It can take several months, or even years, for the brain of a young animal to develop and refine the complex neural networks which underpin cognitive abilities such as memory, planning, and decision making. While the properties that support these functions have been well-documented, less is known about how they emerge during development.

Domínguez, Ma, Yu et al. therefore set out to determine when exactly these properties began to take shape in mice, using lightweight nets of electrodes to record brain activity in sleeping newborn pups. The nets were designed to avoid disturbing the animals or damaging their fragile brains.

The recordings showed that patterns of brain activity similar to those seen in adults emerged during the first couple of weeks after birth. Just before, however, the brains of the pups went through a brief period of reduced activity: this lull seemed to mark a transition from an immature to a more mature mode of operation. After this pause, neurons in the mouse brains showed coordinated patterns of firing reminiscent of those seen in adults. By monitoring the brains of human babies using scalp sensors, Domínguez, Ma, Yu et al. showed that a similar transition also occurs in infants during their first few months of life, suggesting that brains may mature via a process retained across species.

Overall, the relative lull in activity before transition may mark when neural networks gain mature properties; in the future, it could therefore potentially be used to diagnose and monitor individuals with delayed cognitive development.

Multiple cognitive functions emerge rapidly during early development. Neural networks enable precise spatiotemporal coordination of synaptic and cellular activity in mature brain functions (Fries et al., 1997; Maingret et al., 2016; Peyrache et al., 2009; Khodagholy et al., 2017). How immature neural networks develop into their mature form needed for the complex computations underlying cognition remains poorly understood. The first organized and predominant pattern of neural activity that appears in cortical circuits across species is a spindle-like oscillation (10–20 Hz) that occurs intermittently on a background of relative neural inactivity. Known as spindle bursts in rodents and delta brushes in humans, this immature network activity is commonly triggered by peripheral stimuli (An et al., 2014; Hanganu et al., 2006; Khazipov et al., 2004). They demarcate cortical columns or pre-columns and have been linked to neuronal survival (Golbs et al., 2011), establishing sensory ensembles, and critical period plasticity (Khazipov et al., 2004; Colonnese et al., 2010; Khazipov et al., 2013; Winnubst et al., 2015). Spindle bursts are characteristic of the first postnatal week of rodent development, and delta brushes disappear shortly after term in human neonates (Torres and Anderson, 1985), emphasizing their transient role in network maturation. In contrast, mature cortex exhibits perpetual, complex patterns of neural activity that appear and interact across a wide range of frequencies (Tort et al., 2009; Fujisawa and Buzsáki, 2011; Bosman et al., 2012). The organization of this activity creates precise spatiotemporal windows for neural synchronization, enabling plasticity processes and generation of neural sequences (Masquelier et al., 2009; Geisler et al., 2010; Foster and Wilson, 2006). Such network properties facilitate information processing, and resultant activity patterns have been causally linked to cognitive processes from stimulus perception to learning and memory (Fries et al., 1997; Maingret et al., 2016; Peyrache et al., 2009; Khodagholy et al., 2017). Understanding when and how these properties develop in the immature brain is critical given that delay or failure to express mature brain activity is a strong risk factor for subsequent impaired cognition (Khazipov et al., 2004; Anderson et al., 1985; Seelke and Blumberg, 2008; Holmes and Lombroso, 1993).

We hypothesized that emergence of these advanced neural network properties could be heralded by the disappearance of immature activity patterns. During this developmental epoch, cortical microcircuits are still undergoing dramatic changes in anatomical connectivity (Khazipov et al., 2004) and functional connectivity between and within cortical layers is initiated (Bureau et al., 2004) in part through an abrupt increase in synaptogenesis in superficial layers (Naskar et al., 2019). Robust GABA-mediated fast inhibition also appears (Daw et al., 2007; Goldberg et al., 2011) as inhibitory network influences of interneurons increase (Modol et al., 2020; Tuncdemir et al., 2016; Favuzzi et al., 2019). These local circuit changes are paralleled with the gradual ingrowth of subcortical neurotransmitters that assist in setting brain state changes. Together, such modifications could prime a shift in network operating mode, setting the stage for re-emergence of a different oscillation that occupies a similar frequency band, thalamocortical sleep spindles. In contrast to spindle bursts, these oscillations are characteristic of the offline state of non-rapid eye movement (NREM) sleep only, and they facilitate memory consolidation (Maingret et al., 2016; Latchoumane et al., 2017).

To investigate this hypothesis, we examined neural network dynamics across early rodent and human development. We targeted epochs spanning disappearance of immature activity and emergence of mature sleep patterns in each species. For rodents, we developed conformable, minimally invasive, high-resolution electrocorticography arrays coupled with high-density implantable probes and recorded spontaneous in vivo electrophysiological patterns from unanesthetized mice across the first two postnatal weeks. This approach enabled simultaneous recording of large-scale synaptic and cellular activity without damaging fragile cortical circuits. We found that the transition between immature and mature network dynamics was characterized by unexpectedly decreased coordinated cellular and synaptic activity at the beginning of the second postnatal week. After this timepoint, precise neuronal synchronization and oscillatory coupling in space and time robustly emerged. Analysis of continuous electroencephalography (EEG) recordings from human subjects, 36–69 weeks post-gestation, revealed a similar developmental trajectory characterized by a transient decrease in neural activity prior to onset of spatiotemporal oscillatory coupling. Therefore, a shift from local, loosely correlated, prominently sensory-driven patterns to internally organized, spatially distributed, and temporally precise activity is preceded by a transient quiescent period. These findings suggest that mechanisms to developmentally regulate functional network capacity may be evolutionarily conserved and manifest on the systems level as similar, discrete states.

To identify and characterize emergence of advanced neural network properties in the developing brain, we acquired high spatiotemporal resolution electrophysiological recordings from somatosensory cortex of unanesthetized mouse pups aged postnatal day (P) 5 to 14 (n = 108 pups), an epoch that spans the transition from immature to mature activity patterns during sleep. We used minimally invasive surface electrocorticography arrays (NeuroGrids, n = 70 pups) to permit a spatially extensive survey of cortex. These customized NeuroGrids (Khodagholy et al., 2015) consisted of 119 electrodes regularly spaced on a diagonal square-centered lattice embedded in 4-µm-thick parylene C to conform to the cortical surface (Figure 1A). Recordings were made following recovery from surgery to eliminate any influence of anesthesia (Figure 1—figure supplement 1 and Figure 1—figure supplement 2). Mice at these ages enter into a cyclical pattern of sleep and wakefulness as assayed by peripheral indicators, such as muscular tone, movements, and heart rate (Figure 1—figure supplement 1). We ensured electrodes used for analysis were located in somatosensory cortex by postmortem histology. Following NeuroGrid recordings, the location of the array was marked on the surface of the brain using a biocompatible fluorescent material (Rauhala et al., 2020) and tissue was harvested for immunohistochemistry. vGlut2 staining identified the location of primary somatosensory and visual cortices relative to NeuroGrid placement on flattened axial slices (Figure 1B, upper), and the spatially extensive nature of the array compared to the size of the pups’ brains further facilitated anatomical targeting. This histology-based electrode grouping corresponded to spatial localization of electrophysiological activity patterns, identifying electrodes recording from somatosensory cortex (Figure 1B, lower; Figure 1—figure supplement 3). To additionally capture neural spiking patterns across cortical layers, we stereotactically implanted silicon probes (linear electrode arrangement) in somatosensory cortex of a separate cohort (n = 38 pups). Immunohistochemistry of coronal slices from these brains verified probe placement within somatosensory cortex, permitted allocation of recording electrodes to cortical layers (Figure 1C, upper), and demonstrated transcortical neural spiking patterns (Figure 1C, lower; Figure 1—figure supplement 4).

