Search

Abstract Machine learning has been applied in recent years to categorize sleep stages (NREM, REM, and wake) using electroencephalogram (EEG) recordings; however, a well-validated sleep scoring automatic pipeline in rodent research is still not publicly available. Here, we present IntelliSleepScorer, a software package with a graphic user interface to score sleep stages automatically in mice. IntelliSleepScorer uses the light gradient boosting machine (LightGBM) to score sleep stages for each epoch of recordings. We developed LightGBM models using a large cohort of data, which consisted of 5776 h of sleep EEG and electromyogram (EMG) signals across 519 unique recordings from 124 mice. The LightGBM model achieved an overall accuracy of 95.2% and a Cohen’s kappa of 0.91, which outperforms the baseline models such as the logistic regression model (accuracy = 93.3%, kappa = 0.88) and the random forest model (accuracy = 94.3%, kappa = 0.89). The overall performance of the LightGBM model as well as the performance across different sleep stages are on par with that of the human experts. Most importantly, we validated the generalizability of the LightGBM models: (1) The LightGBM model performed well on two publicly available, independent datasets (kappa > = 0.80), which have different sampling frequency and epoch lengths; (2) The LightGBM model performed well on data recorded at a lower sampling frequency (kappa = 0.90); (3) The performance of the LightGBM model is not affected by the light/dark cycle; and (4) A modified LightGBM model performed well on data containing only one EEG and one EMG electrode (kappa > = 0.89). Taken together, the LightGBM models offer state-of-the-art performance for automatic sleep stage scoring in mice. Last, we implemented the IntelliSleepScorer software package based on the validated model to provide an out-of-box solution to sleep researchers (available for download at https://sites.broadinstitute.org/pan-lab/resources ).

Abstract Schizophrenia is a heterogeneous psychiatric disorder with a strong genetic basis, whose etiology and pathophysiology remain poorly understood. Exome sequencing studies have uncovered rare, loss-of-function variants that greatly increase risk of schizophrenia [1], including loss-of-function mutations in GRIN2A (aka GluN2A or NR2A, encoding the NMDA receptor subunit 2A) and AKAP11 (A-Kinase Anchoring Protein 11). AKAP11 and GRIN2A mutations are also associated with bipolar disorder [2], and epilepsy and developmental delay/intellectual disability [1, 3, 4], respectively. Accessible in both humans and rodents, electroencephalogram (EEG) recordings offer a window into brain activity and display abnormal features in schizophrenia patients. Does loss of Grin2a or Akap11 in mice also result in EEG abnormalities? We monitored EEG in heterozygous and homozygous knockout Grin2a and Akap11 mutant mice compared with their wild-type littermates, at 3- and 6-months of age, across the sleep/wake cycle and during auditory stimulation protocols. Grin2a and Akap11 mutants exhibited increased resting gamma power, attenuated auditory steady-state responses (ASSR) at gamma frequencies, and reduced responses to unexpected auditory stimuli during mismatch negativity (MMN) tests. Sleep spindle density was reduced in a gene dose-dependent manner in Akap11 mutants, whereas Grin2a mutants showed increased sleep spindle density. The EEG phenotypes of Grin2a and Akap11 mutant mice show a variety of abnormal features that overlap considerably with human schizophrenia patients, reflecting systems-level changes caused by Grin2a and Akap11 deficiency. These neurophysiologic findings further substantiate Grin2a and Akap11 mutants as genetic models of schizophrenia and identify potential biomarkers for stratification of schizophrenia patients.

Sensations from viscera, like fullness, easily become painful if the stimulus persists. Mice lacking α 1GT-type [Ca.sup.2+] channels show hyperalgesia to visceral pain. Thalamic infusion of a T-type blocker induced similar hyperalgesia in wild-type mice. In response to visceral pain, the ventroposterolateral thalamic neurons evoked a surge of single spikes, which then slowly decayed as T type-dependent burst spikes gradually increased. In α 1G-deficient neurons, the single-spike response persisted without burst spikes. These results indicate that T-type [Ca.sup.2+] channels underlie an antinociceptive mechanism operating in the thalamus and support the idea that burst firing plays a critical role in sensory gating in the thalamus., Low voltage-activated (LVA) T-type [Ca.sup.2+] channels play crucial roles in the control of cellular excitability under diverse physiological and pathological processes (1, 2). Recently, studies revealed a novel role of [...]

