A reverse-translational approach to bipolar disorder: rodent and human studies in the Behavioral Pattern Monitor

Abstract

Mania is the defining feature of bipolar disorder (BD). There has been limited progress in understanding the neurobiological underpinnings of BD mania and developing novel therapeutics, in part due to a paucity of relevant animal models with translational potential. Hyperactivity is a cardinal symptom of mania, traditionally measured in humans using observer-rated scales. Multivariate assessment of unconditioned locomotor behavior using the rat Behavioral Pattern Monitor (BPM) developed in our laboratory has shown that hyperactivity includes complex multifaceted behaviors. The BPM has been used to demonstrate differential effects of drugs on locomotor activity and exploratory behavior in rats. Studies of genetically engineered mice in a mouse BPM have confirmed its utility as a cross-species tool. In a "reverse-translational" approach to this work, we developed the human BPM to characterize motor activity in BD patients. Increased activity, object interactions, and altered locomotor patterns provide multi-dimensional phenotypes to model in the rodent BPM. This unique approach to modeling BD provides an opportunity to identify the neurobiology underlying BD mania and test novel antimanic agents.

Figure 1. Stimulant effects on rat locomotor patterns in the Behavioral Pattern Monitor

The effects of saline (a), amphetamine (b), phencyclidine (c) and scopolamine (d) on rat behavioral organization. a) The rat treated with saline explored little of the environment, making two or three excursions around the chamber, exhibiting short meandering movement in one or two areas then moving on. b) The amphetamine-treated rat however made numerous circuits of the chamber, crossing from one area to another in relatively straight lines, crossing the center as often as being close to the chamber walls, leading to a great variability of movement. The rat also exhibited more than one area of focused activity [‘home corner’ (Geyer 1982) or ‘home base’ (Eilam and Golani 1989)] c) The phencyclidine-treated rat exhibited sweeping movement patterns, from one corner to another, often circling from a home base that covered the right side of the chamber, with repetitive movements. d) The scopolamine-treated rat displayed increased long and straight movements particularly close to the chamber walls, deviating very little from this path. The level of activity was as great as amphetamine but with scopolamine the rat did not spend time focusing on any areas in particular, nor did it cross the center very often, displaying very repetitive movements. Data are provided for measures of locomotor activity (transitions), locomotor pattern (spatial d) and exploration (holepokes).

Figure 2. Effects of amphetamine on mouse locomotor patterns in the Behavioral Pattern Monitor

The effects of saline (a) and amphetamine (b) on the behavioral organization of a representative mouse are presented. a) Mice treated with saline exhibit exploration throughout the chamber, but only perform a limited number of excursions around the chamber. This animal spent most of the time in the bottom right hand side of the chamber (the “home corner”). b) Mice treated with amphetamine however, exhibit a large number of excursions around the chamber, covering the chamber floor many times in a variety of paths. They also display several areas where their behavior is concentrated, suggesting several home bases as opposed to the one home corner observed in the saline administered mouse. While the x-y plots represented here are genuine movements, they also reflect a limitation of the number of photobeams used to identify a subject's position. Data are also provided for measures of locomotor activity (transitions), locomotor pattern (spatial d), and exploration (holepokes). Mice treated with amphetamine display a hyperactive phenotype, with lower spatial d and lower exploratory behavior when compared to control animals.

Figure 3. Locomotor patterns of dopamine transporter knockdown and wildtype littermate mice in the Behavioral Pattern Monitor

Representative locomotor patterns of dopamine transporter wildtype (WT; a) and knockdown (KD; b) mice are shown. a) WT mice spend most of their time near the chamber walls and while some movement is made to explore the center, activity is concentrated in the left wall, where the mouse circles back and forth. b) In contrast to the time spent in the home corner by the WT mouse, this KD mouse displayed numerous areas of interest and exhibited more varied paths of activity. Locomotor activity (transitions), pattern (spatial d), and exploration (holepokes) data are provided, with the KD mice displaying greater activity, exploratory behavior, and straighter line movements (lower spatial d) when compared to their WT littermates.

Figure 4. Effects of the selective dopamine transporter uptake inhibitor GBR 12909 on mouse locomotor patterns in the Behavioral Pattern Monitor

The representative locomotor patterns are shown for mice treated with saline (a) or GBR 12909 (b). a) Mice treated with saline display very limited activity, making only one or two excursions around the chamber with limited exploration into the center of the chamber. b) Mice treated with GBR 12909 display far greater levels of activity, with numerous areas of focused activity and greater variety of paths taken. Data are also presented for locomotor activity (transitions), pattern (spatial d), and exploratory behavior (holepokes) with mice treated with GBR 12909 displaying hyperactivity, increased exploratory behavior, and straighter line movements (lower spatial d) than mice treated with saline.

Figure 5. Locomotor patterns of human subjects in the human Behavioral Pattern Monitor

The layout of the room as observed through a fisheye lens is outlined in red. The locomotor pattern of a representative healthy subject (a), or manic BD (b) and schizophrenia (c) patients are shown in black. The location of the subject's upper torso (specifically, the LifeShirt vest) in x- and y-coordinates was recorded by tracking software (Clever Systems, Inc) as the subject examined the room and the objects located therein. An accelerometer embedded in a wearable ambulatory monitoring device (Lifeshirt) also recorded levels of motor activity in digital units for each subject. a) The healthy comparison subject walked around the room once, investigated the window, which is covered, examined some objects placed on the bookshelves farthest from the door, and finally moved to the desk, spending the remainder of time examining that area and the objects found there. b) The manic BD patient conducted numerous excursions around the room, often concentrating movement at specific locations such as the window and bookshelves farthest from the door and the small filing cabinet. Apart from the obvious quantity of movements that differentiate this subject from the healthy comparison subject, the manic BD subject also clearly failed to exhibit a preference for one location, spending time in numerous areas. This subject also displayed longer tracks of movement from one area of the room to another, with a large variability in the paths chosen. c) The schizophrenia patient displayed a virtual lack of exploratory behavior. This subject remained at the desk for the duration of the session. Some objects were investigated on the desk, but all exploration was specific and within a limited area. The quantitative data are shown for these representative subjects' acceleration, transitions, spatial d, and exploratory behavior (object interactions).

Figure 6. Observed walking and entropy-derived probability

The close correspondence between walking behavior as observed by video ratings (gray line) and entropy-derived probability of walking (black line) for one healthy human subject across a 15-minute session in the human BPM. The entropy-derived probability of walking corresponded exactly to the observed walking at each time-frame.