.64, p ¼ 0.001: the average collapsed across 3 pre-running baseline days was 20.09 ± 0.69 g, and the average collapsed over 12 running days (4 days  3 phases) was 19.51 ± 0.62 g. 4.2.3. Wheel turns The number of wheel turns largely differed across the conditions, because the speed of wheel rotations were manipulated with the running time being equal (60 min) across the conditions. Fig. 5 depicts the relationship between the number of wheel turns and the kaolin intake for each rat: each data point represents the score averaged over 4 running days. There were a little fluctuations even in the same condition, because the wheel motors were manually turned off one by one. In 5 out of the 7 rats, the kaolin intake is a clear positive function of the amount of running, while the data lines stayed almost flat for the remaining 2 (A4 and D2). 4.2.4. Body weights The rats’ body weights were unaffected by wheel running. The average weight increased throughout the experimental period from 303.7 ± 5.7 g to 385.0 ± 11.0 g, reflecting rats’ growth. 5. Experiment 3 Experiments 1 and 2 have shown that voluntary and forced running, respectively, generates pica behavior in rats, as measured

90

S. Nakajima / Appetite 105 (2016) 85e94

disengaged activity wheels. The first squad of rats (Group V-F) were allowed to voluntarily run for 60 min, while the dial of the motor was set to initiate rotation for the second squad of rats (Group F-V). This phase was were followed by a 4-day baseline and then a 4-day running treatment, where the natures of wheel running (voluntary vs. forced) were interchanged between the squads. The motor speed for forced running was originally planned to be determined by matching each rat under the forced running condition to its counterpart under the voluntary running condition, to equate the average speed of running (i.e., the number of wheel turns per hour). However, as the running speed of voluntary running was, in practice, less than the lower limit of the speed controller for the wheel motor, so the dial knob of the controller was set at the lower limit (80 m/h) for the forced running treatment. 5.1.3. Measurement and analysis The amounts of food and kaolin consumed in the home cages during the 23-h period were recorded every day as in Experiments 1 and 2. Fig. 5. Relationship between the number of wheel turns (i.e., the amount of forced running) and the kaolin intake in Experiment 2. The scores are averages over the 4 running days of each phase: the data points represent the individual rats.

by kaolin consumption in the home cages. Experiment 3 directly compared the size of pica behavior between these two running treatments, in both between-groups and within-subject experimental designs. 5.1. Method 5.1.1. Subjects and apparatus In order to amplify the sensitivity to detect effect of running on pica behavior, 4 good kaolin eaters were chosen from each of Experiments 1 and 2 (N ¼ 8 in total). The subjects were assigned to two squads of rats, matched with respect to amount of kaolin intake and treatment of the previous experiments. In this experiment, a hanging stainless container (8 cm wide, 4.5 cm long, 6 cm deep), instead of the bowl used in Experiments 1 and 2, was employed for presenting kaolin pellets for each rat. This modification was intended to obtain more precise kaolin consumption data by preventing hoarding and shattering of kaolin pellets. The container filled with 12e16 kaolin pellets (about 80e100 g in total) was installed in the home cage with its end apertures 10 cm above the cage floor. The wheels were identical to those of Experiment 2, but the motor was turned off for the voluntary running condition: the wheels could be turned in both directions, and the minimum torque to initiate the movement was around 50 cN when measured as in Experiment 1. For the forced running condition, the wheel rotated in one direction and the speed of wheel rotation of each session was set by a dial knob as in Experiment 2. 5.1.2. Procedure Experiment 3 started one month after the end of Experiments 1 and 2, which had been executed currently as aforementioned. Daily procedures were executed in two squads of 4 rats each as in Experiment 2: the starting time of the first squad was 1315 h and that of the second squad was 1430 h. As in Experiments 1 and 2, rats were weighed and then moved to the individual compartments of the cart. On the initial 4 baseline days, the cart was kept in the vivarium for 60 min. On the next 4 days, animals were transferred to the experimental room, where they were confined in the