A crucial pathophysiological issue concerning central neuropathic pain is the modification of sensory processing by abnormally increased low-frequency brain rhythms. Here we explore the molecular mechanisms responsible for such abnormal rhythmicity and its relation to neuropathic pain syndrome. Toward this aim, we investigated the behavioral and electrophysiological consequences of trigeminal neuropathic pain following infraorbital nerve ligations in CaV3.1 T-type Ca2+ channel knockout and wild-type mice. CaV3.1 knockout mice had decreased mechanical hypersensitivity and reduced low-frequency rhythms in the primary somatosensory cortex and related thalamic nuclei than wild-type mice. Lateral inhibition of gamma rhythm in primary somatosensory cortex layer 4, reflecting intact sensory contrast, was present in knockout mice but severely impaired in wild-type mice. Moreover, cross-frequency coupling between low-frequency and gamma rhythms, which may serve in sensory processing,was pronounced inwild-typemice but not in CaV3.1 knockout mice. Our results suggest that the presence of CaV3.1 channels is a key element in the pathophysiology of trigeminal neuropathic pain. [ABSTRACT FROM AUTHOR]

In unconscious status (e.g., deep sleep and anesthetic unconsciousness) where cognitive functions are not generated there is still a significant level of brain activity present. Indeed, the electrophysiology of the unconscious brain is characterized by well-defined thalamocortical rhythmicity. Here we address the ionic basis for such thalamocortical rhythms during unconsciousness. In particular, we address the role of CaV3.1 T-type Ca2+ channels, which are richly expressed in thalamic neurons. Toward this aim, we examined the electrophysiological and behavioral phenotypes of mice lacking CaV3.1 channels (CaV3.1 knockout) during unconsciousness induced by ketamine or ethanol administration. Our findings indicate that CaV3.1 KO mice displayed attenuated low-frequency oscillations in thalamocortical loops, especially in the 1- to 4-Hz delta band, compared with control mice (CaV3.1 WT). Intriguingly, we also found that CaV3.1 KO mice exhibited augmented highfrequency oscillations during unconsciousness. In a behavioral measure of unconsciousness dynamics, CaV3.1 KO mice took longer to fall into the unconscious state than controls. In addition, such unconscious events had a shorter duration than those of control mice. The thalamocortical interaction level between mediodorsal thalamus and frontal cortex in CaV3.1 KO mice was significantly lower, especially for delta band oscillations, compared with that of CaV3.1 WT mice, during unconsciousness. These results suggest that the CaV3.1 channel is required for the generation of a given set of thalamocortical rhythms during unconsciousness. Further, that thalamocortical resonant neuronal activity supported by this channel is important for the control of vigilance states. [ABSTRACT FROM AUTHOR]

The rhythmic motor pathway activation by pacemaker neurons or circuits in the brain has been proposed as the mechanism for the timing of motor coordination, and the abnormal potentiation of this mechanism may lead to a pathological tremor. Here, we show that the potentiation of Cav3.1 T-type Ca2+ channels in the inferior olive contributes to the onset of the tremor in a pharmacological model of essential tremor. After administration of harmaline, 4to 10-Hz synchronous neuronal activities arose from the 10 and then propagated to cerebellar motor circuits in wild-type mice, but those rhythmic activities were absent in mice lacking Cav3.1 gene. Intracellular recordings in brain-stem slices revealed that the Cav3.1-deficient inferior olive neurons lacked the subthreshold oscillation of membrane potentials and failed to trigger 4to 10-Hz rhythmic burst discharges in the presence of harmaline. In addition, the selective knockdown of Cav3.l gene in the inferior olive by shRNA efficiently suppressed the harmaline-induced tremor in wild-type mice. A mathematical model constructed based on data obtained from patch-clamping experiments indicated that harmaline could efficiently potentiate Cav3.l channels by changing voltage-dependent responsiveness in the hyperpolarizing direction. Thus, Cav3.1 is a molecular pacemaker substrate for intrinsic neuronal oscillations of inferior olive neurons, and the potentiation of this mechanism can be considered as a pathological cause of essential tremor. [ABSTRACT FROM AUTHOR]