Figure 1 with 5 supplements see all

Download asset Open asset

We expected that physiologically significant features of neural network maturation would be conserved across species. Thus, in parallel, we obtained continuous EEG recordings from normal human subjects 36–69 weeks post-gestation (1 day to 7 months after birth), an epoch that similarly spans the transition from immature to mature activity patterns of human sleep and can be mapped to corresponding timepoints of rodent brain development (Jouvet-Mounier et al., 1969; Clancy et al., 2001; Milh et al., 2007). We first identified epochs of artifact-free behavioral immobility consistent with sleep in the pups (Figure 1—figure supplement 5) and compared the activity present with human sleep EEG. Sleep in the youngest mouse pups (P5–7) was characterized by irregular, discontinuous bursts of neural activity (Figure 1D, E, upper traces). The predominant activity was organized into spindle bursts, oscillations characteristic of developing rodent sensory cortex (Figure 1D, left spectrogram). Sleep in the youngest human subjects recorded was also discontinuous, and contained delta brushes, an oscillatory pattern that develops by 26 weeks post-gestation and disappears around 40–44 weeks (Khazipov and Luhmann, 2006; Figure 1E, left spectrogram). Delta brushes share morphological and frequency characteristics with spindle bursts. Sleep patterns changed profoundly with maturation in both species. More mature mouse pups exhibited nearly continuous patterns during sleep, an absence of spindle bursts, and the appearance of sleep spindles, which are well-characterized oscillations of NREM sleep (Figure 1D, lower trace, right spectrogram). Human sleep was similarly continuous, and sleep spindles appear during NREM between 1 and 2 months after term in infants (Figure 1E, lower trace, right spectrogram). This mature sleep in rodents and humans is predominantly governed by internally generated dynamics (Levenstein et al., 2019). In contrast, self-generated or evoked movements during sleep are closely associated with local field potential (LFP) activity in somatosensory cortex of immature rodents and humans (Khazipov et al., 2004; Dooley et al., 2020). We confirmed that the association between muscular twitches and changes in LFP power (Figure 1F; n = 6716 twitches from n = 52 pups, sampled randomly to ensure uniform group size; Kruskal–Wallis ANOVA, chi-square = 12.93, p=0.0048) or detectable spindle band oscillations (Figure 1G; cross-correlation of twitches and spindle band oscillations exceeded 95% confidence interval, black dashed lines, only in pups aged P5–7) decreased as the pups matured. These marked similarities in the electrophysiological features of neural activity during sleep across development suggest the possibility of evolutionarily conserved mechanisms of neural network maturation.

To investigate how this maturation occurs, we examined the macrostructure of sleep LFP with fine temporal precision over early pup development (P5–14). Mature sleep can be classified into distinct stages (NREM and REM) based on electrophysiological criteria, but these indicators do not reliably differentiate the immature analogues of these states (quiet and active sleep, respectively) until after P11 (Seelke and Blumberg, 2008). Therefore, we focused on periods of behavioral and EMG quiescence that lasted for longer than 10 s (Figure 1—figure supplement 5), with the goal of preferentially analyzing the transition of quiet sleep (which contains spindle bursts) into NREM sleep (which contains sleep spindles). Unexpectedly, recordings across development revealed an epoch of relative neural quiescence at the beginning of the second postnatal week (Figure 2A) that corresponded to decreased spectral power and absence of prominent activity at physiological frequencies (Figure 2B, Figure 2—figure supplement 1). We hypothesized that this quiescence could be related to a transient decrease in the continuity of the neural signal or the power of the activity that was present. Continuity of neural activity was quantified by determining the duration of neural activity above the wideband noise floor. Although continuity increased markedly from the youngest to most mature animals, we found a decrease in the duration of continuous neural activity that corresponded to the identified period of spectral quiescence (Figure 2C). When we extracted the average wideband power of the continuous activity that was present at each developmental day, there was also a transient nadir at this timepoint (Figure 2D). Because we aimed to characterize the trajectory of the electrophysiological features over time without making a priori assumptions about the nature of the temporal relationship or the rate of change, we employed a modeling approach that considered age as a continuous variable. This approach avoided arbitrary grouping of data from different aged pups without sacrificing statistical power and enabled characterization of the developmental profile beyond testing at individual timepoints. Data were fit with linear and polynomial regression models, and model performance was evaluated using leave-one-out cross-validation (LOOCV) to avoid overfitting (Figure 2—figure supplement 2). Polynomial regression provided the best fit for continuity and power data as quantified by the mean squared error (MSE; Figure 2—figure supplement 2B). In addition, linear regression was insufficient to model these parameters because linear fit residuals systematically deviated from zero and consistently overestimated observed values at the beginning of the second postnatal week (Figure 2—figure supplement 2C). The best fit models exhibited local minima, and to estimate the age at which these nadirs occurred, we employed a bootstrapping method. Bootstrapping data within each developmental day localized probability of nadir in continuity and power at the beginning of the second postnatal week with likelihood far exceeding the distribution obtained by bootstrapping across days (Figure 2C, D inset, Figure 2—figure supplement 2D–F, Figure 2—figure supplement 3). We tested the possibility that the quiescence could be related to a shift in proportion of time spent in each sleep state, a decrease in self-generated sleep movements, or differential recovery from anesthesia across ages. In keeping with previous observations, the proportion of active relative to quiet sleep decreased gradually with maturation (Figure 2—figure supplement 4A) and we also found a trend toward fewer self-generated sleep movements in the more mature animals, though this shift was not statistically significant (Figure 2—figure supplement 4B). Recovery from anesthesia was determined individually for each pup, and there was no difference in post-anesthesia stability of oscillation occurrence rate or neural spiking across pups (Figure 1—figure supplement 2). The absence of a marked shift in any of these features during the observed quiescent period makes them unlikely drivers of the phenomenon. Therefore, sleep LFP macrostructure expresses a nonlinear pattern of development that is characterized by a discrete period of relative neural quiescence occurring at the beginning of the second postnatal week.

Figure 2 with 4 supplements see all

Download asset Open asset

Rate and synchrony of neural spiking strongly contribute to expression of LFP patterns. We hypothesized that changes in neural spiking patterns could accompany the transient quiescent period. Spiking activity was detected in recordings obtained using implantable silicon probes from two non-contiguous sites (>200 µm vertical separation) within the deep layers of cortex (IV–VI) and one site within the superficial layers of cortex (I–III) per pup recording (n = 38 pups). We first calculated rates of neural spiking only during epochs of above-threshold LFP activity to eliminate effects related to LFP discontinuity. We found that neural spiking increased nonlinearly over the course of development, again with a significant nadir at the beginning of the second postnatal week as quantified using a bootstrapping method (Figure 3A, Figure 3—figure supplement 1). These data suggest that lowered intracortical synaptic drive could contribute to the nadir in LFP patterns at the beginning of the second postnatal week. Consistent spiking within neuronal integration time is critical for plasticity processes (Mainen and Sejnowski, 1995). To estimate this property, we calculated the inter-spike interval (ISI) between each unique action potential. ISI duration decreased nonlinearly across development, with the longest ISI duration occurring in pups at the beginning of the second postnatal week (Figure 3B, Figure 3—figure supplement 2). Population neural spiking at fine time scales is thus both transiently decreased and desynchronized at this time in both superficial and deep cortical layers, resulting in a brief period of relative quiescence prior to the progressive increase in spiking rate and synchrony that accompany subsequent maturation. During mature NREM sleep, there is alternation between periods of neural spiking and hyperpolarization that serves to synchronize neural populations and is implicated in memory consolidation (Ji and Wilson, 2007; Steriade et al., 1993). We examined population spiking dynamics in the developing cortex by computing the autocorrelation of spiking activity, which permits quantification of neural synchrony over durations relevant to this slow oscillation (Figure 3C). Significant short time scale (<50 ms) interactions (above the 95% confidence interval) robustly emerged only after P9 and were prominent in the most mature pups (Figure 3D, left; n = 38 pups, rank-sum z = −4.2, p=2.5×10⁻⁵). All age groups demonstrated significant suppression of spiking activity within the first 500 ms, becoming more marked as development progressed (Figure 3D, right; n = 38 pups, rank-sum z = 3.0, p=0.002). These data implicate that the transient quiescent period demarcates a transition in population spiking that initiates development of a cycle of longer scale network synchronization characteristic of the dynamics observed during mature sleep (Steriade et al., 1993).