5.2. Results and discussion 5.2.1. Kaolin intakes On average, voluntary and forced running generated pica behavior, and the amount of pica was greater for the voluntary running, as shown in the third and fourth pairs of bars in Fig. 1, paired t (7) ¼ 9.03, p < 0.001; paired t (7) ¼ 2.99, p ¼ 0.020, respectively. Readers might be surprised by the fact that statistically reliable pica was caused by the slow forced running of 80 m/h, because forced running at 98 m/h was not effective in generating reliable pica behavior in Experiment 2 (see Fig. 4). Selection of good reacting animals from the subjects of the preceding experiments is a most probable reason for this success. The top panel of Fig. 6 illustrates daily changes in the kaolin intakes along the time course of this experiment. The rats showed some pica behavior on the first day of the first baseline phase, reflecting spontaneous recovery of formerly habituated exploratory biting of novel objects: presenting kaolin pellets in an unfamiliar hanging container might have potentiated this recovery. On the second day and onward, the two groups of rats equally consumed very small amount of kaolin during the first baseline. A 2 (group)  4 (day) ANOVA, applied to the first baseline data, yielded a significant main effect of day, F (3, 18) ¼ 10.48, p < 0.001, reflecting the modest intake on the first day. In the first running phase, voluntary running provoked large kaolin intake, compared with the forced running, as indicated by the significant main effects of group, F (1, 6) ¼ 6.13, p ¼ 0.048, and day, F (3, 18) ¼ 11.46, p < 0.090, and their interaction, F (3, 18) ¼ 12.87, p < 0.001, of the 2  4 ANOVA. The simple group effect was significant on the second, third, and fourth day of this phase, Fs (1, 24) > 5.18, ps < 0.032; the simple day effect was significant only in the voluntary running rats (i.e., Group V-F), F (3, 18) ¼ 24.28, p < 0.001. Group V-F consumed slightly more kaolin than Group F-V in the second baseline, suggesting some carryover effect. This difference was statistically reliable: a 2 (group)  4 (day) ANOVA yielded a significant main effect of group, F (1, 6) ¼ 9.70, p ¼ 0.021. The difference between Groups V-F and F-V was reversed in the second running phase. The main group effect of a simple 2 (group)  4 (day) ANOVA, applied to the data of the second running phase, just missed the statistical significance, F (1, 6) ¼ 5.39, p ¼ 0.059, but the main effect of day was significant, F (3, 18) ¼ 7.24, P ¼ 0.002. Furthermore, if we include the second baseline data in the analysis as a 2 (group)  2 (phase: baseline

S. Nakajima / Appetite 105 (2016) 85e94

91

Fig. 6. Mean amount of kaolin intake (top panel) and mean number of wheel turns (bottom panel) across the days in Experiment 3. Group V-F (n ¼ 4) was allowed to run voluntarily for 1 h in the first running phase and forced to run at 80 m/h for 1 h in the second running phase. The order of these running phases was exchanged for Group F-V (n ¼ 4). The error bars indicate the standard errors across subjects (Some error bars are smaller than the symbols).

vs. running)  4 (day) ANOVA, the reversal of kaolin intake shown in the right two sections of the top panel of Fig. 6 was statistically supported by a significant main effect of phase, F (1, 6) ¼ 31.67, p ¼ 0.001, and more importantly its interaction with group, F (1, 6) ¼ 9.21, p ¼ 0.023. Subsequent analyses revealed that the two groups significantly differed in the second running phase, F (1, 12) ¼ 10.03, p ¼ 0.008, and that the phase effect was significant only in Group F-V, F (1, 6) ¼ 37.52, p < 0.001. The ANOVA noted above also found a significant interaction of phase  day, F (3, 18) ¼ 6.32, p ¼ 0.004, and subsequent analyses found a significant phase effect in every day, Fs (1, 24) > 7.43, ps < 0.012, and a significant day effect only in the second running phase, F (3, 36) ¼ 8.18, p < 0.001. The larger kaolin intake by voluntary running than forced running is also supported by a 2 (group)  2 (type of running: voluntary vs. forced)  4 (day) ANOVA, which yielded significant main effects of type, F (1, 6) ¼ 55.62, p < 0.001, and day, F (3, 18) ¼ 18.32, p < 0.001, and their interaction, F (3, 18) ¼ 10.73, P < 0.001. Subsequent analyses of this interaction supported the claim that the voluntary running evoked large kaolin ingestion, compared with the forced running on the second day and onward, Fs (1, 24) > 23.7, ps < 0.001, and that the increasing day effect was significant only under the voluntary running condition, F (3, 36) ¼ 26.33, p < 0.001. Notably, the group  type interaction of the three-factor ANOVA was also significant, F (1, 6) ¼ 15.03, p ¼ 0.008, suggesting some effect of execution of the two running treatments. This effect, however, was not very marked, because subsequent analyses found that voluntary running generated significantly more kaolin intake both in Group V-F, F (1, 6) ¼ 6.42, p ¼ 0.044, and in Group F-V, F (1, 6) ¼ 64.28, p < 0.001.

5.2.2. Food intakes Voluntary running slightly, but statistically significantly, decreased rats’ food intake, paired t (7) ¼ 2.89, p ¼ 0.023: the average over the pre-running baseline days was 21.01 ± 0.56 g, and the average over the 4 running days was 20.00 ± 0.58 g. On the other hand, forced running had no effect on the food intake, paired t < 1: 20.63 ± 0.44 and 20.74 ± 0.69 g, respectively, for the baseline and running days.

