Learning and memory are potent metrics in studying either developmental, disease-dependent, or environmentally induced cognitive impairments. Most cognitive assessments require specialized equipment and extensive time commitments. However, the shuttle box assay is an associative learning tool that utilizes a conventional gel box for rapid and reliable assessment of adult zebrafish cognition.
Cognitive deficits, including impaired learning and memory, are a primary symptom of various developmental and age-related neurodegenerative diseases and traumatic brain injury (TBI). Zebrafish are an important neuroscience model due to their transparency during development and robust regenerative capabilities following neurotrauma. While various cognitive tests exist in zebrafish, most of the cognitive assessments that are rapid examine non-associative learning. At the same time, associative-learning assays often require multiple days or weeks. Here, we describe a rapid associative-learning test that utilizes an adverse stimulus (electric shock) and requires minimal preparation time. The shuttle box assay, presented here, is simple, ideal for novice investigators, and requires minimal equipment. We demonstrate that, following TBI, this shuttle box test reproducibly assesses cognitive deficit and recovery from young to old zebrafish. Additionally, the assay is adaptable to examine either immediate or delayed memory. We demonstrate that both a single TBI and repeated TBI events negatively affect learning and immediate memory but not delayed memory. We, therefore, conclude that the shuttle box assay reproducibly tracks the progression and recovery of cognitive impairment.
Learning and memory are routinely used as metrics of cognitive impairment, which happens due to aging, neurodegenerative disease, or injury. Traumatic brain injuries (TBIs) are the most common injury that results in cognitive deficits. TBIs are of growing concern because of their association with several neurodegenerative disorders, such as frontotemporal dementia and Parkinson's disease1,2. In addition, the increased beta-amyloid aggregations observed in some TBI patients suggest that it may also be associated with the development of Alzheimer's disease3,4. TBIs are often the result of blunt-force trauma and span a range of severities5, with mild brain injuries (miTBI) being the most common. However, miTBIs are often unreported and misdiagnosed because they result in minor cognitive impairments for only a short period, and the injured individuals usually recover fully6. In contrast, repeated miTBI events have been a growing concern because it is highly prevalent in young and middle-aged adults, can accumulate over time7, can impair cognitive development, and exacerbate neurodegenerative diseases1,2,3,4,5, similar to individuals who experience either a moderate or severe TBI8.
Zebrafish (Danio rerio) is a useful model for exploring a variety of topics in neuroscience, including the ability to regenerate lost or damaged neurons throughout the central nervous system9,10,11,12,13. Neural regeneration was also demonstrated in the telencephalon, which contains the archipallium in the dorsal-inner region. This neuroanatomical region is analogous to the hippocampus and is likely required for cognition in fish and for the short-time memory in humans14,15,16. Furthermore, zebrafish behavior has been extensively characterized and cataloged17. Learning has been studied through various techniques, including habituation to the startle response18, which can represent a rapid form of non-associative learning when performed in short blocks and with attention to the rapid decay time19. More complex tests of associative learning, such as T-boxes, plus-mazes, and visual discrimination20,21 are used but often are time-consuming, require days or weeks of preparation, and rely on shoaling or positive reinforcement. Here, we describe a rapid paradigm to assess both associative learning and either immediate or delayed memory. This shuttle box assay uses an aversive stimulus and negative reinforcement conditioning to assess cognitive deficits and recovery following blunt-force TBI. We demonstrate that undamaged control adult zebrafish (8-24 months) reproducibly learn to avoid the red light within 20 trials (<20 min of assessment) in the shuttle box, with a high degree of consistency across observers. Additionally, using the shuttle box we demonstrate that learning and memory abilities across adult (8-24 months old) are consistent and are useful for assaying cognition with significant impairments between either different TBI severities or repeated TBI. Furthermore, this method could be rapidly employed as a metric to track a wide range of disease progressions or efficacy of drug interventions impacting maintenance or recovery of cognition in adult zebrafish.