Figure 3 with 2 supplements see all

Download asset Open asset

We then sought to understand the functional relevance of this quiescent transition period to network operations. Key LFP properties (spatial extent, waveform, and power) can provide an index of the underlying neural population activity and its computational purpose – the distance over which neurons can be concurrently recruited, as well as the strength and temporal patterning of this activity (Lowet et al., 2015; Strüber et al., 2017; Cole and Voytek, 2017; Stark et al., 2014). We detected spindle band oscillations, which were present in pups of all ages, to facilitate comparisons. We quantified spatial extent using the two-dimensional organization of the NeuroGrid to determine the proportion of electrodes expressing co-occurring spindle band oscillations with a reference electrode located in primary somatosensory cortex (Figure 4A; Dahal et al., 2019). Waveform was characterized using an asymmetry index (Figure 4B) and power was extracted from the filtered Hilbert envelope (Figure 4C). Spindle band oscillations in the most immature pups were high power and had a substantially asymmetric waveform, but were expressed over a restricted cortical area (~480 µm diameter). In contrast, spindle band oscillations in the most mature pups were lower power and more symmetric, but could variably be detected over a larger spatial extent (~1.4 mm diameter). Given these observed differences in oscillatory parameters, we asked to what extent they could be used to classify the developmental stage at which the oscillations occurred. A coarse tree model based on spatial extent and waveform asymmetry of manually identified oscillations was effectively trained to classify individual spindle band oscillations as originating from immature (P5–7) or mature (P10–14) pups with 92.4% accuracy (Figure 4D), suggesting that these parameters capture defining features of the oscillations at the ends of the developmental spectrum. Spindle band oscillations that occurred during the transient quiescent period at the beginning of the second postnatal week were not consistently classified into either group, instead displaying a unique combination of oscillatory parameters: large spatial extent, low power, and intermediate waveform asymmetry (Figure 4A–C, yellow bars). When examined in a combinatorial fashion, a significant inflection point in oscillation properties emerged between P8 and 9 (Figure 4E). Polynomial regression provided the best fit for these data (Figure 4—figure supplement 1). Bootstrapping data within each developmental day confirmed localization of a nadir at the beginning of the second postnatal week with likelihood far exceeding the distribution obtained by bootstrapping across days (Figure 4F). These data indicate that oscillations are transformed at the beginning of the second postnatal week, potentially reflecting a transition in neural network function from consistently highly localized cortical processing to capacity for broader, tuned recruitment of neural activity.

Figure 4 with 1 supplement see all

Download asset Open asset

Oscillations in the adult brain precisely organize neural activity across spatial and temporal scales. A key marker for this organization is cross-frequency coupling, which is important for regulating strength of inputs from different afferent pathways (Schomburg et al., 2014) and directing flow of information processing (Buschman and Miller, 2007). The marked shift in oscillatory properties that we observed at the beginning of the second postnatal week suggested that this organizational capacity may also be developmentally regulated. We first investigated the relationship of spindle band activity with other frequencies by detecting epochs of sustained increase in spindle band power and calculating measures of amplitude cross-frequency coupling. Spindle band epochs in the most immature animals (P5–7) demonstrated high comodulation across a broad, non-specific range of frequencies, in contrast to frequency-specific comodulation expressed in mature animals (P13–14; Figure 5A). Specific, discrete frequency bands that demonstrated significant temporal coupling with spindle band oscillations emerged only after P8–9. Discrete coupling with the gamma band (45–60 Hz) first occurred at P10–12, with an additional high gamma band (60–90 Hz) arising at P13–14 (Figure 5B; n = 1020 spindle band oscillations from 51 pups; area under the curve [AUC] for 45–60 Hz chi-square = 10.13, p=0.02; AUC for 60–90 Hz chi-square = 24.29, p<0.001). Taken together, these results suggest that immature networks lack the capacity to couple discrete frequency bands, whereas after the beginning of the second postnatal week such capacity emerges, potentially signaling a developmental shift in the function of oscillations.

Figure 5 with 1 supplement see all

Download asset Open asset

Furthermore, oscillations facilitate neural communication by biasing spike occurrence to particular phases of the oscillatory waveform. Spindle band oscillations recruited neural spiking above baseline rates to a variable extent across development, with the lowest amount of recruitment occurring at the beginning of the second postnatal week (Figure 5C, Figure 5—figure supplement 1). This trajectory significantly deviated from linearity with a discrete nadir confirmed by bootstrapping during this epoch (Figure 5—figure supplement 1A). P8 and P9 also exhibited the lowest proportion of recording sessions with significant phase-locking (kappa > 0.1 and alpha < 0.05), and consistency of phase-locking abruptly increased at P10 (Figure 5D, E; n = 38 pups implanted with silicon probes). These results reinforce the notion of enhanced information processing capacity in cortical networks after a rapid transitional period at the beginning of the second postnatal week.

Thus, multiple LFP and neural spiking measures followed a similar highly nonlinear developmental trajectory marked by a key transition at the beginning of the second postnatal week in mouse pups. We next aimed to investigate whether this pattern of neural maturation was conserved across species. To accomplish this goal, we analyzed clinically acquired continuous EEG data from 54 human subjects ranging from 36 to 69 weeks post-gestation. EEG recordings were included for analysis only if they were reported as normal by the reading epileptologists and the subject had no known underlying neurological or genetic condition. The majority of subjects received a discharge diagnosis of normal movements/behaviors (e.g., sleep myoclonus) or brief resolved unexplained events (BRUE; Figure 6—figure supplement 1). We analyzed epochs of clinically determined quiet/NREM sleep from each subject using data derived from electrodes overlying human somatosensory cortex (parietal or central electrodes; no differences were found between these localizations; Figure 6—figure supplement 1D). Raw traces and normalized spectral analysis suggested an epoch of transient neural quiescence between 42 and 47 weeks (Figure 6A, B), akin to what we observed at the beginning of the second postnatal week in mice. To quantify these observations, we again used regression fitting to determine whether a nonlinear trajectory was present and bootstrapping of data within and between ages to identify any significant nadirs in neural activity patterns. Wideband power of oscillatory activity followed a highly nonlinear trajectory, with a significant nadir between 42 and 47 weeks (Figure 6C, Figure 6—figure supplement 2A). Power spectra revealed a lack of prominent frequency peaks during this epoch, as well as an overall steeper decay in power with increasing frequency (corresponding to a larger exponent for the modeled power law; Figure 6—figure supplement 3A–C). Quantification of this aperiodic component of the power spectrum across ages similarly revealed a transient increase around the time of nadir in wideband power (Figure 6—figure supplement 3D, E). As expected from previous work (Torres and Anderson, 1985), duration of oscillatory epochs generally increased over development, but we did observe a period of relative stationarity coinciding with the decrease in wideband power (Figure 6D, Figure 6—figure supplement 2B). To assess properties of oscillatory activity in these subjects, we detected spindle band oscillations. Such oscillations were present in all ages, consistent with delta brushes in the youngest subjects and sleep spindles in the older subjects (Figure 6E). We calculated spatial extent of these oscillations as the co-occurrence of spindle band activity across EEG electrodes and found that this activity shifted from highly localized to broader cortical expression around an inflection point between 42 and 47 weeks (Figure 6F, Figure 6—figure supplement 2C). We also found evidence for significant cross-frequency coupling of higher frequencies with the spindle band emerging after 47 weeks post-gestation (Figure 6G, H; n = 702 spindle band oscillations from 13 subjects; Kruskal–Wallis chi-square for AUC = 138.4; p<0.001). Taken together, these data support the existence of a conserved transient quiescent period in development that bridges a dramatic shift in functional properties of neural networks.

