Limits...
Irregular spiking of pyramidal neurons organizes as scale-invariant neuronal avalanches in the awake state.

Bellay T, Klaus A, Seshadri S, Plenz D - Elife (2015)

Bottom Line: As the animal transitions from the anesthetized to awake state, spontaneous single neuron firing increases in irregularity and assembles into scale-invariant avalanches at the group level.In vitro spike avalanches emerged naturally yet required balanced excitation and inhibition.This demonstrates that neuronal avalanches are linked to the global physiological state of wakefulness and that cortical resting activity organizes as avalanches from firing of local PN groups to global population activity.

View Article: PubMed Central - PubMed

Affiliation: Section on Critical Brain Dynamics, National Institute of Mental Health, Bethesda, United States.

ABSTRACT
Spontaneous fluctuations in neuronal activity emerge at many spatial and temporal scales in cortex. Population measures found these fluctuations to organize as scale-invariant neuronal avalanches, suggesting cortical dynamics to be critical. Macroscopic dynamics, though, depend on physiological states and are ambiguous as to their cellular composition, spatiotemporal origin, and contributions from synaptic input or action potential (AP) output. Here, we study spontaneous firing in pyramidal neurons (PNs) from rat superficial cortical layers in vivo and in vitro using 2-photon imaging. As the animal transitions from the anesthetized to awake state, spontaneous single neuron firing increases in irregularity and assembles into scale-invariant avalanches at the group level. In vitro spike avalanches emerged naturally yet required balanced excitation and inhibition. This demonstrates that neuronal avalanches are linked to the global physiological state of wakefulness and that cortical resting activity organizes as avalanches from firing of local PN groups to global population activity.

No MeSH data available.


Related in: MedlinePlus

Neuronal avalanche dynamics recorded in vitro using GCaMP3 in organotypic cortex cultures.(A) GCaMP3 covers a wide dynamic range of spike bursts but fails to reliably capture burst sizes <3 APs. Loose-patch recordings combined with 2-PI of cortex slice cultures grown for ∼2 weeks. Fluorescence traces triggered on spontaneously occurring AP bursts and sorted by the number of spontaneous APs/250 ms. Note the relative insensitivity of GCaMP3 to small bursts establishing a natural threshold of λthr in the data acquisition. Single PN. (B) Summary of change in fluorescent intensity ΔF/F, which increases linearly with spontaneous APs/250 ms, but is undetectably low at sizes <3 APs (n = 8 cells; color codes). Broken lines: regression for individual neurons. (C) Average λnorm distribution for all ROIs. Vertical line, λnorm = 1. (D) Corresponding average distribution in normalized quiescent time intervals, IBInorm. Vertical line, IBInorm = 1. (E) Mean λ autocorrelation function. Note the strong decay in autocorrelation demonstrating temporal correlations for up to 10 s. (F) CV in AP firing is much larger than 1 and independent of λthr. (G) Representative traces of changes in fluorescent intensity (ΔF/F) simultaneously recorded over time from GCaMP3-expressing L2/3 PNs in vitro. Note the relatively sparse activity compared to YC2.60 recordings and large fluctuations in burst amplitudes. (H) Cluster rate as a function of λthr. (I) Cluster size distributions of individual cultures based on GCaMP3 recordings. (J) Average power-law distributions in normalized cluster size s for different λthr (color scale). Observed cascade sizes with the burst indicator GCaMP3 also follow a power law up to the cut-off of s = 1. Dashed line, slope = −1.5.DOI:http://dx.doi.org/10.7554/eLife.07224.017
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4492006&req=5

fig7s3: Neuronal avalanche dynamics recorded in vitro using GCaMP3 in organotypic cortex cultures.(A) GCaMP3 covers a wide dynamic range of spike bursts but fails to reliably capture burst sizes <3 APs. Loose-patch recordings combined with 2-PI of cortex slice cultures grown for ∼2 weeks. Fluorescence traces triggered on spontaneously occurring AP bursts and sorted by the number of spontaneous APs/250 ms. Note the relative insensitivity of GCaMP3 to small bursts establishing a natural threshold of λthr in the data acquisition. Single PN. (B) Summary of change in fluorescent intensity ΔF/F, which increases linearly with spontaneous APs/250 ms, but is undetectably low at sizes <3 APs (n = 8 cells; color codes). Broken lines: regression for individual neurons. (C) Average λnorm distribution for all ROIs. Vertical line, λnorm = 1. (D) Corresponding average distribution in normalized quiescent time intervals, IBInorm. Vertical line, IBInorm = 1. (E) Mean λ autocorrelation function. Note the strong decay in autocorrelation demonstrating temporal correlations for up to 10 s. (F) CV in AP firing is much larger than 1 and independent of λthr. (G) Representative traces of changes in fluorescent intensity (ΔF/F) simultaneously recorded over time from GCaMP3-expressing L2/3 PNs in vitro. Note the relatively sparse activity compared to YC2.60 recordings and large fluctuations in burst amplitudes. (H) Cluster rate as a function of λthr. (I) Cluster size distributions of individual cultures based on GCaMP3 recordings. (J) Average power-law distributions in normalized cluster size s for different λthr (color scale). Observed cascade sizes with the burst indicator GCaMP3 also follow a power law up to the cut-off of s = 1. Dashed line, slope = −1.5.DOI:http://dx.doi.org/10.7554/eLife.07224.017

