A novel variable delay Go/No-Go task to study attention, motivation and working memory in the head-fixed rodent

Introduction

The use of operant behavioral tasks that utilize the head-immobilized condition in rodents have some advantages over tasks that occur under freely moving conditions by limiting the range of possible non task-relevant behaviors and distractors. While the head-restrained behaving rodent has been examined for many years, the relevant literature is minimal in comparison to similar behavioral tasks involving freely moving animals. Head immobilization is a traditional and common method in electrophysiological studies utilizing behaving monkeys, in which skilled intentional forelimb movements (e.g. grasping) are widely used as conditional responses. Compared with monkey studies, despite a growing number of studies that utilize head-fixed rodents, very few experiments have assessed psychological and physiological processes simultaneously occurring during operant behavioral tasks that require intentional skilled movements (e.g. lever pressing) as conditional responses in the awake, head-restrained rodent^1,2.

The Go/No-Go task, a canonical paradigm in animal behavior and psychology, requires the subject to initiate an operant conditioned response during one stimulus (CS+) and withhold from such a response during the presence of the opposite type of stimulus (CS-)^1,3. Variations of the Go/No-go task have long been applied to electrophysiological, lesion and pharmacological studies^4–7. This type of behavioral task has a high level of construct validity and clinical utility when studying the brain motivation/reward pathways and behaviors such as impulsivity. Impulsivity and motivational deficits are complex behavioral phenotypes implicated in a number of psychiatric disorders including attention deficit hyperactivity disorder (ADHD), mania and depression associated with bipolar disorder, schizophrenia, pathological gambling, borderline personality disorder, and substance abuse³.

The Go/No-Go task we developed allows for simultaneous measurement of the subject’s motivational level, impulsivity and short-term memory. The task consists of four periods: initiation, delay, challenge and reward. First, rats are required to press the lever in response to distinct Go or No-Go tones to initiate a trial (initiation). Performance during this period can be used as a measurement of motivational level. After initiating a trial, a variable delay follows and the rats must maintain information about the cue (tone frequency). After the delay, rats must make a correct Go or No-Go response according to the preceding tones (challenge). A delay period can be used to assess short-term memory and performance during the challenge period can be used as a measure of impulsivity. Finally, a reward tone of the same frequency as the initiation tone is presented following a correct response during challenge period. Data obtained from these studies provide the opportunity for development of novel future pharmacotherapies and behavioral interventions⁸.

The head-immobilized rodent Go/No-Go task we have developed is a procedure that provides the framework for performing repeated electrophysiological recordings and simultaneous molecular and pharmacologic interventions during ongoing operant behavioral tasks. In addition, this technique allows for accurate stereotaxic localization during electrophysiological recordings, daily anesthesia-free recordings in various brain regions, rapid electrode repositioning during chronic recording sessions, and high quality stabilized recordings⁸. With the advent of optogenetics, the head-restrained procedure allows for these recordings to be coupled with cell-type specific silencing and activation, free from electrophysiological artifacts⁹.

Methods

Behavioral conditioning procedure

Acquisition of lever-pressing and lever/tone association

Following successful sucrose consumption under non-contingent reward delivery conditions, subjects were trained to press a right forepaw lever to receive sucrose reinforcement (Figure 1). This stage of operant behavioral training was intended to facilitate the contingent relationship between lever pressing and reward delivery. This phase of training persisted until the subject responded with >1000 lever presses for 2–3 consecutive days. The next phase of training was the Lever/Tone association phase during which subjects learned to lever press in the presence of an audible tone. The two tones randomly presented were either 3000 Hz or 9000 Hz, which corresponded to the same frequencies used in the Go and No-Go constituents of the later stages of the training procedure. This stage of training began with a 5 s tone presentation with 8 µl sucrose delivered for every lever press occurring during the tone (15 s ITI). Subjects were switched to the next stage of training after >70% correct responding during the presence of the tone. The next phase of training was a reduction in the duration of the tone presentation (5 s to 3 s). After achieving >70% correct responding using these parameters, tone duration was finally decreased to 2 s with a single lever press contingent sucrose delivery of 40 µl. The final phase persisted until subjects achieved >80% correct responding for 2 consecutive sessions.

Figure 1. Schematic showing the head-fixed Go/No Go task under a No-Go trial and Go trial.

Lever presses, initiation tone, challenge period, reward tone, and sucrose reward delivery components are shown for both trials. The black boxes represent the relative timing of the rat’s response (lever press) in response to the auditory cues. ITI: inter-trial interval.