Figure 6 with 3 supplements see all

Download asset Open asset

We demonstrate that maturation of cortical network dynamics is preceded by a discrete, relatively quiescent, evolutionarily conserved transition period during early postnatal development. During this period, oscillations and neural spiking display a nadir that signals the shift from local, loosely correlated, putatively sensory-driven patterns to internally organized, spatially distributed, and temporally precise activity. Such a developmental inflection point may reflect a rapid and confluent engagement of mechanisms that enable mature computational processes within cortical networks.

Self-organization in complex systems is often mediated by processes that can facilitate sudden transitions in system behavior. Our results provide evidence for such a transition in developing cortex of rodents and humans. Because rodent neurodevelopment is temporally compressed relative to humans, an inflection epoch that lasts 1–2 days in mouse pups could conceivably extend over weeks in humans (Clancy et al., 2001). Prior to this inflection point, oscillatory activity was intermittent with non-uniform cycle-to-cycle waveforms and high broadband frequency coherence. Oscillatory activities after the transition became nearly uninterrupted, exhibiting coupling at discrete frequency bands and increased waveform symmetry. Based on these similarities in oscillatory patterns, we infer that the neural spiking features we observe through invasive recordings in mouse pups could also extend to normally developing human infants, where such recordings are not possible. Neural spiking was more efficiently recruited to oscillatory activity with consistent, strong phase-locking, and higher temporal precision. These features are characteristic of adult neural networks engaged in information processing (Masquelier et al., 2009). Thus, a transient period of developmental perturbation may be an evolutionarily conserved indicator of a substantial shift in cortical network operations.

Our results suggest that the cortical network dynamics shift to permit temporally precise neural spiking after the transition period. Significantly synchronized firing within 10–50 ms intervals was absent prior to this period and robustly emerged afterward. This result is consistent with observed increase in pair-wise correlations of spike activity in visual cortex over development (Colonnese et al., 2017). Although recruitment of neural spiking during epochs of spindle burst oscillations in immature rodents likely co-activates neurons that share similar afferent inputs (Khazipov et al., 2004; Colonnese et al., 2010; Khazipov et al., 2013; Winnubst et al., 2015), more precise coordination is required for establishment of flexible, defined neural assemblies (Dragoi and Buzsáki, 2006), and the capacity of circuits to express sequential activation of these assemblies is developmentally regulated (Farooq and Dragoi, 2019).

We found that the neural population activity demonstrated correlated suppression over a progressively shorter duration (from approximately 500 to 200 ms) with maturation. This process is in keeping with the progressively decorrelated activity patterns observed when analysis is performed on the time scale of calcium indicators (Golshani et al., 2009; Rochefort et al., 2009). When oscillations occur in discrete frequency bands, the possibility for complex comodulation arises. This cross-frequency coupling has been linked to information processing in the mature brain, though the precise anatomical substrates and physiological mechanisms that enable its expression are often unknown (Tort et al., 2009; Fujisawa and Buzsáki, 2011; Bosman et al., 2012). We demonstrated that prior to the transient quiescent period the dominant cortical oscillations (centered on spindle band) exhibit high comodulation extending from 1 to 100 Hz. During this period, broadband comodulation abruptly decreased and was subsequently replaced by discrete gamma oscillations. Therefore, the oscillatory infrastructure to support cross-frequency coupling arises concomitantly with fine time-scale neural spiking synchronization and robust phase-locking of these spikes to LFP, suggesting an emergence of adult-like information processing capacity at this developmental timepoint.

Despite sharing a peak frequency, immature and mature spindle band oscillations are highly differentiable on the joint basis of waveform, power, and spatial extent. Correspondingly, the cellular mechanisms that permit expression of these oscillations, as well as their functional purpose, are different. Comparison of oscillatory features can shed light on the functional capacity of the network. We found that although of high amplitude, spindle band frequency oscillations in the most immature mouse pups were highly and uniformly confined to localized regions of cortex (200–500 µm) roughly corresponding to cortical columns (Yang et al., 2013; Feldmeyer et al., 2013). Similarly, human delta brushes exhibited restricted localization even given the limited spatial resolution of clinical EEG. During the transition period, spindle band oscillations in mouse pups had higher spatial extent, but markedly reduced amplitude, perhaps facilitated by increased intracortical synaptogenesis and maturation of corticothalamic afferents (Naskar et al., 2019). The transition period in humans was marked by transiently decreased spatial extent of spindle band oscillations, which we hypothesize is related to attenuation of low-power activity by the intervening tissue between cortical surface and scalp electrode. We observed that mature spindle band activity across species was characterized by increased and customizable spatial extent. Thus, the end of the transient quiescent period marks capacity for oscillatory frequencies that were previously confined to local processing units to extend into broader cortical areas with the potential to establish intercortical communication.

Our study was limited to somatosensory cortex, and the existence of a transient quiescent epoch outside of this brain region is unknown. Similar to our results in the spindle band, early gamma oscillations disappear around the beginning to the second postnatal week, to be later replaced with gamma rhythms characteristic of adult animals (Minlebaev et al., 2011). In visual cortex, a transient nadir in continuity was not identified, but continuity did rise sharply after P8 with a nonlinear trajectory (Shen and Colonnese, 2016). Furthermore, the least synchronous neural spiking activity in visual cortex occurred at the beginning of the second postnatal week, potentially in keeping with our decrease in temporal precision of spiking at this time (Colonnese et al., 2017). Higher order cortices, which do not directly receive sensory thalamic inputs, may express a different developmental trajectory, in keeping with their distinctive role in information processing, though oscillatory expression during this same epoch has been identified as critical for intact adult function (Bitzenhofer et al., 2015; Bitzenhofer et al., 2021; Bitzenhofer et al., 2020). Therefore, multiple lines of evidence support an abrupt change in cortical network function at the beginning of the second postnatal week. It is possible that our use of surface arrays, which sample summated LFP activity from the undisturbed superficial cortical layers (I–III), highlights the transient quiescent state compared to penetrating probes that disrupt the cortical surface upon implantation. Furthermore, our conducting polymer-based electrodes have lower impedance than the silicon probes used in most neonatal rodent studies (Khodagholy et al., 2015), potentially increasing the sensitivity to changes in oscillatory power and continuity.

Given the multitude of anatomic and functional changes that occur during the early postnatal epoch, it is likely that the shift in network operating mode we observe is mechanistically multifactorial. A potential driver for this state is maturation of functional cortical inhibition. Functional inhibitory connections regulate physiological oscillations in mature neural networks enabling submillisecond precision of spike timing and facilitating inter-regional synchronization (Buzsáki and Chrobak, 1995; Cobb et al., 1995; Pouille and Scanziani, 2001). In contrast, interneurons do not substantially pace immature oscillations during the first postnatal week, though they can be recruited to and contribute to spatial properties of this activity (Khazipov et al., 2004; Minlebaev et al., 2011). Functional feedforward inhibition rapidly matures around the beginning of the second postnatal week in mice, resulting in emergence of adult-like hyperpolarizing synaptic potentials and precisely timed action potentials in glutamatergic neurons within the microcircuit (Daw et al., 2007). This timepoint, derived from in vitro data, temporally overlaps with the onset of adult-like neural network activity we observe in vivo. In the human brain, when mature functional inhibition arises is not known, but ongoing migration of cells that differentiate into interneurons in the months after birth (Paredes et al., 2016) suggests the possibility of a postnatal transition to such network activity. The aperiodic component of the power spectrum has been linked to excitatory-inhibitory balance (Gao et al., 2017), and our finding that this value transiently increases during the quiescent period could support a temporary increase in inhibition in the human brain. Such a mechanism could underlie the known postnatal disappearance of the EEG bursts characteristics of trace alternant (Ellingson and Peters, 1980), resulting in the observed low-voltage state with lack of organized oscillatory activity prior to emergence of mature NREM sleep patterns.