Here, we provide an instructional overview of a rapid cognitive assessment that can examine both complex associative learning (section 1) and memory in terms of both immediate and delayed memory.This paradigm provides an assessment of the short and long-term memory of a learned associative cognitive task (section 2).
Zebrafish were raised and maintained in the Notre Dame Zebrafish facility in the Freimann Life Sciences Center. The methods described in this manuscript were approved by the University of Notre Dame Animal Care and Use Committee (Animal Welfare Assurance Number A3093-01).
1. Shuttle box learning paradigm (Figure 1A)
NOTE: The learning paradigm provides a rapid assessment of cognition regarding associative learning.
2. Memory paradigm (Figure 1A)
NOTE: This paradigm provides an assessment of the short and long-term memory of a learned associative cognitive task.
The learning paradigm, outlined in the protocol and schematic (Figure 1), provides a rapid assessment of cognition with respect to associative learning. In addition, this paradigm has a high level of stringency, by defining learning as a repeated and consistent display of 5 consecutive positive trials. This paradigm is also applicable to a range of ages and injuries. Undamaged fish at 8 months (young adult), 18 months (middle-aged adult), and 24 months (elderly adult) required a similar number of trials to learn the behavior of avoiding the red light (Undamaged 8 m: 15.28 ± 4.92 trials, 18 m: 17.66 ± 5.5 trials, 24 m: 16.2 ± 4.79 trials, 8 m vs. 18 m p=0.92, 8 m vs. 24 m p=0.98, 18 m vs. 24 m p=0.97, Figure 2A). We also utilized a severe blunt-force traumatic brain injury (sTBI) model22 and observed that fish at different ages required similar number of trials to master the assay across 1-5 days post-injury (dpi; 8 m vs 18 m, p=0.09, 8 m vs 24 m, p=0.96, 18 m vs 24 m, p=0.12, Figure 2A). At Day 1 following sTBI, fish of all ages (8, 18, and 24 m) required a similar number of trials to learn the behavior (8 m: 73.3 ± 9.45 trials, 18 m: 79.33 ± 6.35 trials, 24 m: 68.25 ± 6.65 trials, 8 m vs. 18 m p=0.71, 8 m vs. 24 m p=0.76, 18 m vs. 24 m p=0.28, Figure 2A) and they were all significantly greater than the undamaged controls (p<0.01). Collectively, these data demonstrate that the shuttle box can be utilized to examine injury-induced cognitive deficits across age ranges and suggest that adult zebrafish can recover cognitively following blunt-force injury.
Because repeated miTBI events can increasingly impair cognitive function, we used the shuttle box assay as a metric to track dose-dependent progression using repetitive TBI. We employed this assay to assess learning following a miTBI blunt force injury22 that is repeated daily for the different lengths of time. As previously observed, undamaged fish rapidly mastered the shuttle-box achieving 5 consecutive positive trials in 16.4 ± 3.5 trials (Figure 2B). One day following a single miTBI, fish display a significant increase in the number of trials to learn the behavior (40.25 ± 12.65 trials, p<0.05, Figure 2B). This deficit increased after 2 miTBI events (48 ± 14.9 trials) and was further elevated after 3 miTBI injuries (56.63 ± 12.75 trials, Figure 2B). Additionally, we observed a significant increase in cognitive impairment between miTBI fish which received a singular injury and 3 injuries (p<0.05).
