Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2018 May 15;12:18.
doi: 10.3389/fnsys.2018.00018. eCollection 2018.

Open Source Tools for Temporally Controlled Rodent Behavior Suitable for Electrophysiology and Optogenetic Manipulations

Affiliations
Free PMC article

Open Source Tools for Temporally Controlled Rodent Behavior Suitable for Electrophysiology and Optogenetic Manipulations

Nicola Solari et al. Front Syst Neurosci. .
Free PMC article

Abstract

Understanding how the brain controls behavior requires observing and manipulating neural activity in awake behaving animals. Neuronal firing is timed at millisecond precision. Therefore, to decipher temporal coding, it is necessary to monitor and control animal behavior at the same level of temporal accuracy. However, it is technically challenging to deliver sensory stimuli and reinforcers as well as to read the behavioral responses they elicit with millisecond precision. Presently available commercial systems often excel in specific aspects of behavior control, but they do not provide a customizable environment allowing flexible experimental design while maintaining high standards for temporal control necessary for interpreting neuronal activity. Moreover, delay measurements of stimulus and reinforcement delivery are largely unavailable. We combined microcontroller-based behavior control with a sound delivery system for playing complex acoustic stimuli, fast solenoid valves for precisely timed reinforcement delivery and a custom-built sound attenuated chamber using high-end industrial insulation materials. Together this setup provides a physical environment to train head-fixed animals, enables calibrated sound stimuli and precisely timed fluid and air puff presentation as reinforcers. We provide latency measurements for stimulus and reinforcement delivery and an algorithm to perform such measurements on other behavior control systems. Combined with electrophysiology and optogenetic manipulations, the millisecond timing accuracy will help interpret temporally precise neural signals and behavioral changes. Additionally, since software and hardware provided here can be readily customized to achieve a large variety of paradigms, these solutions enable an unusually flexible design of rodent behavioral experiments.

Keywords: head-fixed; measurement noise; reinforcement; sensory stimulus; sound attenuation; temporal control.