Whether a period of quiescence is required for physiological transition to mature cortical processing remains unclear. Manipulations that could modulate its expression, such as alteration of the neuromodulatory milieu, activation/suppression of specific cell-type activity, or modulation of the sensory environment (Smith et al., 2018), could address this question. Forging linkages to activity patterns observed using calcium imaging (Wilson et al., 2018) and enabling high spatiotemporal resolution chronic in vivo electrophysiology will also be critical. In humans, neuronal dynamics suggestive of mature information processing mechanisms are established by 6 months of age (Jannesari et al., 2020), and changes in sleep patterns have been identified as markers for brain connectivity from toddlers to adolescents (Kurth et al., 2017). Extending such measures to earlier timepoints could help to clarify functional ramifications of the quiescent period in humans. Furthermore, the manifestation of a developmental epoch of relative quiescence in electrophysiological activity suggests the possibility for use as a network-level biomarker that could be altered in disorders of neurodevelopment, from mouse models to human subjects.

Neural processes in the adult brain are characterized by a specific constellation of network dynamics that are necessary for execution of complex behaviors and cognition (Tolner et al., 2012). Our results suggest that the capacity for these processes simultaneously emerges abruptly after a distinct transition period characterized by relative neural quiescence. This transient quiescent period may be a critical stage that indicates a shift in the neural network operating mode from a predominantly external stimulus-driven state to one that facilitates the internal representations associated with planning and executive function.

Source data are presented in Figure supplements and uploaded to Dryad. Data pertaining to human subjects is governed by IRB policy and can be accessed through application to the IRB. Pooled, processed human subject data are uploaded to Dryad.

1. 1. Khodagholy D

   (2021) Dryad Digital Repository

   A transient postnatal quiescent period precedes emergence of mature cortical dynamics.

   https://doi.org/10.5061/dryad.15dv41nxp

1. 1. An S
   2. Kilb W
   3. Luhmann HJ

   (2014) Sensory-evoked and spontaneous gamma and spindle bursts in neonatal rat motor cortex

   Journal of Neuroscience 34:10870–10883.

   https://doi.org/10.1523/JNEUROSCI.4539-13.2014

   • PubMed
   • Google Scholar
2. 1. Anderson CM
   2. Torres F
   3. Faoro A

   (1985) The EEG of the early premature

   Electroencephalography and Clinical Neurophysiology 60:95–105.

   https://doi.org/10.1016/0013-4694(85)90015-X

   • PubMed
   • Google Scholar
3. 1. Bitzenhofer SH
   2. Sieben K
   3. Siebert KD
   4. Spehr M
   5. Hanganu-Opatz IL

   (2015) Oscillatory activity in developing prefrontal networks results from theta-gamma-modulated synaptic inputs

   Cell Reports 11:486–497.

   https://doi.org/10.1016/j.celrep.2015.03.031

   • PubMed
   • Google Scholar
4. 1. Bitzenhofer SH
   2. Pöpplau JA
   3. Hanganu-Opatz I

   (2020) Gamma activity accelerates during prefrontal development

   eLife 9:e56795.

   https://doi.org/10.7554/eLife.56795

   • Google Scholar
5. 1. Bitzenhofer SH
   2. Pöpplau JA
   3. Chini M
   4. Marquardt A
   5. Hanganu-Opatz IL

   (2021) A transient developmental increase in prefrontal activity alters network maturation and causes cognitive dysfunction in adult mice

   Neuron 109:1350–1364.

   https://doi.org/10.1016/j.neuron.2021.02.011

   • PubMed
   • Google Scholar
6. 1. Blumberg MS
   2. Sokoloff G
   3. Tiriac A
   4. Del Rio-Bermudez C

   (2015) A valuable and promising method for recording brain activity in behaving newborn rodents

   Developmental Psychobiology 57:506–517.

   https://doi.org/10.1002/dev.21305

   • PubMed
   • Google Scholar
7. 1. Bosman CA
   2. Schoffelen J-M
   3. Brunet N
   4. Oostenveld R
   5. Bastos AM
   6. Womelsdorf T
   7. Rubehn B
   8. Stieglitz T
   9. De Weerd P
   10. Fries P

   (2012) Attentional stimulus selection through selective synchronization between monkey visual Areas

   Neuron 75:875–888.

   https://doi.org/10.1016/j.neuron.2012.06.037

   • Google Scholar
8. 1. Bureau I
   2. Shepherd GM
   3. Svoboda K

   (2004) Precise development of functional and anatomical columns in the neocortex

   Neuron 42:789–801.

   https://doi.org/10.1016/j.neuron.2004.05.002

   • PubMed
   • Google Scholar
9. 1. Buschman TJ
   2. Miller EK

   (2007) Top-down versus bottom-up control of attention in the prefrontal and posterior parietal cortices

   Science 315:1860–1862.

   https://doi.org/10.1126/science.1138071

   • PubMed
   • Google Scholar
10. 1. Buzsáki G
    2. Chrobak JJ

    (1995) Temporal structure in spatially organized neuronal ensembles: a role for interneuronal networks

    Current Opinion in Neurobiology 5:504–510.

    https://doi.org/10.1016/0959-4388(95)80012-3

    • PubMed
    • Google Scholar
11. 1. Clancy B
    2. Darlington RB
    3. Finlay BL

    (2001) Translating developmental time across mammalian species

    Neuroscience 105:7–17.

    https://doi.org/10.1016/S0306-4522(01)00171-3

    • PubMed
    • Google Scholar
12. 1. Cobb SR
    2. Buhl EH
    3. Halasy K
    4. Paulsen O
    5. Somogyi P

    (1995) Synchronization of neuronal activity in Hippocampus by individual GABAergic interneurons

    Nature 378:75–78.

    https://doi.org/10.1038/378075a0

    • PubMed
    • Google Scholar
13. 1. Cole SR
    2. Voytek B

    (2017) Brain oscillations and the importance of waveform shape

    Trends in Cognitive Sciences 21:137–149.

    https://doi.org/10.1016/j.tics.2016.12.008

    • PubMed
    • Google Scholar
14. 1. Colonnese MT
    2. Kaminska A
    3. Minlebaev M
    4. Milh M
    5. Bloem B
    6. Lescure S
    7. Moriette G
    8. Chiron C
    9. Ben-Ari Y
    10. Khazipov R

    (2010) A conserved switch in sensory processing prepares developing neocortex for vision

    Neuron 67:480–498.

    https://doi.org/10.1016/j.neuron.2010.07.015

    • PubMed
    • Google Scholar
15. 1. Colonnese MT
    2. Shen J
    3. Murata Y

    (2017) Uncorrelated neural firing in mouse visual cortex during spontaneous retinal waves

    Frontiers in Cellular Neuroscience 11:1–14.

    https://doi.org/10.3389/fncel.2017.00289

    • PubMed
    • Google Scholar
16. 1. Corner MA
    2. Kwee P

    (1976) Cyclic EEG and motility patterns during sleep in restrained infant rats

    Electroencephalography and Clinical Neurophysiology 41:64–72.