Mentions: LFP recordings in cortex slice cultures (Beggs and Plenz, 2003; Stewart and Plenz, 2007; Gireesh and Plenz, 2008) have shown avalanche dynamics to emerge spontaneously in superficial layers. Similarly, spike avalanches have been identified in extracellular unit recordings from dissociated cultures of hippocampus (Mazzoni et al., 2007) and cortex (Pasquale et al., 2008; Tetzlaff et al., 2010; Vincent et al., 2012), although the mesoscopic organization of the tissue was not preserved. Yet, these studies are limited by the unknown composition of the LFP population signal (see ‘Introduction’) and cell types recorded from. For extracellular unit activity, strongly bursting interneurons can dominate large spike clusters in the neuronal population, in which case heavy-tailed cluster size distributions reflect neuronal differences rather than neuronal interactions. In order to demonstrate that avalanche dynamics also capture spatiotemporal activity in L2/3 PNs in vitro, we conducted studies in GECI-expressing cortical slices, co-cultured with VTA to ensure proper maturation of superficial cortical layers (Gireesh and Plenz, 2008) (Figure 7A–D). We recorded AP activity from local groups of L2/3 PNs in vitro (n = 15–80 ROIs) monitored with YC2.60 (Δt = 250 ms; n = 129 movies, n = 35 cultures; Figure 7B–D) and compared the activity to conditions when GABAA (5 µM PTX, n = 8) or AMPA and NMDA-receptor mediated (0.5 µM DNQX, 5 µM AP5, n = 6) synaptic transmission were slightly reduced. Neuronal activity was stable throughout the recording for each condition (Figure 7—figure supplement 1A,B). At the single neuron level, AP firing was irregular, in line with our in vivo results (Figure 7E,F; Figure 7—figure supplement 1C,D). Temporal clustering was present under normal conditions (ACSF) but was reduced during disinhibition or disfacilitation (Figure 7G, PTX and DNQX/AP5, respectively). An intermediate level in correlated AP firing was found under normal conditions (Figure 7H). Correlations between neighboring and distant neurons were highly similar and as expected increased during disinhibition but decreased during disfacilitation (Figure 7H). As described in our in vivo results, the mean rate smoothly declined with increase in λthr (Figure 7—figure supplement 1E), and the number of AP cascades of PN groups peaked in rate at an intermediate threshold λthr (Figure 7—figure supplement 1F) for all three conditions. When processed at the corresponding , cascade sizes under normal conditions distributed according to a power law that was robust to changes in λthr (Figure 7I, left). As expected, the power law was destroyed when spiking activity was shuffled (Figure 7I, right). As previously shown for LFP-based analysis (Plenz, 2012), AP-based cluster size distributions became strongly bimodal during pharmacological disinhibition and slightly bimodal during disfacilitation (Figure 7J, PTX and DNQX/AP5, respectively; Figure 7—figure supplement 2). YC2.60, while being sensitive to single APs, tends to saturate for very strong spike bursts (Yamada et al., 2011). In contrast, the GECI GCaMP3 (Tian et al., 2009) naturally has a higher threshold for AP detection (>3 APs) but reports even strong bursts linearly (Yamada et al., 2011) (Figure 7—figure supplement 3A,B). In line with our expectation of threshold invariance for LFP-based avalanches in the AW monkey (Petermann et al., 2009) and our YC2.60 measurements, we found that AP bursts measured with GCaMP3 were irregular at the single neuron level (Figure 7—figure supplement 3C–H), while AP cascades formed a clear power law (Figure 7—figure supplement 3I,J; n = 9 cultures). These in vitro results demonstrate neuronal avalanches to describe the spatiotemporal spike activity in L2/3 PN groups, that is, sensitive to the balance of excitation and inhibition and can be detected using high-threshold GECIs.10.7554/eLife.07224.014Figure 7.Spatiotemporal clustering in ongoing spiking activity recorded from groups of L2/3 PNs in vitro.


Irregular spiking of pyramidal neurons organizes as scale-invariant neuronal avalanches in the awake state.

Bellay T, Klaus A, Seshadri S, Plenz D - Elife (2015)

