Sex-specific effects of appetite suppressants on stereotypy in rats

Abstract

This study investigated the sex-specific effects of commonly prescribed appetite suppressants on body weight and the manifestation of motor side effects, specifically stereotypy. Employing video recordings and DeepLabCut (DLC) for precise behavioral quantification, we analyzed stereotypy, defined as purposeless, repetitive motor behaviors, in male and female rats. Under control (saline) conditions, male rats exhibited a greater propensity for weight gain compared to females. However, in contrast, female rats demonstrated greater and more homogenous weight loss than males following the administration of diethylpropion and tesofensine. Phentermine and mazindol induced comparable weight loss in both sexes, whereas cathine elicited weight reduction exclusively in males. 5-HTP and d-amphetamine administration only prevented weight gain relative to controls. Analysis of motor side effects revealed that drugs primarily targeting dopamine pathways - specifically, phentermine, mazindol, diethylpropion, cathine, and d-amphetamine - induced pronounced stereotypies, particularly head-weaving, in both sexes. Interestingly, tesofensine elicited head-weaving behavior exclusively in female subjects, albeit to a lesser extent than that observed with other dopaminergic agents; conversely, tesofensine was most frequently associated with orolingual dyskinesia. Male subjects treated with these same drugs exhibited an unexpected effect: spontaneous ejaculations, potentially attributable to the combined effects on dopamine and serotonin signaling in brain regions regulating sexual function. Network analysis and Markov transition matrices revealed distinct behavioral profiles associated with head-weaving, which emerged as the dominant attractor state, suggesting potential mechanistic differences among these drugs. Collectively, this study provides a valuable database characterizing the behavioral side effects of appetite suppressants.

Fig 1. Effects of appetite suppressants on body weight change in male and female rats.

The change in body weight of control naïve rats daily injected intraperitoneally with saline (n = 6 males(▲ triangles), n = 6 females (● circles)) from day 1 to day 7 compared with rats injected intraperitoneally with phentermine at 20 mg/kg (n = 6 males, n = 6 females), mazindol at 10 mg/kg (n = 6 males, n = 6 females), diethylpropion at 20 mg/kg (n = 6 males, n = 6 females), cathine at 20 mg/kg (n = 6 males, n = 6 females), d-amphetamine at 10 mg/Kg (n = 3 males, n = 3 females), tesofensine (subcutaneous) at 2 mg/kg (n = 6 males, n = 6 females), and 5-HTP/CB at 31 mg/kg (n = 6 males, n = 6 females). The break in the axis indicates where the treatment was stopped. For comparison reasons, the saline group was replotted for the rest of the panels. The body weight was measured again 7 days after the last injection (withdrawal). The horizontal dotted line represents the baseline body weight before treatment. * p < 0.05 for treatment days where there was a significant difference between sexes within the drug. Data is presented as mean ± SEM. # one-way ANOVA, p < 0.05, significantly different from the saline group and same-sex on withdrawal.

Fig 2. Automatic quantification of behavioral states induced by appetite suppressants.

A) Experimental protocol for video recording rat’s behavior. For the data acquisition, rats were filmed from top and bottom views. The first 15 minutes were used as a baseline, followed by the injection of an appetite suppressant at 15 minutes. The recording continues until completing four hours. Images were converted into videos, and a machine learning algorithm, ResNet-50, was used for training these key points: the nose, front paws, hind paws, and tail base. Subsequently, the coordinates of these key points were extracted. After correcting errors, behaviors such as locomotion, quiet-awake state, rearing, grooming, and stereotypy were defined based on these coordinates and quantitatively analyzed across all frames. B) Ethograms were built of the following behavioral states: 1) Head weaving, 2) Quiet sleep state, 3) Locomotion, 4) Grooming, and 5) Rearing. C) Pose estimation images extracted from DLC show behaviors in the same order as in panel (B). D) Representative ethogram of male rats. Left panel: saline-injected rat; right panel: rat injected with Phentermine at 20 mg/kg. Behavioral states are color-coded as follows: 1) Stereotypy (black), 2) Quiet-awake state (red), 3) Locomotion (blue), 4) Grooming (cyan), and 5) Rearing (magenta). In S1–S8 Figs, the complete ethograms for all rats and drugs can be seen, and the entire database can be found in [87].

Fig 3. Effect of appetite suppressants on head-weaving stereotypy in male rats.

Drugs were ranked in descending order of potency on stereotypy. Each behavior is quantified as a proportion of the time spent in that state during the 4 h duration of each session. Each dot within the violin plots represents an individual rat, and the colored dot corresponds to the same rat across various days and behaviors. A) Saline: As expected, under physiological conditions, rats spend most of their daytime in a quiet-awake time (immobile but mainly in a sleeping-like posture) or exploring their environment (locomotion). No head-weaving stereotypy was observed. B) Phentermine (20 mg/kg): Stereotypy strengthens over the days, occupying more than 60% of session time from the fourth day onwards. Moreover, we observe a progressively shorter onset of stereotypy across days. C) Mazindol (10 mg/kg) also induced significant head weaving stereotypy, although with more variability than phentermine. D) Diethylpropion (20 mg/kg) was the third appetite suppressant with a robust stereotypy, although its stereotypy was more homogeneous than phentermine (CV for DEP = 17.85% vs CV for PHEN = 22.01%). E) Cathine 20 mg/kg (d-Norpseudoephedrine (NPE)): Stereotypy was not as pronounced as with phentermine, and rats exhibited high variability (CV = 37.81%), but it tended to strengthen across days. F) d-amphetamine (10 mg/kg): High but variable levels of stereotypy were observed from the first day of treatment (CV = 53.35%). Additionally, this drug induced a quiet-awake state (including sleep-like behaviors) on the first and third days of treatment, suggesting that their effects disappeared more rapidly than other appetite suppressants. G) Tesofensine (2 mg/kg): Unlike other appetite suppressants, males do not exhibit significant head weaving stereotypy, as previously reported [25], except for one subject on the third day. Tesofensine-treated rats exhibited prolonged periods of immobility (quite-awake state) while remaining awake, alternating with brief sleep-like posture, suggesting this drug may have insomnia-inducing properties. H) 5-HTP (31 mg/kg): 5-HTP/CB treatment did not induce head-weaving stereotypy. Rats treated with 5-HTP exhibited behavioral profiles similar to saline controls, characterized by prolonged periods of inactivity, primarily sleep-like behavior. For all the drugs, the symbols # and + and the color next to each dot indicate the rat and the day on which a subject presented spontaneous ejaculation (+) and/or backward locomotion (# moonwalk stereotypy).