    https://doi.org/10.1016/0013-4694(76)90215-7

    • PubMed
    • Google Scholar
17. 1. Dahal P
    2. Ghani N
    3. Flinker A
    4. Dugan P
    5. Friedman D
    6. Doyle W
    7. Devinsky O
    8. Khodagholy D
    9. Gelinas JN

    (2019) Interictal epileptiform discharges shape large-scale intercortical communication

    Brain 142:3502–3513.

    https://doi.org/10.1093/brain/awz269

    • PubMed
    • Google Scholar
18. 1. Daw MI
    2. Ashby MC
    3. Isaac JT

    (2007) Coordinated developmental recruitment of latent fast spiking interneurons in layer IV barrel cortex

    Nature Neuroscience 10:453–461.

    https://doi.org/10.1038/nn1866

    • PubMed
    • Google Scholar
19. 1. Dooley JC
    2. Glanz RM
    3. Sokoloff G
    4. Blumberg MS

    (2020) Self-Generated whisker movements drive State-Dependent sensory input to developing barrel cortex

    Current Biology 30:2404–2410.

    https://doi.org/10.1016/j.cub.2020.04.045

    • PubMed
    • Google Scholar
20. 1. Dragoi G
    2. Buzsáki G

    (2006) Temporal encoding of place sequences by hippocampal cell assemblies

    Neuron 50:145–157.

    https://doi.org/10.1016/j.neuron.2006.02.023

    • PubMed
    • Google Scholar
21. 1. Ellingson RJ
    2. Peters JF

    (1980) Development of EEG and daytime sleep patterns in Trisomy-21 infants during the first year of life: longitudinal observations

    Electroencephalography and Clinical Neurophysiology 50:457–466.

    https://doi.org/10.1016/0013-4694(80)90012-7

    • PubMed
    • Google Scholar
22. 1. Farooq U
    2. Dragoi G

    (2019) Emergence of preconfigured and plastic time-compressed sequences in early postnatal development

    Science 363:168–173.

    https://doi.org/10.1126/science.aav0502

    • PubMed
    • Google Scholar
23. 1. Favuzzi E
    2. Deogracias R
    3. Marques-Smith A
    4. Maeso P
    5. Jezequel J
    6. Exposito-Alonso D
    7. Balia M
    8. Kroon T
    9. Hinojosa AJ
    10. F Maraver E
    11. Rico B

    (2019) Distinct molecular programs regulate synapse specificity in cortical inhibitory circuits

    Science 363:413–417.

    https://doi.org/10.1126/science.aau8977

    • PubMed
    • Google Scholar
24. 1. Feldmeyer D
    2. Brecht M
    3. Helmchen F
    4. Petersen CC
    5. Poulet JF
    6. Staiger JF
    7. Luhmann HJ
    8. Schwarz C

    (2013) Barrel cortex function

    Progress in Neurobiology 103:3–27.

    https://doi.org/10.1016/j.pneurobio.2012.11.002

    • PubMed
    • Google Scholar
25. 1. Foley PL
    2. Kendall LV
    3. Turner PV

    (2019) Clinical management of pain in rodents

    Comparative Medicine 69:468–489.

    https://doi.org/10.30802/AALAS-CM-19-000048

    • PubMed
    • Google Scholar
26. 1. Foster DJ
    2. Wilson MA

    (2006) Reverse replay of behavioural sequences in hippocampal place cells during the awake state

    Nature 440:680–683.

    https://doi.org/10.1038/nature04587

    • Google Scholar
27. 1. Fries P
    2. Roelfsema PR
    3. Engel AK
    4. König P
    5. Singer W

    (1997) Synchronization of oscillatory responses in visual cortex correlates with perception in interocular rivalry

    PNAS 94:12699–12704.

    https://doi.org/10.1073/pnas.94.23.12699

    • PubMed
    • Google Scholar
28. 1. Fujisawa S
    2. Buzsáki G

    (2011) A 4 hz oscillation adaptively synchronizes prefrontal, VTA, and hippocampal activities

    Neuron 72:153–165.

    https://doi.org/10.1016/j.neuron.2011.08.018

    • PubMed
    • Google Scholar
29. 1. Gao R
    2. Peterson EJ
    3. Voytek B

    (2017) Inferring synaptic excitation/inhibition balance from field potentials

    NeuroImage 158:70–78.

    https://doi.org/10.1016/j.neuroimage.2017.06.078

    • PubMed
    • Google Scholar
30. 1. Geisler C
    2. Diba K
    3. Pastalkova E
    4. Mizuseki K
    5. Royer S
    6. Buzsáki G

    (2010) Temporal delays among place cells determine the frequency of population theta oscillations in the Hippocampus

    PNAS 107:7957–7962.

    https://doi.org/10.1073/pnas.0912478107

    • Google Scholar
31. 1. Golbs A
    2. Nimmervoll B
    3. Sun JJ
    4. Sava IE
    5. Luhmann HJ

    (2011) Control of programmed cell death by distinct electrical activity patterns

    Cerebral Cortex 21:1192–1202.

    https://doi.org/10.1093/cercor/bhq200

    • PubMed
    • Google Scholar
32. 1. Goldberg EM
    2. Jeong HY
    3. Kruglikov I
    4. Tremblay R
    5. Lazarenko RM
    6. Rudy B

    (2011) Rapid developmental maturation of neocortical FS cell intrinsic excitability

    Cerebral Cortex 21:666–682.

    https://doi.org/10.1093/cercor/bhq138

    • PubMed
    • Google Scholar
33. 1. Golshani P
    2. Goncalves JT
    3. Khoshkhoo S
    4. Mostany R
    5. Smirnakis S
    6. Portera-Cailliau C

    (2009) Internally mediated developmental desynchronization of neocortical network activity

    Journal of Neuroscience 29:10890–10899.

    https://doi.org/10.1523/JNEUROSCI.2012-09.2009

    • Google Scholar
34. 1. Hanganu IL
    2. Ben-Ari Y
    3. Khazipov R

    (2006) Retinal waves trigger spindle bursts in the neonatal rat visual cortex

    Journal of Neuroscience 26:6728–6736.

    https://doi.org/10.1523/JNEUROSCI.0752-06.2006

    • PubMed
    • Google Scholar
35. 1. Holmes GL
    2. Lombroso CT

    (1993) Prognostic value of background patterns in the neonatal EEG

    Journal of Clinical Neurophysiology 10:323–352.

    https://doi.org/10.1097/00004691-199307000-00008

    • Google Scholar
36. 1. Jannesari M
    2. Saeedi A
    3. Zare M
    4. Ortiz-Mantilla S
    5. Plenz D
    6. Benasich AA

    (2020) Stability of neuronal avalanches and long-range temporal correlations during the first year of life in human infants

    Brain Structure and Function 225:1169–1183.

    https://doi.org/10.1007/s00429-019-02014-4

    • PubMed
    • Google Scholar
37. 1. Ji D
    2. Wilson MA

    (2007) Coordinated memory replay in the visual cortex and Hippocampus during sleep

    Nature Neuroscience 10:100–107.

    https://doi.org/10.1038/nn1825

    • PubMed
    • Google Scholar
38. 1. Jouvet-Mounier D
    2. Astic L
    3. Lacote D

    (1969) Ontogenesis of the states of sleep in rat, cat, and guinea pig during the first postnatal month

    Developmental Psychobiology 2:216–239.

    https://doi.org/10.1002/dev.420020407

    • Google Scholar
39. 1. Khazipov R
    2. Sirota A
    3. Leinekugel X
    4. Holmes GL
    5. Ben-Ari Y
    6. Buzsáki G

    (2004) Early motor activity drives spindle bursts in the developing somatosensory cortex

    Nature 432:758–761.

