An Open-Source, Automated Home-Cage Sipper Device for Monitoring Liquid Ingestive Behavior in Rodents

Abstract

Measuring ingestive behavior of liquids in rodents is commonly used in studies of reward, metabolism, and circadian biology. Common approaches for measuring liquid intake in real time include computer-tethered lickometers or video-based systems. Additionally, liquids can be measured or weighed to determine the amount consumed without real-time sensing. Here, we built a photobeam-based sipper device that has the following advantages over traditional methods: (1) it is battery powered and fits in vivarium caging to allow home-cage measurements; (2) it quantifies the intake of two different liquids simultaneously for preference studies; (3) it is low cost and easily constructed, enabling high-throughput experiments; and (4) it is open source so that others can modify it to fit their experimental needs. We validated the performance of this device in three experiments. First, we calibrated our device using time-lapse video-based measurements of liquid intake and correlated sipper interactions with liquid intake. Second, we used the sipper device to measure preference for water versus chocolate milk, demonstrating its utility for two-bottle choice tasks. Third, we integrated the device with fiber photometry, establishing its utility for measuring neural activity in studies of ingestive behavior. This device requires no special equipment or caging, and is small, battery powered, and wireless, allowing it to be placed directly in rodent home cages. The total cost of fabrication is less than $100, and all design files and code are open source. Together, these factors greatly increase scalability and utility for a variety of behavioral neuroscience applications.

Keywords: Arduino; open source hardware; two-bottle choice.

Graphical abstract

Figure 1.

Construction and implementation of a home-cage drinking monitor. A, Circuit wiring diagram of electronic components; 3 V power is supplied to the photo-interrupters, which are connected to a ground pin. Each photo-interrupter is also attached to a digital output pin (shown as D9 and D10). B, 3D rendering of the 3D printed housing for the drinking monitor; views of the front tube assembly (left) and rear battery casing (right) are shown. C, D, Photo of the assembled device (C) and the assembled device operating in the home cage (D). E, Example data as written to the SD card.

Figure 2.

Functional validation of liquid consumption with the home-cage drinking monitor. A, Schematic of the experimental setup. B, Histogram depicting the distribution of R ² values of the correlation between measurements (counts or duration) registered on the individual sipper devices with the volume of liquid consumed from the same device, as measured via time-lapse video. C, D, Across all 11 mice tested, sipper counts and durations of interactions with the sipper device weakly correlated with the amount of water consumed by visual quantification with time-lapse video. E, Circadian rhythms in lick duration and volume of liquid consumed were evident over the 5 d of recording.

Figure 3.

Two-bottle choice task with the home-cage drinking monitor. A, Experimental schematic; mice had free access to two liquids in the drinking monitor device: water or chocolate milk. B–D, All mice exhibited a clear preference for chocolate milk over water. Mice exhibited a longer total duration of sipper interactions (B), an increased duration of sipper interaction bouts (C), and an increased number of sipper approaches (D) for chocolate milk over water. **p < 0.01.

Figure 4.

Integration of the drinking monitor with in vivo fiber photometry. A, Schematic of the experiment; photo-interrupter beam breaks registered on digital Arduino pins triggered a TTL pulse to an in vivo fiber photometry system. B, Representative raw trace of gCamp signal recorded in the DMS (top) and concomitant sipper interaction bouts (bottom). C, Normalized gCamp traces were averaged across trials for all mice and were aligned to the onset of the lick bout. Black line, Mean; red, SEM. D, The duration of sipper interactions is aligned to the onset of the lick bout; the mean probability of a sipper interaction is indicated in black, and SEM is depicted in blue. Both C and D are aligned to the onset of lick bout.