Small GTPase Rab is a member of a large family of Ras-related proteins, highly conserved in eukaryotic cells, and thought to regulate specific type(s) and/or specific step(s) in intracellular membrane trafficking. Given our interest in synaptic transmission. we addressed the possibility that Rab27 (a close isoform of Rab3) could be involved in cytosolic synaptic vesicle mobilization. Indeed, preterminal injection of a specific antibody against squid Rab27 (anti-sqRab27 antibody) combined with confocal microscopy demonstrated that Rab27 is present on squid synaptic vesicles. Electrophysiological study of injected synapses showed that the anti-sqRab27 antibody inhibited synaptic release in a stimulation-dependent manner without affecting presynaptic action potentials or inward Ca[sup2+] current. This result was confirmed in in vitro synaptosomes by using total internal reflection fluorescence microscopy. Thus. synaptosomal Ca[sup2+]-stimulated release of FM1-43 dye was greatly impaired by intraterminal anti-sqRab27 antibody. Ultrastructural analysis of the injected giant preterminal further showed a reduced number of docked synaptic vesicles and an increase in nondocked vesicular profiles distant from the active zone. These results, taken together, indicate that Rab27 is primarily involved in the maturation of recycled vesicles and/or their transport to the presynaptic active zone in the squid giant synapse. [ABSTRACT FROM AUTHOR]

Thalamocortical in vivo and in vitro function was studied in mice lacking P/Q-type calcium channels (Cav2.1), in which N-type calcium channels (Cav2.2) supported central synaptic transmission. Unexpectedly, in vitro patch recordings from thalamic neurons demonstrated no y-band subthreshold oscillation, and voltage-sensitive dye imaging demonstrated an absence of cortical γ-band-dependent columnar activation involving cortical inhibitory interneuron activity. In viva electroencephalogram recordings showed persistent absence status and a dramatic reduction of γ-band activity. Pharmacological block of T-type calcium channels (Cav3), although not noticeably affecting normal control animals, left the knockout mice in a coma-like state. Hence, although N-type calcium channels can rescue P/Q-dependent synaptic transmission, P/Q calcium channels are essential in the generation of γ-band activity and resultant cognitive function. [ABSTRACT FROM AUTHOR]

Alterations in thalamic T-type Ca2+ channels are thought to contribute to the pathogenesis of absence seizures. Here, we found that mice with a null mutation for the pore-forming α1A subunits of P/Q-type channels (α1A -/- mice) were prone to absence seizures characterized by typical spike-and-wave discharges (SWDs) and behavioral arrests. Isolated thalamocortical relay (TC) neurons from these mice showed increased T-type Ca2+ currents in vitro. To examine the role of increased T-currents in αA -/- TC neurons, we cross-bred α1A 2+ mice with mice harboring a null mutation for the gene encoding α1G, a major isotype of T-type Ca2+ channels in TC neurons. α1A 2+ /α1G -/- mice showed a complete loss of T-type Ca2+ currents in TC neurons and displayed no SWDs. Interestingly, α1A -/- / α1G +/- mice had 75% of the T-type Ca2+ currents in TC neurons observed in α1A +/+ α1G +/+ mice and showed SWD activity that was quantitatively similar to that in α1A -/- /α1G +/+ mice. Similar results were obtained using double-mutant mice harboring the α1G mutation plus another mutation also used as a model for absence seizures, i.e., lethargic (β4lh/lh), tottering (α1Atg/tg), or stargazer (γ2stg/stg). The present results reveal that α1G T-type Ca2+ channels play a critical role in the genesis of spontaneous absence seizures resulting from hypofunctioning P/Q-type channels, but that the augmentation of thalamic T-type Ca2+ currents is not an essential step in the genesis of absence seizures. [ABSTRACT FROM AUTHOR]