    https://doi.org/10.1038/nature03132

    • PubMed
    • Google Scholar
40. 1. Khazipov R
    2. Minlebaev M
    3. Valeeva G

    (2013) Early gamma oscillations

    Neuroscience 250:240–252.

    https://doi.org/10.1016/j.neuroscience.2013.07.019

    • PubMed
    • Google Scholar
41. 1. Khazipov R
    2. Luhmann HJ

    (2006) Early patterns of electrical activity in the developing cerebral cortex of humans and rodents

    Trends in Neurosciences 29:414–418.

    https://doi.org/10.1016/j.tins.2006.05.007

    • PubMed
    • Google Scholar
42. 1. Khodagholy D
    2. Gelinas JN
    3. Thesen T
    4. Doyle W
    5. Devinsky O
    6. Malliaras GG
    7. Buzsáki G

    (2015) NeuroGrid: recording action potentials from the surface of the brain

    Nature Neuroscience 18:310–315.

    https://doi.org/10.1038/nn.3905

    • PubMed
    • Google Scholar
43. 1. Khodagholy D
    2. Gelinas JN
    3. Buzsáki G

    (2017) Learning-enhanced coupling between ripple oscillations in association cortices and Hippocampus

    Science 358:369–372.

    https://doi.org/10.1126/science.aan6203

    • Google Scholar
44. 1. Kurth S
    2. Riedner BA
    3. Dean DC
    4. O’Muircheartaigh J
    5. Huber R
    6. Jenni OG
    7. Deoni SCL
    8. LeBourgeois MK

    (2017) Traveling slow oscillations during sleep: a marker of brain connectivity in childhood

    Sleep 40:10–12.

    https://doi.org/10.1093/sleep/zsx121

    • Google Scholar
45. 1. Latchoumane CV
    2. Ngo HV
    3. Born J
    4. Shin HS

    (2017) Thalamic spindles promote memory formation during sleep through triple Phase-Locking of cortical, thalamic, and hippocampal rhythms

    Neuron 95:424–435.

    https://doi.org/10.1016/j.neuron.2017.06.025

    • PubMed
    • Google Scholar
46. 1. Levenstein D
    2. Buzsáki G
    3. Rinzel J

    (2019) NREM sleep in the rodent neocortex and Hippocampus reflects excitable dynamics

    Nature Communications 10:1–12.

    https://doi.org/10.1038/s41467-019-10327-5

    • PubMed
    • Google Scholar
47. 1. Lowet E
    2. Roberts M
    3. Hadjipapas A
    4. Peter A
    5. van der Eerden J
    6. De Weerd P

    (2015) Input-Dependent frequency modulation of cortical gamma oscillations shapes spatial synchronization and enables phase coding

    PLOS Computational Biology 11:e1004072.

    https://doi.org/10.1371/journal.pcbi.1004072

    • Google Scholar
48. 1. Mainen ZF
    2. Sejnowski TJ

    (1995) Reliability of spike timing in neocortical neurons

    Science 268:1503–1506.

    https://doi.org/10.1126/science.7770778

    • PubMed
    • Google Scholar
49. 1. Maingret N
    2. Girardeau G
    3. Todorova R
    4. Goutierre M
    5. Zugaro M

    (2016) Hippocampo-cortical coupling mediates memory consolidation during sleep

    Nature Neuroscience 19:959–964.

    https://doi.org/10.1038/nn.4304

    • PubMed
    • Google Scholar
50. 1. Masquelier T
    2. Hugues E
    3. Deco G
    4. Thorpe SJ

    (2009) Oscillations, phase-of-firing coding, and spike timing-dependent plasticity: an efficient learning scheme

    Journal of Neuroscience 29:13484–13493.

    https://doi.org/10.1523/JNEUROSCI.2207-09.2009

    • PubMed
    • Google Scholar
51. 1. Milh M
    2. Kaminska A
    3. Huon C
    4. Lapillonne A
    5. Ben-Ari Y
    6. Khazipov R

    (2007) Rapid cortical oscillations and early motor activity in premature human neonate

    Cerebral Cortex 17:1582–1594.

    https://doi.org/10.1093/cercor/bhl069

    • PubMed
    • Google Scholar
52. 1. Minlebaev M
    2. Colonnese M
    3. Tsintsadze T
    4. Sirota A
    5. Khazipov R

    (2011) Early γ oscillations synchronize developing thalamus and cortex

    Science 334:226–229.

    https://doi.org/10.1126/science.1210574

    • PubMed
    • Google Scholar
53. 1. Modol L
    2. Bollmann Y
    3. Tressard T
    4. Baude A
    5. Che A
    6. Duan ZRS
    7. Babij R
    8. De Marco García NV
    9. Cossart R

    (2020) Assemblies of perisomatic GABAergic neurons in the developing barrel cortex

    Neuron 105:93–105.

    https://doi.org/10.1016/j.neuron.2019.10.007

    • PubMed
    • Google Scholar
54. 1. Naskar S
    2. Narducci R
    3. Balzani E
    4. Cwetsch AW
    5. Tucci V
    6. Cancedda L

    (2019) The development of synaptic transmission is time-locked to early social behaviors in rats

    Nature Communications 10:1–12.

    https://doi.org/10.1038/s41467-019-09156-3

    • Google Scholar
55. 1. Paredes MF
    2. James D
    3. Gil-Perotin S
    4. Kim H
    5. Cotter JA
    6. Ng C
    7. Sandoval K
    8. Rowitch DH
    9. Xu D
    10. McQuillen PS
    11. Garcia-Verdugo JM
    12. Huang EJ
    13. Alvarez-Buylla A

    (2016) Extensive migration of young neurons into the infant human frontal lobe

    Science 354:aaf7073.

    https://doi.org/10.1126/science.aaf7073

    • PubMed
    • Google Scholar
56. 1. Peyrache A
    2. Khamassi M
    3. Benchenane K
    4. Wiener SI
    5. Battaglia FP

    (2009) Replay of rule-learning related neural patterns in the prefrontal cortex during sleep

    Nature Neuroscience 12:919–926.

    https://doi.org/10.1038/nn.2337

    • PubMed
    • Google Scholar
57. 1. Pouille F
    2. Scanziani M

    (2001) Enforcement of temporal fidelity in pyramidal cells by somatic feed-forward inhibition

    Science 293:1159–1163.

    https://doi.org/10.1126/science.1060342

    • PubMed
    • Google Scholar
58. 1. Rauhala OJ
    2. Dominguez S
    3. Spyropoulos GD
    4. Ferrero JJ
    5. Boyers TR
    6. Jastrzebska‐Perfect P
    7. Cea C
    8. Khodagholy D
    9. Gelinas JN

    (2020) Chitosan‐based, biocompatible, solution processable films for in vivo localization of neural interface devices

    Advanced Materials Technologies 5:1900663.

    https://doi.org/10.1002/admt.201900663

    • Google Scholar
59. 1. Rochefort NL
    2. Garaschuk O
    3. Milos RI
    4. Narushima M
    5. Marandi N
    6. Pichler B
    7. Kovalchuk Y
    8. Konnerth A

    (2009) Sparsification of neuronal activity in the visual cortex at eye-opening

    PNAS 106:15049–15054.

    https://doi.org/10.1073/pnas.0907660106

    • PubMed
    • Google Scholar
60. 1. Schomburg EW
    2. Fernández-Ruiz A
    3. Mizuseki K
    4. Berényi A
    5. Anastassiou CA
    6. Koch C
    7. Buzsáki G

    (2014) Theta phase segregation of input-specific gamma patterns in entorhinal-hippocampal networks

    Neuron 84:470–485.