5.2.3. Wheel turns The bottom panel of Fig. 6 shows the mean number of wheel turns, separately shown for the two groups of rats. Because the speed of motor was set at 80 m/h and the running time was always 60 min, the score of the forced running condition was around 80 (1 turn ¼ 1 m) with a minimum variations due to short time lags across the wheels in turning off the motor switches. As the homoscedasticity assumption is not satisfied for the data summarized in this panel, so 2 (group)  4 (day) ANOVAs are inadequate here. Alternatively, a paired-t test was applied to the number of wheel turns averaged over days of each condition for the individual subjects, by collapsing the group factor, to compare the voluntary and forced running conditions. The average of the voluntary condition was 57.4 ± 8.9, and the corresponding score of the forced running was 78.3 ± 1.0, paired t (7) ¼ 2.36, p ¼ 0.050. In other words, the number of wheel rotations was 27% shorter in the voluntary running than the forced running. Taken together with the kaolin intake data noted above, this result means that voluntary running evokes strong pica behavior, compared with forced running at 80 m/h, despite the shorter distance travelled by the voluntary running.

92

S. Nakajima / Appetite 105 (2016) 85e94

Notably, the numbers of wheel turns recorded in this experiment were much smaller than those in Experiment 1. This is most probably due to the fact that the minimum torque to initiate the wheel movement was twice as much as that of Experiment 1 (see also the similar data in Experiment 2 of Nakajima, 2011). 5.2.4. Body weights The types of wheel running had a little effect on rats’ body weights. Because one of the rats of Group F-V was extraordinarily (100 g or 23%) heavier than the average of the other rats of this experiment, the homogeneity of variances between the two groups was not met. Therefore, the body weight of each day was transformed for each rat as a percentage with regard to the value of the first day of this experiment, before applying a 2 (group)  4 (day) ANOVAs to the data of each phase. Although almost all main or interactive effects were nonsignificant in this series of ANOVAs, the group  day interaction was significant in the first running phase, F (3, 18) ¼ 10.66, p < 0.001. Subsequent analyses of this interaction revealed that Group V-F slightly decreased their weight in this phase, F (3, 18) ¼ 3.97, p ¼ 0.025, while Group F-V slightly increased their weight, F (3, 18) ¼ 7.09, p ¼ 0.002. Inspection of the individual data found that all V-F rats decreased, while all F-V rats increased, their body weights during this phase. Thus, voluntary running in this experiment might have been more physically demanding than the forced running at 80 m/h, even when the number of wheel turns (i.e., the distance travelled) was much less in the former case (see the bottom panel of Fig. 6). 6. General discussion The present research successfully replicated the previous report that voluntary running in an activity wheel generates pica behavior (i.e., kaolin ingestion) in laboratory rats (Nakajima & Katayama, 2014). A principal new finding of this study is that the amount of kaolin intake is a positive function of the available time of voluntary running within the range of 20e60 min, although this relationship was blunted by a descending series of test execution (Experiment 1). The positive temporal mass-action principle observed here corresponds to the studies reporting that running-based CTA was a positive function of voluntary wheel running within the range of 5e30 min (Hayashi et al., 2002) or 15e120 min (Masaki & Nakajima, 2006), hence supporting the hypothesis that running induces nausea, which in turn produces CTA in rats. Another major new finding of the present research is that pica is also generated by forced running in a motorized wheel, as a positive function of the speed of wheel rotations (Experiment 2). This fact disagrees with the claim that forced running induces little nausea, because it has little rocking movements (Forristall et al., 2007). The new finding, however, is congruent with the report that antiemetic drug alleviates CTA caused by forced running (Eccles et al., 2005). The size of pica behavior was a positive function of the speed of running within the range of 98e365 m/h. It is notable, however, that the total number of wheel turns (i.e., the distance travelled), rather than the speed of running itself, might be the genuine critical factor affecting the size of pica, because this factor was not equated across conditions by changing the duration of rats’ wheel confinement (e.g., shortening the duration for the highspeed running condition). Experiment 3 showed that voluntary wheel running generated greater pica behavior than forced wheel running at 80 m/h, though the distance travelled was 27% shorter in the voluntary running condition. This result, thus, superficially disagrees with the positive relationship between the amount of running and the size of pica found in Experiments 1 and 2, and appears to partly support the claim of Forristall et al. (2007) that back-and-forth rocking