Results

Subjects rapidly acquired the ability to consume sucrose delivered non-contingently through a reward delivery nozzle. Following acquisition of sucrose consummatory behavior, subjects quickly learned (<3 days) an operant response by lever pressing (Lever) to receive sucrose reinforcement (Figure 2). On average, subjects reached the specified criteria of >1000 presses/90 minute training session in <3 consecutive daily training sessions (n=24). Next, subjects reliably acquired operant lever pressing behavior during the presence of a tone (Lever/Tone association) to receive sucrose reinforcement (Figure 2). The Lever/Tone association stage of training required an average of 8 training sessions to achieve the criteria of >70% correct responding for two consecutive days (n=20). This requisite stage of training was intended to facilitate the association between tone presentation, lever pressing, and concomitant primary reinforcement. No-Go training (No-Go) followed the Lever/Tone association training (Figure 2). The No-Go constituent of the behavioral task required an average of 22 daily training sessions to achieve criteria of >80% correct challenge period responding (n=14). Finally, subjects were trained to perform the Go constituent (Go) of the Go/No-Go task (Figure 2). This component of the task required an average of 14 daily training sessions to achieve a criteria of >80% correct 2 s challenge period responding for 2 consecutive days (n=13). Raster plots of five consecutive representative No-Go trials (Figure 3A) and histograms of lever pressing and licking during the initiation tone, challenge period, and reward tone are shown (Figure 3B). The histograms represent cumulative responding across a 90 minute training session. Subjects lever pressed in response to an initiation tone and received an initial small sucrose reinforcement (8 µl), withheld from lever pressing (No-Go) during the challenge period, and emitted a final lever press during the final reward tone to receive a larger 40 µl sucrose reinforcement. One example of a failed No-Go trial is shown (4^th trial on the raster plot). The subject emitted a lever press during the challenge period of the No-Go trial, resulting in a failed trial and no final 40 µl sucrose delivery. Raster plots of five consecutive Go trials (Figure 4A) and histograms of lever pressing and licking during the initiation tone, challenge period, and reward tone are shown (Figure 4B). Subjects initiated a trial by lever pressing during the initiation tone, emitted a lever press during the challenge period (Go), and lever pressed during the final reward tone to receive sucrose reinforcement. One example of a failed Go trial is shown (1^st trial on the raster plot). The subject failed to lever press during the challenge period of the Go trial and the final 40 µl sucrose reward was not delivered. Once the specified criteria of >80% correct responding for 2 consecutive days was achieved for No-Go and Go training sessions, subjects (n=10) maintained stable challenge period responding between subsequent training sessions (Figure 5A). Similarly, subjects stably responded during the challenge period within individual No-Go and Go training sessions (Figure 5B). The data set from all rats is presented.

Figure 2. The average and standard error of the mean training sessions required to achieve specified criteria for acquisition of each head-fixed training condition of the behavioral task is shown.

Lever: n=24, Lever/Tone: n=20, No-Go: n=14, Go: n=13.

Figure 3.

Raster plot of five consecutive No-Go trials (A) and histograms of lever pressing and licking during initiation tone, challenge period, and reward tone (B). Green bar: initiation tone and reward tone, orange bar: challenge period, magenta bar: reward delivery, black: lick, blue: lever press. (Histogram bin size=100 ms).

Figure 4.

Raster plot of five consecutive Go trials (A) and histograms of lever pressing and licking during the initiation tone, challenge period, and reward tone (B). Green bar: initiation tone and reward tone, orange bar: challenge period, magenta bar: reward delivery, black: lick, blue: lever press. (Histogram bin size=100 ms). The second peak on the histogram of lever pressing during the challenge period represents lever presses during the final reward tone. Note the emergence of non cued lever presses and licking during the challenge period that represents the idiosyncratic variable initiation during the Go challenge tone and subsequent reward tone delivery under the Go condition.

Figure 5.

Stability of the task performance between (A) and within (B) sessions after reaching criteria. A. Session 0 represents the day when the rats (n=10) achieved the criteria (two consecutive training sessions with >80% (dashed horizontal grey line) correct challenge period responding). B. Within session challenge period responding for 200 trials, divided into 20 trial bins, is shown for the fourth training session after achieving criteria. Error bars indicate the standard error of the mean.

Data availability

figshare: Variable delay Go/No-Go task data, http://dx.doi.org/10.6084/m9.figshare.654019¹⁰