    https://doi.org/10.1016/j.neuron.2014.08.051

    • PubMed
    • Google Scholar
61. 1. Seelke AM
    2. Blumberg MS

    (2008) The microstructure of active and quiet sleep as cortical Delta activity emerges in infant rats

    Sleep 31:691–699.

    https://doi.org/10.1093/sleep/31.5.691

    • PubMed
    • Google Scholar
62. 1. Shen J
    2. Colonnese MT

    (2016) Development of activity in the mouse visual cortex

    Journal of Neuroscience 36:12259–12275.

    https://doi.org/10.1523/JNEUROSCI.1903-16.2016

    • PubMed
    • Google Scholar
63. 1. Smith GB
    2. Hein B
    3. Whitney DE
    4. Fitzpatrick D
    5. Kaschube M

    (2018) Distributed network interactions and their emergence in developing neocortex

    Nature Neuroscience 21:1600–1608.

    https://doi.org/10.1038/s41593-018-0247-5

    • PubMed
    • Google Scholar
64. 1. Stark E
    2. Roux L
    3. Eichler R
    4. Senzai Y
    5. Royer S
    6. Buzsáki G

    (2014) Pyramidal cell-interneuron interactions underlie hippocampal ripple oscillations

    Neuron 83:467–480.

    https://doi.org/10.1016/j.neuron.2014.06.023

    • PubMed
    • Google Scholar
65. 1. Steriade M
    2. Nuñez A
    3. Amzica F

    (1993)

    A novel slow (< 1 hz) oscillation of neocortical neurons in vivo: Depolarizing and hyperpolarizing components

    The Journal of Neuroscience 13:3252–3265.

    • PubMed
    • Google Scholar
66. 1. Strüber M
    2. Sauer JF
    3. Jonas P
    4. Bartos M

    (2017) Distance-dependent inhibition facilitates focality of gamma oscillations in the dentate gyrus

    Nature Communications 8:758.

    https://doi.org/10.1038/s41467-017-00936-3

    • PubMed
    • Google Scholar
67. 1. Tolner EA
    2. Sheikh A
    3. Yukin AY
    4. Kaila K
    5. Kanold PO

    (2012) Subplate neurons promote spindle bursts and thalamocortical patterning in the neonatal rat somatosensory cortex

    Journal of Neuroscience 32:692–702.

    https://doi.org/10.1523/JNEUROSCI.1538-11.2012

    • PubMed
    • Google Scholar
68. 1. Torres F
    2. Anderson C

    (1985) The normal EEG of the human newborn

    Journal of Clinical Neurophysiology 2:89–104.

    https://doi.org/10.1097/00004691-198504000-00001

    • Google Scholar
69. 1. Tort AB
    2. Komorowski RW
    3. Manns JR
    4. Kopell NJ
    5. Eichenbaum H

    (2009) Theta-gamma coupling increases during the learning of item-context associations

    PNAS 106:20942–20947.

    https://doi.org/10.1073/pnas.0911331106

    • PubMed
    • Google Scholar
70. 1. Tuncdemir SN
    2. Wamsley B
    3. Stam FJ
    4. Osakada F
    5. Goulding M
    6. Callaway EM
    7. Rudy B
    8. Fishell G

    (2016) Early somatostatin interneuron connectivity mediates the maturation of deep layer cortical circuits

    Neuron 89:521–535.

    https://doi.org/10.1016/j.neuron.2015.11.020

    • PubMed
    • Google Scholar
71. Book

    1. Turek F

    (1999)

    Regulation of Sleep and Circadian Rhythms

    CRC Press.

    • Google Scholar
72. 1. Wilson DE
    2. Scholl B
    3. Fitzpatrick D

    (2018) Differential tuning of excitation and inhibition shapes direction selectivity in ferret visual cortex

    Nature 560:97–101.

    https://doi.org/10.1038/s41586-018-0354-1

    • PubMed
    • Google Scholar
73. 1. Winnubst J
    2. Cheyne JE
    3. Niculescu D
    4. Lohmann C

    (2015) Spontaneous activity drives local synaptic plasticity in vivo

    Neuron 87:399–410.

    https://doi.org/10.1016/j.neuron.2015.06.029

    • PubMed
    • Google Scholar
74. 1. Yang JW
    2. An S
    3. Sun JJ
    4. Reyes-Puerta V
    5. Kindler J
    6. Berger T
    7. Kilb W
    8. Luhmann HJ

    (2013) Thalamic network oscillations synchronize ontogenetic columns in the newborn rat barrel cortex

    Cerebral Cortex 23:1299–1316.

    https://doi.org/10.1093/cercor/bhs103

    • PubMed
    • Google Scholar

Author details

1. Soledad Domínguez

   Institute for Genomic Medicine, Columbia University Medical Center, New York, United States

   Contribution

   Conceptualization, Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft

   Contributed equally with

   Liang Ma and Han Yu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-0378-2442

2. Liang Ma

   1. Institute for Genomic Medicine, Columbia University Medical Center, New York, United States
   2. Department of Biomedical Engineering, Columbia University, New York, United States

   Contribution

   Data curation, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing

   Contributed equally with

   Soledad Domínguez and Han Yu

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0003-3883-1902

3. Han Yu

   Department of Electrical Engineering, Columbia University, New York, United States

   Contribution

   Software, Formal analysis, Validation, Investigation, Visualization, Methodology, Writing - original draft, Writing - review and editing

   Contributed equally with

   Soledad Domínguez and Liang Ma

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-7110-7716

4. Gabrielle Pouchelon

   The Stanley Center at the Broad, Cambridge, United States

   Contribution

   Conceptualization, Validation, Investigation, Visualization, Methodology

   Competing interests

   No competing interests declared

5. Christian Mayer

   Max Planck Institute of Neurobiology, Martinsried, Germany

   Contribution

   Conceptualization, Data curation, Validation, Visualization, Methodology

   Competing interests

   No competing interests declared

6. George D Spyropoulos

   Department of Electrical Engineering, Columbia University, New York, United States

   Contribution

   Data curation, Visualization, Methodology

   Competing interests

   No competing interests declared

7. Claudia Cea

   Department of Electrical Engineering, Columbia University, New York, United States

   Contribution

   Data curation, Visualization, Methodology

   Competing interests

   No competing interests declared

8. György Buzsáki

   1. Neuroscience Institute and Department of Neurology New York University Langone Medical Center, New York, United States
   2. Center for Neural Science, New York University, New York, United States

   Contribution

   Conceptualization, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-3100-4800

9. Gordon Fishell

   1. The Stanley Center at the Broad, Cambridge, United States
   2. Department of Neurobiology, Harvard Medical School, Boston, United States

   Contribution

   Conceptualization, Writing - review and editing

   Competing interests

   No competing interests declared

   "This ORCID iD identifies the author of this article:" 0000-0002-9640-9278

10. Dion Khodagholy

    Department of Electrical Engineering, Columbia University, New York, United States

    Contribution

    Conceptualization, Resources, Data curation, Software, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing

    For correspondence

    dk2955@Columbia.edu

    Competing interests

    No competing interests declared

11. Jennifer N Gelinas

    1. Institute for Genomic Medicine, Columbia University Medical Center, New York, United States
    2. Department of Biomedical Engineering, Columbia University, New York, United States
    3. Department of Neurology, Columbia University Medical Center, New York, United States

    Contribution

    Conceptualization, Resources, Data curation, Formal analysis, Supervision, Funding acquisition, Validation, Investigation, Visualization, Methodology, Writing - original draft, Project administration, Writing - review and editing

    For correspondence

    jng2146@cumc.columbia.edu

    Competing interests

    No competing interests declared

    "This ORCID iD identifies the author of this article:" 0000-0002-1164-638X

• 1,875

  views

• 323

  downloads

• 11

  citations

Views, downloads and citations are aggregated across all versions of this paper published by eLife.