movements of the free wheel induces nausea in rats. It is noteworthy, however, that the wheels employed in Experiment 3 generated no rocking movements even in the voluntary running condition because of their high frictional resistance. Thus, Forristall et al.’s claim is not applicable here. As noted in the introduction of this article, both voluntary and forced running generate CTA in rats probably due to nausea induced by running (Dwyer et al., 2008; Eccles et al., 2005; Nakajima et al., 2006). To the best of my knowledge, only two studies have directly compared the size of CTA generated by voluntary and forced running treatments. Masaki and Nakajima (2006) reported that CTA generated by 60-min running was numerically greater for rats trained under forced running than voluntary running rats. This difference, however, failed to reach the statistical significance, and the distance travelled by the forced running rats was twice as long as that of the voluntary running rats. On the other hand, Forristall et al. (2007) have reported two experiments showing large CTA in voluntary running rats, compared with forced rats, when the distance travelled was matched between the two groups. Experiment 3 of the present research fortifies Forristall et al.’s report of stronger CTA by voluntary running. It is not clear why voluntary wheel running was more effective than forced running in generating pica behavior. Wheel running occurs as bouts of activity separated by pauses (Eikelboom & Mills, 1988; Premack & Schaeffer, 1962, 1963). Thus, one possible explanation is that break-and-run patterns of the voluntary running induced stronger gastrointestinal discomfort than the constant movement of the forced running. The irregularity rather than simple phasic nature of the break-and-run patterns might have been more critical. Another related possibility is that the top speed, rather than the average speed, is the critical factor in generating pica behavior. Whatever the reasons, voluntary running is seemingly more stressful than slow forced running, as reflected in the gradual decrease of body weight in the rats in the first running phase of Experiment 3. Contrary to these arguments, the exact nature of “enforceability” in the forced running might be more stressful for rats, because wheel rotations cannot be controlled by them. The uncontrollability might bring a feeling of “uncertainty” in rats, which is characterized as an aversive state (Imada & Nageishi, 1982). In other words, forced, rather than voluntary, running might be more stressful, and thus, by disturbing the rats, which could hinder generating adaptive pica behavior (see below for the discussion of the adaptive nature of pica behavior). It has been reported that acute elevation of serum corticosterone by wheel activity is large in forced running rats, compared with voluntary running rats (Ploughman et al., 2005, 2007). A chronic running study has also shown that forced, but not voluntary, running increased rats’ emotional defecation in the wheel and anxiety-like behaviors in the open field (Leasure & Jones, 2008). In addition, a recent study with mice provides evidence that forced running is more stressful by showing that forced running on a treadmill exacerbates intestinal inflammation with ulcerative colitis, whereas voluntary running in an activity wheel attenuated the symptom (Cook et al., 2013). These effects are probably caused by differentially altering the intestinal microbiome of mice (Allen et al., 2015). Some features of the pica results documented in the present research merit special attention. First, as in our previous study (Nakajima & Katayama, 2014), the intake of kaolin increased gradually over the 4-day running phase in all experiments reported here, suggesting that the observed pica is a product of the cumulative effect of running. Second, pica behavior was carried over to the next two days in many cases. These two features have been also reported in our previous study, where we ascribed the cause to Pavlovian conditioning of nausea to the environmental cues. The

S. Nakajima / Appetite 105 (2016) 85e94

rats might have felt malaise when re-encountering such cues, resulting in enhanced kaolin intake during the next 23-h period in their home cages. Such conditioned malaise would have added to the unconditioned nausea to produce the cumulative effect on the later days of the running phase. This might have also evoked some kaolin intake without running on the day after the wheel treatment phase. This scenario, however, is not likely as an account of carryover effect in the present study, where the rats were not moved to the experimental room in the baseline days. Although handling cues might have acquired such signal function in our previous study, this is highly unlikely in the present research. These cues were exposed to rats throughout the experimental period, and thus, they might not function as reliable conditioned stimuli evoking pica behavior. It is also noteworthy that we observed a slight, but statistically significant, reduction in food intake by forced wheel running in Experiment 2 and by voluntary wheel running in Experiment 3. A similar reduction in food intake, in conjunction with increase in kaolin intake, has been reported by our previous research on voluntary wheel running (Nakajima & Katayama, 2014), as well as by many studies on emetic treatments (DeJonghe & Horn, 2008; De Jonghe et al., 2009; Horn et at, 2009; Liu et al., 2005; Mitchell et al., 1976; Mitchell, Krusemark, et al., 1977; Tohei et al., 2011; Yamamoto et al., 2002, 2004, 2007). These studies imply the complementary nature of pica and anorexia in rats (Mitchell et al., 1976). Another account of the reduction in food intake by running is activity-based anorexia (Boakes, 2007; Epling & Pierce, 1996), although the durations of running opportunity employed in our studies were much shorter than the ordinary protocol of activity-based anorexia, where rats are kept in wheels for 23 h per day. Pica is an adaptive response to dietary toxin (De Jonghe et al., 2009), because kaolin clay absorbs toxin (Dominy, Davoust, & Minekus, 2004) and prevents diarrhea (Beck, Jenkins, Thurber, & Ambrus, 1977). Intestinal discomfort induced by running is probably fortuitously similar to toxin-induced nausea, thus, provoking the instinctive behavior of kaolin consumption. Tracking the internal state of this process might be revealed by a finer time-course analysis of pica by using an automatic monitoring system, such as that developed by Yamamoto et al. (2011). As concluded in previous studies of running-based pica (Nakajima & Katayama, 2014), further studies should examine the effect of anti-emetic drugs in such behavior, because drug administrations have been shown to be effective in attenuating runningbased CTA (Eccles et al., 2005) and pica caused by irradiation, motion sickness, or administering emetics (see the references listed in the introduction section). Another challenge posed by Nakajima and Katayama is exploring the possibility that swimming generates pica behavior in rats, because CTA can also be established by swimming (Masaki & Nakajima, 2004a, 2004b, 2005, 2010; Nakajima & Masaki, 2004; Nakajima, 2004). Although the present research has been successfully accomplished to reveal some features of pica based on voluntary running as well as forced running, many issues remain to be resolved. Acknowledgements This study was supported by JSPS KAKENHI grant (15K04201) and MEXT Strategic Project to Support the Formation of Research Bases at Private Universities. This research project and the animal facility were approved by the Animal Care and Use Committee of Kwansei Gakuin University, based on a Japanese law (the Act on Welfare and Management of Animals) and the guideline published by the Science Council of Japan (the Guidelines for Proper Conduct of Animals Experiments) in 2006, which is shown in the following web site. http://www.scj.go.jp/ja/info/kohyo/pdf/kohyo-20-k16-2e.pdf.

