Sleep- related electrophysiological activity of cortical cultures on MEA

• ¹ Fondazione Istituto Italiano di Technologia, Italy
• ² Università di Genova, Italy

Motivation Neuronal assemblies plated on MEAs show spontaneously synchronized, low frequency firing patterns, which could resemble the slow wave oscillations that characterize non-rapid eye movement (NREM) sleep in vivo [1]. The lack of the awake state counterpart limited the investigation of the main physiological aspects of sleep. To this end, in a previous work we stimulated our in vitro neuronal networks using the cholinergic agonist Carbachol (CCh, 20 µM) to suppress the sleep features. We observed for the first time that CCh treatment affected both the high and the low frequency components of the signal, causing a suppression of the classical sleep-like properties of activity [2]. Starting from this finding, we were then interested in understand how the electrophysiological signal changed in pathologies that affect sleep. Prader-Willi syndrome (PWS) is a rare neurodevelopmental disorder that is associated with a paternally-expressed genomic imprinting defects within the human chromosome region 15q11-13. One of the candidate genes is the small nucleolar ribonucleic acid-116 (SNORD116). PWS patients are often affected by sleep-wake disturbance associated with alteration during REM sleep, including abnormalities in theta waves. The same identical sleep/EEG defects were found using mutant mice (PWS) carrying a deletion in the gene Snord116 at the orthologous locus [3]. Then we wanted to answer one main question: can our model recapitulate the essential features in a pathological model of sleep? More specifically, it is possible to observe the same abnormalities in a simplified and accessible model in vitro? We performed preliminary experiments using cortical cultures of PWS mice and we applied Carbachol (CCh, 20 µM) to suppress the sleep waves. Interestingly, we found the same difference in the theta waves upon CCh administration, suggesting the idea that our model can be used to study and manipulate sleep properties in a very controlled way. Material and Methods Cell cultures prepared from embryonic mice at gestational day 18 were plated onto 60-channel MEAs (Multichannel Systems, MCS, Reutlingen, Germany) previously coated with poly-L-lysine in buffer borate to promote cell adhesion (final density around 1200 cells/mm2) (Figure 1A). We recorded the electrophysiological activity of four WT cultures and two PWS cultures. The experimental protocol adopted for these experiments included 2 hours of recording in culture solution defined as a basal condition. Then we chemically stimulated our cultures with CCh (20 µM) for 4 hours (Figure 2A). Starting from the raw data (i.e. wide band signal) we performed a dual process to analyse both MUA (Multi-Unit Activity, f>300 Hz) and LFP (Local Field Potential, f<300 Hz) (Figure 1B). In order to capture only MUA activity, we high pass filtered the raw signal. Once spikes (i.e. single over threshold peaks) and bursts (i.e. groups of tightly packed spikes) activity were detected from MUA recordings, we computed the following parameters: Mean Firing Rate, MFR [spike/sec], Inverse Burst Ratio, IBR, (percentage of spike outside the burst) and Burstiness Index, BI, (index of the burstiness level of the network) (Figure1 B, light blue boxes in the diagram) [4]. To select the LFP components, we low pass filtered the raw data between 1-300 Hz (Figure 1B, light grey boxes in the diagram). We then computed the power spectral density of the decimated signal (sampling frequency 1 kHz) (μV²/Hz), using the Welch method (Windows=5s, overlap=50%). We only considered the lower frequency bands of the signal, which are of particular interest for studying the sleep-wake cycle, in particular delta (1-4 Hz), theta (4-11 Hz), beta (11-30 Hz) and gamma (30-55 Hz) bands. Results We performed experiments using PWS cultures plated on MEA, according to the protocol depicted in Figure 2A. We normalized each experiment with respect to the mean of the selected parameter (i.e. MFR, IBR and BI) during the basal recording. Regarding the WT cultures (Figure 2B, blue line) the activity level, evaluated by means of the MFR, increased after CCh administration with respect to the basal condition. At the same time, CCh application caused an increased number of isolated spikes (i.e. a higher level of the IBR) and a decrease of the BI with respect to the basal phase, indicating a loss of bursting activity and then of synchronicity. Indeed, the results obtained on WT cultures treated with CCh are in line with the previous results obtained by using cortical cultures of embryonic rat coupled to MEA [5]. PWS cultures (Figure 2B, red line) showed a decrease of MFR only in the first hours after the CCh treatment. The number of isolated spikes (i.e. IBR) and BI did not show any variation upon CCh administration with respect to the their basal condition. When comparing the behaviour of WT vs PWS cultures, we observed the following. WT cultures showed higher values of MFR and IBR with respect to the PWS cultures during CCh administration. Conversely, the BI of WT cultures was lower compared to the PWS cultures. This suggests a reduced effect of CCh in PWS cultures, but more experiments and statistical analyses are necessary to confirm our preliminary results. The analysis of the LFP in WT and PWS mice revealed that CCh application caused a strong suppression of all main waves (i.e. delta, theta and, partly, beta), which characterize the classical sleep-wake-cycle. Interestingly, we found a difference between the two set of cultures (i.e. WT vs PWS) only in the theta waves during CCh application (Figure 2 C). These results are in line with a previous study that revealed abnormalities in theta waves during REM sleep [6]. Conclusion In this study, we presented electrophysiological evidence that primary cortical cultures, usually displaying synchronized low-frequency firing patterns under spontaneous conditions, are able to encompass some essential features of sleep also in a pathological model. We presented preliminary experiments and analyses suggesting that our results can replicate those obtained from in vivo animals and PWS patients. We need to increase the number of experiments in order to assess the reproducibility of the results. In conclusion, MEA recordings coupled to cortical cultures seem to represent a possible model to investigate the essential features of sleep in both physiological and