Sleep deprivation in rats: effects on learning and cognitive flexibility

Cathalijn Leenaars
The Netherlands Institute for Neuroscience, Amsterdam

A lack of sleep has very clear consequences for cognition; sleep deprivation induces cognitive impairments. However, the mechanism underlying these impairments remains elusive. This dissertation describes a number of experiments that contribute to the understanding of the mechanism(s) responsible for cognitive impairments after sleep disturbance.
The introduction of this thesis starts with a resume of the general knowledge and hypotheses on sleep, and its role in cognitive functioning. It includes a description of the behavioural and electroencephalographic characteristics that we can observe during sleep, which are used to define sleep and distinguish it from wake behaviour. This is followed by the general hypotheses on why we sleep, including sleep's contribution to cognitive functioning, and the specific cognitive domains mostly affected by sleep loss. It provides an overview of common human sleep disorders, which can result in cognitive impairment, and finishes with a description of rodent models for both sleep deprivation and cognitive performance.
Chapter 2 (Leenaars et al., 2011) introduces a new method for inducing sleep deprivation in rats, based on variable forced locomotion. Contrary to some other methods of forced locomotion (e.g. Roman et al., 2006), this method does not induce significant stress, as indicated by the observation that corticosterone levels did not exceed the levels normally seen during the 24-hour day. Moreover, the method did not have the drawback of potential confounding of experimental results by an increase in locomotor activity, as may be the case in some other methods. When our method was applied for 12h of sleep deprivation during the light phase, activity levels did not exceed those normally seen during undisturbed conditions.
When testing behaviour in a sleep-deprived state, other possible confounders have to be addressed as well. Notably, the effects of sleep deprivation on a specific cognitive domain may depend on nonspecific cognitive effects that affect performance on the task of interest. For example, the motivation to "work" for a reward may be decreased, and fatigue may slow motor functioning. These potential problems were investigated using a task on which rats show vast levels of lever pressing to receive food rewards, which makes this task highly sensitive to decreases in motivation and motor impairment. Potential decreases in motivation were limited by imposing a food restriction to 12g/rat/day (Leenaars et al., 2011).
Chapter 3 (Leenaars et al., 2012b) describes the modelling of one sleepless night and one night of disturbed sleep in humans, with 12h of inactive-phase sleep deprivation or sleep disruption in rats. It tests the effect of this sleep disruption on cognitive flexibility and introduces a new switch-task. While 12h of total sleep deprivation during the light (inactive) phase decreases accuracy on switch-task performance, 12h of repetitive sleep disturbance during the inactive phase does not alter task-switching.
Chapter 4 (Leenaars et al., 2013) describes the impairment in instrumental learning; the simple association between lever pressing and food reward, after 3h of active phase nap-prevention. EEG was measured before and between task performance. Learning is accompanied by an increase in REM sleep. Baseline sleep parameters do not predict subsequent individual differences in learning abilities.
In chapter 5 (Leenaars et al., 2012a), both 12h of inactive-phase sleep deprivation (as a model for one sleepless night) and 3h of active-phase nap prevention did not disturb performance on a different cognitive task: spatial reversal learning. Total sleep deprivation for 12h during the inactive phase does not impair the acquisition of a spatial reversal, and 3h of nap-prevention during the active phase does not impair the consolidation of reversal learning. This indicates that also in rats, sleep-related cognitive deficits are not generalized but limited to certain cognitive domains.
In chapter 6 (Leenaars et al. 2012c), rats were exposed to 5 weeks of non-rotating “shiftwork”, comparable to human night-shifts. They showed no learning deficits on an instrumental learning task (the same task as used in chapter 4) in their 5th week on this protocol, which shows that rats may somehow habituate to regular sleep deprivation for 8h per day on 5 days per week (both in the active and in the inactive phase). Furthermore, the undisturbed control groups in this study demonstrate that instrumental learning is similar during the active and the inactive phase.

Concluding remark
Although the effects of sleep deprivation on cognition and weight are regularly overrated in both scientific and popular literature, sleep is important, and further research is essential to find methods that can help people suffering from the consequences of bad sleep.

Leenaars CH, Dematteis M, Joosten RN, Eggels L, Sandberg H, Schirris M, Feenstra MGP, Van Someren EJW, 2011. A new automated method for rat sleep deprivation with minimal confounding effects on corticosterone and locomotor activity. J Neurosci Methods 196: 107-117. (Chapter 2)
Leenaars CH, Girardi CE, Joosten RN, Lako IM, Ruimschotel E, Hanegraaf MA, Dematteis M, Feenstra MG, Van Someren EJ, 2013. Instrumental learning: An animal model for sleep dependent memory enhancement. J Neurosci Methods. In press. doi: 10.1016/j.jneumeth.2013.04.003. [Epub ahead of print] (Chapter 4)
Leenaars CH, Joosten RN, Kramer M, Post G, Eggels L, Wuite M, Dematteis M, Feenstra MG, Van Someren EJ, 2012a Spatial reversal learning is robust to total sleep deprivation. Behav Brain Res. 230(1):40-47.
Leenaars CH, Joosten RN, Zwart A, Sandberg H, Ruimschotel E, Hanegraaf M, Dematteis M, Feenstra MG, Van Someren EJ, 2012b. Switch- task performance in rats is disturbed by 12h of sleep deprivation but not by 12h of sleep fragmentation. Sleep. 35(2):211-221 (Chapter 3)
Leenaars CH, Kalsbeek A, Hanegraaf MA, Foppen E, Joosten RN, Post G, Dematteis M, Feenstra MG, van Someren EJ, 2012c. Unaltered instrumental learning and attenuated body-weight gain in rats during non-rotating simulated shiftwork. Chronobiol Int. 29(3):344-355. (Chapter 6)
Roman V, Hagewoud R, Luiten PG, Meerlo P, 2006. Differential effects of chronic partial sleep deprivation and stress on serotonin-1A and muscarinic acetylcholine receptor sensitivity. J Sleep Res 15: 386-394