93

References Allen, J. M., Miller, M. E. B., Pence, B. D., Whitlock, K., Nehra, V., Gaskins, H. R., … Woods, J. A. (2015). Voluntary and forced exercise differentially alters the gut microbiome in C57BL/6J mice. Journal of Applied Physiology, 118, 1059e1066. http://dx.doi.org/10.1152/japplphysiol.01077.2014. Aung, H., Mehendale, S., Chang, W. T., Wang, C. Z., Xie, J. T., & Yuan, C. S. (2005). Scutellaria baicalensis decreases ritonavir-induced nausea. AIDS Research and Therapy, 2, 12. http://dx.doi.org/10.1186/1742-6405-2-12. Aung, H. H., Mehendale, S. R., Xie, J. T., Moss, J., & Yuan, C. S. (2004). Methylnaltrexone prevents morphine-induced kaolin intake in the rat. Life Sciences, 74, 2685e2691. http://dx.doi.org/10.1016/j.lfs.2003.08.047. Beck, I. T., Jenkins, N., Thurber, L., & Ambrus, J. L. (1977). Methods for the study of antidiarrheal agents: study of commonly used protective and adsorbent agents. Journal of Medicine, 8, 135e158. Boakes, R. A. (2007). Self-starvation in the rat. Running versus eating. Spanish Journal of Psychology, 10, 251e257. Boakes, R. A., & Nakajima, S. (2009). Conditioned taste aversions based on running or swimming. In S. Reilly, & T. R. Schachtman (Eds.), Conditioned taste aversion: Behavioral and neural processes (pp. 159e178). New York: Oxford University Press. Constancio, J., Pereira-Derderian, D. T., Menani, J. V., & De Luca, L. A. (2011). Mineral intake independent from gastric irritation or pica by cell-dehydrated rats. Physiology and Behavior, 104, 659e665. http://dx.doi.org/10.1016/ j.physbeh.2011.07.010. Cook, M. D., Martin, S. A., Williams, C., Whitlock, K., Wallig, M. A., Pence, B. D., et al. (2013). Forced treadmill exercise training exacerbates inflammation and causes mortality while voluntary wheel training is protective in a mouse model of colitis. Brain, Behavior, and Immunity, 33, 46e56. http://dx.doi.org/10.1016/ j.bbi.2013.05.005. De Jonghe, B. C., & Horn, C. C. (2008). Chemotherapy-induced pica and anorexia are reduced by common hepatic branch vagotomy in the rat. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 294, R756eR765. http://dx.doi.org/10.1152/ajpregu.00820.2007. De Jonghe, B. C., Lawler, M. P., Horn, C. C., & Tordoff, M. G. (2009). Pica as an adaptive response: Kaolin consumption helps rats recover from chemotherapy-induced illness. Physiology and Behavior, 97, 87e90. http://dx.doi.org/10.1016/ j.physbeh.2009.02.009. Dominy, N. J., Davoust, E., & Minekus, M. (2004). Adaptive function of soil consumption: an in vitro study modeling the human stomach and small intestine. Journal of Experimental Biology, 207, 319e324. http://dx.doi.org/10.1242/ jeb.00758. Dwyer, D. M., Boakes, R. A., & Hayward, A. J. (2008). Reduced palatability in lithiumand activity-based, but not in amphetamine-based, taste aversion learning. Behavioral