Adaptive and pathological inhibition of neuroplasticity associated with circadian rhythms and sleep

Abstract

The circadian system organizes sleep and wake through imposing a daily cycle of sleep propensity on the organism. Sleep has been shown to play an important role in learning and memory. Apart from the daily cycle of sleep propensity, however, direct effects of the circadian system on learning and memory also have been well documented. Many mechanistic components of the memory consolidation process ranging from the molecular to the systems level have been identified and studied. The question that remains is how do these various processes and components work together to produce cycles of increased and decreased learning abilities, and why should there be times of day when neural plasticity appears to be restricted? Insights into this complex problem can be gained through investigations of the learning disabilities caused by circadian disruption in Siberian hamsters and by aneuploidy in Down's syndrome mice. A simple working hypothesis that has been explored in this work is that the observed learning disabilities are due to an altered excitation/inhibition balance in the CNS. Excessive inhibition is the suspected cause of deficits in memory consolidation. In this article we present the evidence that excessive inhibition in these cases of learning disability involves GABAergic neurotransmission, that treatment with GABA receptor inhibitors can reverse the learning disability, and that the efficacy of the treatment is time sensitive coincident with the major daily sleep phase, and that it depends on sleep. The evidence we present leads us to hypothesize that a function of the circadian system is to reduce neuroplasticity during the daily sleep phase when processes of memory consolidation are taking place.

Figure 1

Sleep intensification following sleep deprivation improves novel object recognition memory in Ts65Dn mice. The protocol for this experiment is shown in panel A. Mice were trained early in the second half of the light phase and were tested 24 hrs. later. One cohort of mice was sleep deprived by gentle handling for 4 hrs. prior to training. Panel B shows the NOR performance in 2N and Ts65Dn mice that were not sleep deprived (baseline) and that were sleep deprived prior to training (SD). Unpublished data from Colas et al. presented as means +/− SEM. ** indicates P<.01 based on t-test for paired samples.

Figure 2

PTZ improves novel object recognition memory in Ts65Dn mice only if the animals are dosed during the light phase of the daily cycle. A. Treatment protocol was 2 weeks of daily ip injections of PTZ at 0.3 mg/kg or of saline. At least 1 week was allowed between end of treatment and testing. B. NOR test results for 2–3 month old mice treated with PTZ (0.3 mg/kg) or saline. Two objects were used, and discrimination index is the time spent with the novel object divided by time spent with novel and familiar objects. Gray bars are training and black bars are novel recognition results 24 hrs. later. C. NOR test results from mice receiving PTZ treatment (0.3 mg/kg) or saline during the dark phase. Data from Colas et al. 2013 presented as means +/− SEM. * indicates P<.05,** indicates P<.01 based on t-test for paired samples.

Figure 3

PTZ rescued memory in circadian-arrhythmic hamsters. The loss of circadian rhythms impaired spatial working memory as assessed by alternation behavior in a T-maze (left panel). The ability to discriminate a familiar object from a novel one was impaired in a novel object recognition test (right panel). Test performance of arrhythmic hamsters (ARR) after a 10-day injection regimen of PTZ was restored to levels observed in normal entrained animals (ENT). PTZ had no effect in ENT hamsters. * P < 0.05, ** P < 0.01, compared to random chance performance. Data from Ruby et al. 2013.

Figure 4

Effect of sleep fragmentation on novel object recognition memory. Mice were trained in the NOR task early in the light phase. Then for 4 hrs they were left undisturbed, were totally sleep deprived, or were subjected to different protocols of sleep fragmentation administered by brief trains of optogenetic stimulation of hypocretinergic cells at 30, 60, 120, or 240 sec. intervals. For the remaining 20 hrs they were allowed to sleep ad lib. until testing 24 hrs after training. Significant NOR memory was seen only in mice that slept ad lib. for 24 hrs. or had their early sleep fragmented at 120 or 240 sec. intervals. Data from Rolls et al. 2011 are presented as means +/− SEM. ***indicates P<.001.

Figure 5

The effects of introducing the Conditioning Stimulus from a foot shock conditioning paradigm during subsequent sleep. A. Mice received foot shocks paired with an odor (the CS) at the beginning of the light phase. During the sleep phase beginning 24 hrs later the mice were re-exposed to the CS or to a control odor during episodes of NREM sleep. They were then tested 48 hrs. after conditioning in a neutral context by reintroduction of the CS or the control odor and measurements of time spent freezing were recorded. The re-exposure to the CS during sleep intensified the subsequent fear response. B. The experiment was repeated with a higher stimulus intensity, but the mice received bilateral injections of anisomycin or vehicle into the amygdala prior to the sleep phase. Responses to the CS in a neutral context were tested 48 hrs following the conditioning. Animals that received the anisomycin and were re-exposed to the CS during sleep showed a reduction in their subsequent fear response to the CS. Data from Rolls et al. 2013.