Fig 4. Effect of appetite suppressants on head-weaving stereotypy in female rats.

Drugs were ranked in descending order of potency on stereotypy. Same conventions as in Fig 3, but for female rats: A) Saline: female rats spend most of their time in the quiet-awake state (mostly sleeping). B) Phentermine (20 mg/kg): Induced stereotypy in females is even more homogeneous than in males from the first day (males had a larger coefficient of variation in treatment CV = 22.2% versus females CV = 10.5%. Moreover, the onsets are quicker from the second day of treatment, with rats exhibiting head-weaving behavior within 15 minutes post-injection. C) Mazindol (10 mg/kg): Similar to males, female rats exhibit marked head weaving stereotypy. However, one subject (green dot) stands out, as it exhibited very low levels of stereotypy but high locomotion during the first three days of treatment. D) Diethylpropion (20 mg/kg): In contrast to the low variability shown by males (CV = 17.8%), females exhibited high variability in the induced levels of stereotypy (CV = 35.9%). A particular case stands out with the cyan dot, which consistently showed stereotypy levels below the mean except for the second day of the treatment. E) Cathine 20 mg/Kg (d-Norpseudoephedrine NPE): Cathine was the drug that induced the fourth-highest average level of stereotypy in both females and males. However, unlike males (CV = 37.8%), there was significant variability between subjects (CV = 49%) and across days. For example, the subjects represented by the red dot showed increasing levels of stereotypy over the days, whereas the subjects represented by the cyan dot transitioned from high levels of stereotypy during the first days of treatment to levels below the mean on days 6 and 7 of treatment. F) d-amphetamine (10 mg/kg): From the first day of treatment, subjects exhibited a level above 50% of time spent on stereotypy, which slightly increased on the second day. Notably, one subject (cyan dot) showed a considerable decrease in stereotypy levels on the third day, quantified after 100 minutes post-injection. G) Tesofensine (2 mg/kg): In females, tesofensine, unlike in males, did induced head weaving behavior, although with high variability between subjects (CV = 116.04%), which was also characterized by its delayed onset (82.62 min), highlighting a gender-specific effect of the drug. H) 5-HTP (31 mg/kg): Like in male rats, serotonin alone does not induce the head-weaving stereotypy. For all the drugs, the symbols # and color next to each dot indicate the rat and the day a subject presented backward locomotion (# moonwalk stereotypy).

Fig 5. t-distributed Stochastic Neighbor Embedding (t-SNE) and hierarchical clustering analysis delineate distinct clusters of drug-induced motor effects in rats, with tesofensine in females demonstrating an intermediate profile.

The analysis reveals two primary clusters: one encompassing amphetamine-like drugs associated with stereotyped behaviors and another comprising drugs that do not elicit such behaviors. t-SNE map illustrating the clustering of motor effects across various treatments. Each data point represents an individual subject, with triangles indicating males and circles indicating females. Colors denote different treatments: saline (pink), phentermine (yellow), mazindol (light green), diethylpropion (beige), cathine (blue), tesofensine (burgundy), and 5-HTP (dark green). Tesofensine-treated females (burgundy circles) are positioned intermediately between the two main clusters. This suggests that tesofensine in females elicits a motor effect profile that shares characteristics with both amphetamine-like and non-stereotypy-inducing drugs.

Fig 6. Appetite suppressants alter state transition probabilities in a Markov chain model of rat behavior.

Markov chain analysis reveals significant shifts in behavioral state transitions induced by appetite suppressants in rats. Saline-treated animals (controls) demonstrate a pattern of transitioning from a quiet-awake state to locomotion, indicative of exploratory behavior. In contrast, appetite suppressant-treated rats exhibit a reduced probability of transitioning from quiet-awake to locomotion and an increased probability of transitioning directly from locomotion to stereotyped behaviors, such as head weaving. (A) Mean percentage of time spent in the head-weaving stereotypy state across seven days of treatment. Data are presented separately for males (left) and females (right), with each point representing an individual animal. (B-I) Directed network graphs illustrate the transition probabilities between behavioral states for each treatment group and sex (i.e., saline, phentermine, mazindol, diethylpropion, cathine, d-amphetamine, tesofensine, and 5-HTP/CB, respectively). Nodes (circles) represent distinct behavioral states: head-weaving stereotypy (black), quiet-awake (red), locomotion (blue), grooming (magenta), and rearing (cyan). Edges represent the transition probability between states, with thickness and color intensity corresponding to probability magnitude. Thicker, more yellow-toned edges signify higher transition probabilities, while thinner, more blue-toned edges indicate lower probabilities.