Nneuroscience, 122, 1051e1060. http://dx.doi.org/10.1037/a0012703. Eccles, S., Kim, E. M., & O’Hare, E. (2005). Granisetron attenuates exercise-induced conditioned taste aversion in the rat. Appetite, 44, 325e328. http://dx.doi.org/ 10.1016/j.appet.2005.02.001. Eikelboom, R., & Mills, R. (1988). A microanalysis of wheel running in male and female rats. Physiology and Behavior, 43, 625e630. http://dx.doi.org/10.1016/ 0031-9384(88)90217-X. Epling, W. F., & Pierce, W. D. (Eds.). (1996). Activity anorexia: Theory, research, and treatment. Mahwah, NJ: Erlbaum. Forristall, J. R., Hookey, B. L., & Grant, V. L. (2007). Conditioned taste avoidance induced by forced and voluntary wheel running in rats. Behavioural Processes, 74, 326e333. http://dx.doi.org/10.1016/j.beproc.2006.12.002. Gillis, G. B., & Biewener, A. A. (2001). Hindlimb muscle function in relation to speed and gait: in vivo patterns of strain and activation in a hip and knee extensor of the rat (Rattus norvegicus). Journal of Experimental Biology, 204, 2717e2731. Grant, V. L., McDonald, S. V., Sheppard, R. C., Caldwell, C. L., Heeley, T. H., Brown, A. R., et al. (2012). Dissociation of conditioned taste avoidance from conditioned disgust reactions induced by wheel running in rats. Behavioural Processes, 90, 223e228. http://dx.doi.org/10.1016/j.beproc.2012.01.011. Han, Z. X., Xu, J., Wang, H. M., Ma, J., Sun, X., & Du, X. P. (2014). Antiemetic role of thalidomide in a rat model of cisplatin-induced emesis. Cell Biochemistry and Biophysics, 70, 361e365. http://dx.doi.org/10.1007/s12013-014-9921-8. Hayashi, H., Nakajima, S., Urushihara, K., & Imada, H. (2002). Taste avoidance caused by spontaneous wheel running: effects of duration and delay of wheel confinement. Learning and Motivation, 33, 390e409. http://dx.doi.org/10.1016/ S0023-9690(02)00004-8. Heth, C. D., Inglis, P., Russell, J. C., & Pierce, W. D. (2001). Conditioned taste aversion induced by wheel running is not due to novelty of the wheel. Physiology and Behavior, 74, 53e56. http://dx.doi.org/10.1016/S0031-9384(01)00553-4. Horn, C. C., De Jonghe, B. C., Matyas, K., & Norgren, R. (2009). Chemotherapyinduced kaolin intake is increased by lesion of the lateral parabrachial nucleus of the rat. American Journal of Physiology-Regulatory, Integrative and Comparative Physiology, 297, R1375eR1382. http://dx.doi.org/10.1152/ajpregu.00284.2009. Horn, C. C., Kimball, B. A., Wang, H., Kaus, J., Dienel, S., Nagy, A., … Andrews, P. L. (2013). Why can’t rodents vomit? A comparative behavioral, anatomical, and physiological study. PLoS One, 8(4), e60537. http://dx.doi.org/10.1371/ journal.pone.0060537. Imada, H., & Nageishi, Y. (1982). The concept of uncertainty in animal experiments using aversive stimulation. Psychological Bulletin, 91, 573e588. http:// dx.doi.org/10.1037/0033-2909.91.3.573.