We also examined how memory was affected following repeated miTBI events using the protocol for immediate and delayed memory paradigms (Figure 1A). Naïve undamaged fish were given a training period and an initial testing period, after which a portion of fish were injured for immediate memory and others were returned to the fish facility for 4 days to access delayed memory (Figure 2C). Undamaged fish exhibit a slight increase in the percent difference of successful trials in both immediate memory (6.22% ± 4.7%) and delayed memory (6.13% ± 5.57%) relative to the initial testing period. We, then examined the effect of multiple blunt-force TBI events had on memory. Significant deficits were observed following miTBI in immediate memory, but not in delayed memory. Following a single miTBI, fish displayed significant immediate memory deficits (-26.77% ± 8.93%) compared to undamaged fish (p<0.0001, Figure 2C). This trend continued with repeated injury with increasing deficits following both 2x miTBI (-37.42% ± 10.01%) and 3x miTBI (-39.71% ± 11.39%). Furthermore, we observed a similar dose-effect between fish treated with a single (1x) miTBI and 3x miTBI (p<0.05, Figure 2C). These data suggest that learning and memory is reduced in miTBI fish with the increasing number of injuries, significantly increasing the deficit and the shuttle box assay and protocols described above are sensitive enough to detect these differences.
Figure 1: The Shuttle Box Assay. (A) Instructional overview of the learning and memory paradigms for cognitive assessment. (B) Schematic of a converted large DNA gel box for the shuttle box assay. (C,D) Graphical representation of stimuli application during trials. Please click here to view a larger version of this figure.
Figure 2: Zebrafish display cognitive deficits following blunt-force TBI. (A) Following sTBI, zebrafish at 8, 18, and 24-months of age exhibit learning deficits that are not significantly different between age groups. Significant increases in the number of trials to learn the shuttle box paradigm compared to age-matched controls were observed at 1 dpi returning to undamaged levels by 4-5 dpi. (B,C) Repeated miTBI fish displayed both learning (B) and memory (C) deficits in a dose-dependent manner. The mean ± SEM is plotted in A and B, while the mean ± Standard deviation is plotted in C. Each data point on all three graphs represents a single adult zebrafish. Statistical analyses were performed with either a One-Way or Two-Way ANOVA followed by a Tukey post-hoc test. # p<0.05, ## p<0.01. Please click here to view a larger version of this figure.
Cognitive impairment can significantly and negatively impact the quality of life. Because of the increased visibility and occurrence of concussions and traumatic brain injuries throughout the population, it is important to understand how they cause cognitive impairment and how the damage can be minimized or reversed. For these reasons, model organisms that can be tested for cognitive decline play a critical role in these studies. Rodents have long been the primary model to investigate neurobehavior and cognition, however, zebrafish have emerged as a useful model with numerous distinct behaviors to investigate a range of developmental, age-related, and acquired cognitive deficits17,20,23,24,25,26. Various methods to assess cognition have been utilized from one-dimensional learning in the form of habituation, to complex learning and spatial memory, novel object and location recognition, and decision making18,19,20,21,27,28. However, these cognitive tests are limited to testing non-associative cognition or require a complex set-up, financial investment in equipment, or an extensive time commitment before tests can be performed. In contrast, the shuttle box and the learning and memory paradigms described here utilize a complex associative learning assay that is cost-effective, a rapidly assessed, and easily employed by a novice investigator. Most importantly, consistent with the other cognitive tests, our assay demonstrates that undamaged fish rapidly learn the associative task and can memory the task days later without intermittent training29.
The adaptability of the assay provides avenues to investigate various time points of learning and memory as a metric of disease progression or mechanistic interventions. There are two primary features of the assay. First, the method is simple. The assay is quickly set up and has clear and distinct end points with respect to successful and failed trials, making it accessible to a range of investigators. We found that because of the simplicity of this assay, there is very little troubleshooting needed to use the shuttle box successfully. Second, the assay is extremely quick in comparison to other cognitive exams, which provides flexibility or the ability to examine a large number of fish rapidly in a single day. The time to assess learning is at a minimum 19.75 min (Figure 1), with the fish requiring 15 minutes to acclimate to the shuttle box (determined by tank exploration), followed by a single failed trial (15 s light stimulus, 15 s aversion stimulus, 30 s between trials) and 5 immediate and consecutive positive trials (<15 s light stimulus). In practice, we observed that undamaged fish require 6-30 trials (19.75 min-43.75 min), while in extreme cases (following a severe blunt-force trauma), the most severe deficits can require 100 trials (113.75 min). Memory studies are also rapidly performed. Following the protocol outline, the minimum time necessary for acclimation, training, and initial testing is 67.5 min (15 min acclimation, 25 iterations of light and shock for 15 s, 30 s rest between trials, and repeat for initial testing without the adverse stimuli). While retesting either immediate or delayed memory requires only 33.75 min (15 min acclimation, 25 iterations of only light stimulus for 15 s, and 30 s rest between trials), regardless of injury, treatment, or cognitive deficit.