Figures

FIGURE 1
FIGURE 1
Firing rate analysis. (A) Left, mice are exposed to temporally controlled sensory cues, reinforcers and external neural stimulation. Right, neural activity (black ticks) can be aligned to these events in raster plots (top) and peri-event time histograms (bottom). (B) Neurons can be categorized according to their tuning to these events. (C) The information carried by neural responses about external events can be quantified by information theory tools.
FIGURE 2
FIGURE 2
The effects of noise on event timing. (A) We simulated an event train without (top) or with Gaussian noise (middle) as well as a spike train consisting of ‘stimulus-evoked’ spikes according to a normal distribution and background Poisson spiking (bottom). (B) Raster plots aligned to noiseless (left) or noisy (right) events. (C) Peri-event time histograms (equivalent to event-spike cross-correlations) corresponding to the raster plots above. Blue, raw traces; black, Gaussian fits. (D) Mutual information between spike times and event times without (left) or with (right) noise. Note the second increase of mutual information around 500 ms corresponding to the information on the event times carried by the lack of event-aligned spikes, demonstrating the power of information theory to detect non-linear correlations. (E) Both linear cross-correlation (left) and non-linear mutual information (right) decreases with the amount of added noise in event timing. Code is available at https://github.com/hangyabalazs/Rodent_behavior_setup/, Experiment_simulation.m.
FIGURE 3
FIGURE 3
Head-fixed setup. (A) The animal is held by an implanted head bar with a pair of metal holders (1), facing a custom-made lick port hosting an IR emitter and an IR receiver (2, 3) for lick detection and a plastic water spout (4). Air-puff is delivered via a cannula placed near the animal’s face (5). Visual and auditory cues are delivered by a central LED (6) and lateral speakers (7). (B) Schematic diagram of the behavior setup. Cue and reinforcement delivery are controlled by Bpod. Motion is monitored with a camera using Bonsai open software.
FIGURE 4
FIGURE 4
Neuronal recordings. (A) Raster plot of lick responses (black ticks) to reward-predicting cues (red line) of a mouse trained on an auditory Pavlovian task. (B) Local filed potential deflections in response to photostimulation (blue squares) in a mouse expressing the light-sensitive channelrhodopsin in parvalbumin-expressing neurons of the HDB. (C) Raster plot of action potentials (black ticks) of two neurons recorded in VP/SI responding to the reward-predicting cues (red line) in Pavlovian conditioning. (D) Raster plot of a neuron from VP/SI selectively responding to punishment but not reward in the same task.
FIGURE 5
FIGURE 5
Sound calibration and delivery. (A) Components: computer (1), miniUSB-USB A cable (2), Bpod (3), RJ45 cable (4), miniUSB-USB A cable (5), Audio Adaptor Board for Teensy + Teensy USB Development Board + SD card (6), 3.5 mm stereo jack to jack cable (7), Adafruit Audio Amplifier (8), 12V power supply (9), Digikey 8 Ohm Magnetic Speakers (10), EMM-6 Electret Measurement Microphone (11), Male–Female three-pin XLR cable (12), AudioBox iOne (13), USB B -USB A cable (14). (B) Schematic of the setup. A sine wave is generated in Matlab and sent to Bpod, which loads it to the Teensy apparatus as a. wav file. When played by the speakers, the sound is detected by the microphone, delivered to the computer and the dB SPL is read using the TrueRTA software. (C) The dB SPL levels at each frequency before (blue) and after (red) the calibration process. Solid black line indicates the calibration target volume (60 dB SPL).
FIGURE 6
FIGURE 6
Delay measurements. (A) Internal delay. Left, signals were sent from the BNC output port (blue) and the RJ45 output connector for communication with the port interface board (red) directly to the oscilloscope. Right, example signals detected by the oscilloscope. Arrow, measured delay. (B) Delay of visual cue. Left, signals were sent from Bpod to the oscilloscope both directly (blue) and via the port interface board (red). Right, example of the signals detected by the oscilloscope. (C) Delay of sound delivery. Left, signals were sent directly (blue) or via the Teensy board (red). The oscilloscope receives the latter signal from the line out pins of the Teensy slave board. Right, example of the signals detected by the oscilloscope. (D) Delay of reinforcement delivery. Left, signals were sent from the BNC output port (blue) directly to the oscilloscope and to two port interface boards. One port was receiving commands to open and close the water valve (red), while the other was receiving similar input for controlling the air valve (yellow) along with a constant PWM signal (orange). The latter was sent to the oscilloscope throughout a circuit that water or air could close or break, respectively, changing the oscilloscope voltage input. Top right, example of the signals detected by the oscilloscope for water delay measurement. Bottom right, example of the signals detected by the oscilloscope for air delay measurement.
FIGURE 7
FIGURE 7
Temporally precise delivery of stimuli and feedback. (A) Distribution of minimal elapsed time between sending signals to the BNC and RJ45 output of Bpod (mean ± SD, 0.045 ± 0.001 ms). (B) Board delay: distribution of delays between the signals from the BNC output port and the LED output wire terminal of the port interface board (mean ± SD, 0.047 ± 0.003 ms) (C) Delay distribution of sound delivery, between the signals from the BNC output port and the Teensy board (mean ± SD, 6.59 ± 0.9 ms). (D) Delay distribution of air puff delivery (mean ± SD, 3.48 ± 0.02 ms). (E) Delay distribution of water delivery (mean ± SD, 8.61 ± 0.81 ms).
FIGURE 8
FIGURE 8
Custom-made sound-attenuated enclosures. (A) A 50-by-50-by-50 cm sound attenuated box designed in the freely available 3D modeling software SketchUp. (B) Picture of the sound attenuated chamber. (C) Cross-section of the box: from left to right, pyramidal foam, sound absorbing foam, stainless steel mesh and medium-density fiberboard (MDF). (D) Configuration #1: sound attenuation by acoustic insulation board with quartz sand filling and pyramidal foam. (E) Configuration #2: Hanno Protecto 20 foam. (F) Configuration #3: Hanno Protecto 50 foam combined with pyramidal foam. (G) Sound attenuation measurements: pure tones of different pitch were played from speakers outside the enclosure and the dB SPL was measured by a microphone placed inside with the door open or closed.

Similar articles

See all similar articles

Cited by 3 articles

References

    1. Abraham N. M., Guerin D., Bhaukaurally K., Carleton A. (2012). Similar odor discrimination behavior in head-restrained and freely moving mice. 7:e51789. 10.1371/journal.pone.0051789 - DOI - PMC - PubMed
    1. Arabzadeh E., Panzeri S., Diamond M. E. (2006). Deciphering the spike train of a sensory neuron?: counts and temporal patterns in the rat whisker pathway. 26 9216–9226. 10.1523/JNEUROSCI.1491-06.2006 - DOI - PMC - PubMed
    1. Beck S. M., Locke H. S., Savine A. C., Jimura K., Braver T. S. (2010). Primary and secondary rewards differentially modulate neural activity dynamics during working memory. 5:e9251. 10.1371/journal.pone.0009251 - DOI - PMC - PubMed
    1. Brunton B. W., Botvinick M. M., Brody C. D. (2013). Rats and humans can optimally accumulate evidence for decision-making. 340 95–98. 10.1126/science.1233912 - DOI - PubMed
    1. Burgess C. P., Lak A., Steinmetz N. A., Zatka-Haas P., Bai Reddy C., Jacobs E. A. K., et al. (2017). High-Yield methods for accurate two-alternative visual psychophysics in head-fixed mice. 20 2513–2524. 10.1016/J.CELREP.2017.08.047 - DOI - PMC - PubMed

LinkOut - more resources

Feedback