94

S. Nakajima / Appetite 105 (2016) 85e94

Leasure, J. L., & Jones, M. (2008). Forced and voluntary exercise differentially affect brain and behavior. Neuroscience, 156, 456e465. http://dx.doi.org/10.1016/ j.neuroscience.2008.07.041. Lett, B. T., & Grant, V. L. (1996). Wheel running induces conditioned taste aversion in rats trained while hungry and thirsty. Physiology and Behavior, 59, 699e702. http://dx.doi.org/10.1016/0031-9384(95)02139-6. Lett, B. T., Grant, V. L., & Gaborko, L. L. (1998). Wheel running simultaneously induces CTA and facilitates feeding in non-deprived rats. Appetite, 31, 351e360. http://dx.doi.org/10.1006/appe.1998.0171. Liu, Y. L., Malik, N., Sanger, G. J., Friedman, M. I., & Andrews, P. L. (2005). Picada model of nausea? Species differences in response to cisplatin. Physiology and Behavior, 85, 271e277. http://dx.doi.org/10.1016/j.physbeh.2005.04.009. Malik, N. M., Liu, Y. L., Cole, N., Sanger, G. J., & Andrews, P. L. (2007). Differential effects of dexamethasone, ondansetron and a tachykinin NK1 receptor antagonist (GR205171) on cisplatin-induced changes in behaviour, food intake, pica and gastric function in rats. European Journal of Pharmacology, 555, 164e173. http://dx.doi.org/10.1016/j.ejphar.2006.10.043. Masaki, T., & Nakajima, S. (2004a). Swimming-induced taste aversion and its prevention by a prior history of swimming. Learning and Motivation, 35, 406e418. http://dx.doi.org/10.1016/j.lmot.2004.04.003. Masaki, T., & Nakajima, S. (2004b). Taste aversion learning induced by delayed swimming activity. Behavioural Processes, 67, 357e362. http://dx.doi.org/ 10.1016/j.beproc.2004.06.005. Masaki, T., & Nakajima, S. (2005). Further evidence for conditioned taste aversion induced by forced swimming. Physiology and Behavior, 84, 9e15. http:// dx.doi.org/10.1016/j.physbeh.2004.09.022. Masaki, T., & Nakajima, S. (2006). Taste aversion in rats induced by forced swimming, voluntary running, forced running, and lithium chloride injection treatments. Physiology and Behavior, 88, 411e416. http://dx.doi.org/10.1016/ j.physbeh.2006.04.013. Masaki, T., & Nakajima, S. (2010). The effect of swimming experience on acquisition and retention of swimming-based taste aversion learning in rats. Learning and Motivation, 41, 32e47. http://dx.doi.org/10.1016/j.lmot.2009.09.001. McCaffrey, R. J. (1985). Appropriateness of kaolin consumption as an index of motion sickness in the rat. Physiology and Behavior, 35, 151e156. http://dx.doi.org/ 10.1016/0031-9384(85)90329-4. McCutcheon, B., Ballard, M., & McCaffrey, R. J. (1992). Intraperitoneally injected cholecystokinin-octapeptide activates pica in rats. Physiology and Behavior, 51, 543e547. http://dx.doi.org/10.1016/0031-9384(92)90177-4. Mitchell, D., Krusemark, M. L., & Hafner, E. (1977). Pica: a species relevant behavioral assay of motion sickness in the rat. Physiology and Behavior, 18, 125e130. http://dx.doi.org/10.1016/0031-9384(77)90103-2. Mitchell, D., Laycock, J. D., & Stephens, W. F. (1977). Motion sickness-induced pica in the rat. The American Journal of Clinical Nutrition, 30, 147e150. Mitchell, D., Wells, C., Hoch, N., Lind, K., Woods, S. C., & Mitchell, L. K. (1976). Poison induced pica in rats. Physiology and Behavior, 17, 691e697. http://dx.doi.org/ 10.1016/0031-9384(76)90171-2. Morita, M., Takeda, N., Kubo, T., & Matsunaga, T. (1988). Pica as an index of motion sickness in rats. ORL: Journal for Otorhinolaryngology and Its Related Specialties, 50, 188e192. Morita, M., Takeda, N., Kubo, T., Yamatodani, A., Wada, H., & Matsunaga, T. (1988). Effects of anti-motion sickness drugs on motion sickness in rats. ORL: Journal for Otorhinolaryngology and Its Related Specialties, 50, 330e333. Nakajima, S. (2004). Conditioned ethanol aversion in rats induced by voluntary wheel running, forced swimming, and electric shock: an implication for aversion therapy of alcoholism. Integrative Physiological and Behavioral Science, 39, 95e104. http://dx.doi.org/10.1007/BF02734275. Nakajima, S. (2011). Calorie supply does not alleviate running-based taste aversion learning in rats. Appetite, 57, 605e614. http://dx.doi.org/10.1016/ j.appet.2011.08.001. Nakajima, S., Hayashi, H., & Kato, T. (2000). Taste aversion induced by confinement in a running wheel. Behavioural Processes, 49, 35e42. http://dx.doi.org/10.1016/ S0376-6357(00)00071-1. Nakajima, S., & Katayama, T. (2014). Running-based pica in rats. Evidence for the gastrointestinal discomfort hypothesis of running-based taste aversion. Appetite, 83, 178e184. http://dx.doi.org/10.1016/j.appet.2014.08.031. Nakajima, S., & Masaki, T. (2004). Taste aversion learning induced by forced swimming in rats. Physiology and Behavior, 80, 623e628. http://dx.doi.org/ 10.1016/j.physbeh.2003.11.006. Nakajima, S., Urata, T., & Ogawa, Y. (2006). Familiarization and cross-familiarization of wheel running and LiCl in conditioned taste aversion. Physiology and