When assessing neurobehavior, various paradigms utilize either positive or adverse stimuli. Positive stimuli in the form of food or social interaction, often used in classical T-box mazes, can aid in a strong response of a learned task. However, assays utilizing positive association do so at the expense of time. In contrast, while conditioning in response to an adverse stimulus provides a rapid association and strong behavioral response, it is at the expense of the adverse stimulus. Undamaged fish often learn the shuttle box assay quickly and are therefore subjected to a minimal number of shocks, and as a result seem to have no adverse events. However, neurologically compromised fish (TBI), with severe cognitive deficits, require a significant number of trials and electrical shocks. These multiple shocks have been observed to occasionally result in tonic-clonic seizures. Any fish experiencing a tonic-clonic seizure while within the shuttle box should be immediately removed and ethically euthanized. All trials for the euthanized fish, up to and including the seizure event, should be excluded in any statistical analysis. Furthermore, it is worth noting that electrical shock to a neurologically damaged subject could impose unintended differences between damaged fish that are and are not resulting from the shuttle box. For that reason, we suggest all fish subjected for neurobehavior assessment should not be used for any other quantitative metric (serum biomarker, IHC, etc.). It is also important to understand that this method of learning is based on a visual stimulus and is not appropriate for damage that may compromise visual circuits, as it will confound the results.
Our results demonstrate that following blunt-force TBI, zebrafish exhibit a rapid cognitive deficit that results in increased trials to master an associative task in the shuttle box assay. Similar immediate deficits are seen in rodent models of TBI, however these deficits can diminish, they often persist and remain significant30. In contrast, zebrafish display cognitive recovery within 7 days following injury. The regenerative capacity of the adult zebrafish is well documented9,10,11,12,13,14,15, with known neurogenic niches in the ventricular/subventricular zones of the telencephalon31,32. The cognitive recovery observed in our assay following TBI provides insight into needed exams to identify if these neurogenic niches are stimulated and play a role in tissue and cognitive recovery.
In conclusion, the shuttle box provides a rapid assessment of cognition in regard to associative learning and memory. The assay utilizes minimal and conventual equipment and is technically simple. Future applications could be utilized to assess genetic and pharmacological interventions to neurologically insulted fish in regard to neuroprotection as well as other injury paradigms or neurodegenerative models.
The authors have nothing to disclose.
The authors would like to thank the Hyde lab members for their thoughtful discussions and the Freimann Life Sciences Center technicians for zebrafish care and husbandry. This work was supported by the Center for Zebrafish Research at the University of Notre Dame, the Center for Stem Cells and Regenerative Medicine at the University of Notre Dame, and grants from National Eye Institute of NIH R01-EY018417 (DRH), the National Science Foundation Graduate Research Fellowship Program (JTH), LTC Neil Hyland Fellowship of Notre Dame (JTH), Sentinels of Freedom Fellowship (JTH), and the Pat Tillman Scholarship (JTH). Figure 1 made with BioRender.com.
Flashlight | Ultrafire | 9145 | |
Instant Ocean | Instant Ocean | SS15-10 | |
Large DNA Gel Box | Fisher Scientific | FB-SB-1316 | Shuttle Box |
Power Supply | Fisher Scientific | FB-105 |