Behavior, 88, 1e11. http://dx.doi.org/10.1016/j.physbeh.2006.02.006. Ploughman, M., Granter-Button, S., Chernenko, G., Attwood, Z., Tucker, B. A., Mearow, K. M., et al. (2007). Exercise intensity influences the temporal profile of growth factors involved in neuronal plasticity following focal ischemia. Brain Research, 1150, 207e216. http://dx.doi.org/10.1016/j.brainres.2007.02.065. Ploughman, M., Granter-Button, S., Chernenko, G., Tucker, B. A., Mearow, K. M., & Corbett, D. (2005). Endurance exercise regimens induce differential effects on brain-derived neurotrophic factor, synapsin-I and insulin-like growth factor I after focal ischemia. Neuroscience, 136, 991e1001. http://dx.doi.org/10.1016/ j.neuroscience.2005.08.037. Premack, D., & Schaeffer, R. W. (1962). Distributional properties of operant-level locomotion in the rat. Journal of the Experimental Analysis of Behavior, 5, 89e95. http://dx.doi.org/10.1901/jeab.1962.5-89. Premack, D., & Schaeffer, R. W. (1963). Some parameters affecting the distributional properties operant-level running in rats. Journal of the Experimental Analysis of Behavior, 6, 473e475. http://dx.doi.org/10.1901/jeab.1963.6-473. Rudd, J. A., Yamamoto, K., Yamatodani, A., & Takeda, N. (2002). Differential action of ondansetron and dexamethasone to modify cisplatin-induced acute and delayed kaolin consumption (“pica”) in rats. European Journal of Pharmacology, 454, 47e52. http://dx.doi.org/10.1016/S0014-2999(02)02472-X. Saeki, M., Sakai, M., Saito, R., Kubota, H., Ariumi, H., Takano, Y., … Kamiya, H. O. (2001). Effects of HSP-117, a novel tachykinin NK1-receptor antagonist, on cisplatin-induced pica as a new evaluation of delayed emesis in rats. Japanese Journal of Pharmacology, 86, 359e362. http://dx.doi.org/10.1254/jjp.86.359. Sharma, S. S., Gupta, S. K., Kochupillai, V., Seth, S. D., & Gupta, Y. K. (1997). Cisplatininduced pica behaviour in rats is prevented by antioxidants with antiemetic activity. Environmental Toxicology and Pharmacology, 3, 145e149. http:// dx.doi.org/10.1016/S1382-6689(97)00151-8. Takeda, N., Hasegawa, S., Morita, M., Horii, A., Uno, A., Yamatodani, A., et al. (1995a). Neuropharmacological mechanisms of emesis. I. Effects of antiemetic drugs on motion-and apomorphine-induced pica in rats. Methods and Findings in Experimental and Clinical Pharmacology, 17, 589e590. Takeda, N., Hasegawa, S., Morita, M., Horii, A., Uno, A., Yamatodani, A., et al. (1995b). Neuropharmacological mechanisms of emesis. II. Effects of antiemetic drugs on cisplatin-induced pica in rats. Methods and Findings in Experimental and Clinical Pharmacology, 17, 647e652. Takeda, N., Hasegawa, S., Morita, M., & Matsunaga, T. (1993). Pica in rats is analogous to emesis: an animal model in emesis research. Pharmacology Biochemistry and Behavior, 45, 817e821. http://dx.doi.org/10.1016/0091-3057(93)90126-E. Tohei, A., Kojima, S. I., Ikeda, M., Hokao, R., & Shinoda, M. (2011). Effects of cyclophosphamide on the kaolin consumption (pica behavior) in five strains of adult male rats. Journal of Veterinary Medical Science, 73, 901e906. http://dx.doi.org/ 10.1292/jvms.10-0433. Watson, P. J., Hawkins, C., McKinney, J., Beatey, S., Bartles, R. R., & Rhea, K. (1987). Inhibited drinking and pica in rats following 2-deoxy-D-glucose. Physiology and Behavior, 39, 745e752. http://dx.doi.org/10.1016/0031-9384(87)90260-5. Watson, P. J., & Leitner, C. (1988). Patterns of increased and decreased ingestive behavior after injections of lithium chloride and 2-deoxy-D-glucose. Physiology and Behavior, 43, 697e704. http://dx.doi.org/10.1016/0031-9384(88)90366-6. Yamamoto, K., Asano, K., Matsukawa, N., Imaizumi, M., & Yamatodani, A. (2011). Time-course analysis of pica in rats using an automatic feeding monitoring system. Journal of Pharmacological and Toxicological Methods, 63, 30e34. http:// dx.doi.org/10.1016/j.vascn.2010.04.011. Yamamoto, K., Asano, K., Tasaka, A., Ogura, Y., Kim, S., Ito, Y., et al. (2014). Involvement of substance P in the development of cisplatin-induced acute and delayed pica in rats. British Journal of Pharmacology, 171, 2888e2899. http:// dx.doi.org/10.1111/bph.12629. Yamamoto, K., Nakai, M., Nohara, K., & Yamatodani, A. (2007). The anti-cancer druginduced pica in rats is related to their clinical emetogenic potential. European Journal of Pharmacology, 554, 34e39. http://dx.doi.org/10.1016/ j.ejphar.2006.09.058. Yamamoto, K., Ngan, M. P., Takeda, N., Yamatodani, A., & Rudd, J. A. (2004). Differential activity of drugs to induce emesis and pica behavior in Suncus murinus (house musk shrew) and rats. Physiology and Behavior, 83, 151e156. Yamamoto, K., Takeda, N., & Yamatodani, A. (2002). Establishment of an animal model for radiation-induced vomiting in rats using pica. Journal of Radiation Research, 43, 135e141. http://dx.doi.org/10.1269/jrr.43.135. Yuan, C. S., Wang, C. Z., Mehendale, S. R., Aung, H. H., Foo, A., & Israel, R. J. (2009). Protease inhibitor-induced nausea and vomiting is attenuated by a peripherally acting, opioid-receptor antagonist in a rat model. AIDS Research and Therapy, 6, 19. http://dx.doi.org/10.1186/1742-6405-6-19.