Neuronal avalanche dynamics recorded in vitro using GCaMP3 in organotypic cortex cultures.(A) GCaMP3 covers a wide dynamic range of spike bursts but fails to reliably capture burst sizes <3 APs. Loose-patch recordings combined with 2-PI of cortex slice cultures grown for ∼2 weeks. Fluorescence traces triggered on spontaneously occurring AP bursts and sorted by the number of spontaneous APs/250 ms. Note the relative insensitivity of GCaMP3 to small bursts establishing a natural threshold of λthr in the data acquisition. Single PN. (B) Summary of change in fluorescent intensity ΔF/F, which increases linearly with spontaneous APs/250 ms, but is undetectably low at sizes <3 APs (n = 8 cells; color codes). Broken lines: regression for individual neurons. (C) Average λnorm distribution for all ROIs. Vertical line, λnorm = 1. (D) Corresponding average distribution in normalized quiescent time intervals, IBInorm. Vertical line, IBInorm = 1. (E) Mean λ autocorrelation function. Note the strong decay in autocorrelation demonstrating temporal correlations for up to 10 s. (F) CV in AP firing is much larger than 1 and independent of λthr. (G) Representative traces of changes in fluorescent intensity (ΔF/F) simultaneously recorded over time from GCaMP3-expressing L2/3 PNs in vitro. Note the relatively sparse activity compared to YC2.60 recordings and large fluctuations in burst amplitudes. (H) Cluster rate as a function of λthr. (I) Cluster size distributions of individual cultures based on GCaMP3 recordings. (J) Average power-law distributions in normalized cluster size s for different λthr (color scale). Observed cascade sizes with the burst indicator GCaMP3 also follow a power law up to the cut-off of s = 1. Dashed line, slope = −1.5.DOI:http://dx.doi.org/10.7554/eLife.07224.017
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4492006&req=5

fig7s3: Neuronal avalanche dynamics recorded in vitro using GCaMP3 in organotypic cortex cultures.(A) GCaMP3 covers a wide dynamic range of spike bursts but fails to reliably capture burst sizes <3 APs. Loose-patch recordings combined with 2-PI of cortex slice cultures grown for ∼2 weeks. Fluorescence traces triggered on spontaneously occurring AP bursts and sorted by the number of spontaneous APs/250 ms. Note the relative insensitivity of GCaMP3 to small bursts establishing a natural threshold of λthr in the data acquisition. Single PN. (B) Summary of change in fluorescent intensity ΔF/F, which increases linearly with spontaneous APs/250 ms, but is undetectably low at sizes <3 APs (n = 8 cells; color codes). Broken lines: regression for individual neurons. (C) Average λnorm distribution for all ROIs. Vertical line, λnorm = 1. (D) Corresponding average distribution in normalized quiescent time intervals, IBInorm. Vertical line, IBInorm = 1. (E) Mean λ autocorrelation function. Note the strong decay in autocorrelation demonstrating temporal correlations for up to 10 s. (F) CV in AP firing is much larger than 1 and independent of λthr. (G) Representative traces of changes in fluorescent intensity (ΔF/F) simultaneously recorded over time from GCaMP3-expressing L2/3 PNs in vitro. Note the relatively sparse activity compared to YC2.60 recordings and large fluctuations in burst amplitudes. (H) Cluster rate as a function of λthr. (I) Cluster size distributions of individual cultures based on GCaMP3 recordings. (J) Average power-law distributions in normalized cluster size s for different λthr (color scale). Observed cascade sizes with the burst indicator GCaMP3 also follow a power law up to the cut-off of s = 1. Dashed line, slope = −1.5.DOI:http://dx.doi.org/10.7554/eLife.07224.017
Mentions: LFP recordings in cortex slice cultures (Beggs and Plenz, 2003; Stewart and Plenz, 2007; Gireesh and Plenz, 2008) have shown avalanche dynamics to emerge spontaneously in superficial layers. Similarly, spike avalanches have been identified in extracellular unit recordings from dissociated cultures of hippocampus (Mazzoni et al., 2007) and cortex (Pasquale et al., 2008; Tetzlaff et al., 2010; Vincent et al., 2012), although the mesoscopic organization of the tissue was not preserved. Yet, these studies are limited by the unknown composition of the LFP population signal (see ‘Introduction’) and cell types recorded from. For extracellular unit activity, strongly bursting interneurons can dominate large spike clusters in the neuronal population, in which case heavy-tailed cluster size distributions reflect neuronal differences rather than neuronal interactions. In order to demonstrate that avalanche dynamics also capture spatiotemporal activity in L2/3 PNs in vitro, we conducted studies in GECI-expressing cortical slices, co-cultured with VTA to ensure proper maturation of superficial cortical layers (Gireesh and Plenz, 2008) (Figure 7A–D). We recorded AP activity from local groups of L2/3 PNs in vitro (n = 15–80 ROIs) monitored with YC2.60 (Δt = 250 ms; n = 129 movies, n = 35 cultures; Figure 7B–D) and compared the activity to conditions when GABAA (5 µM PTX, n = 8) or AMPA and NMDA-receptor mediated (0.5 µM DNQX, 5 µM AP5, n = 6) synaptic transmission were slightly reduced. Neuronal activity was stable throughout the recording for each condition (Figure 7—figure supplement 1A,B). At the single neuron level, AP firing was irregular, in line with our in vivo results (Figure 7E,F; Figure 7—figure supplement 1C,D). Temporal clustering was present under normal conditions (ACSF) but was reduced during disinhibition or disfacilitation (Figure 7G, PTX and DNQX/AP5, respectively). An intermediate level in correlated AP firing was found under normal conditions (Figure 7H). Correlations between neighboring and distant neurons were highly similar and as expected increased during disinhibition but decreased during disfacilitation (Figure 7H). As described in our in vivo results, the mean rate smoothly declined with increase in λthr (Figure 7—figure supplement 1E), and the number of AP cascades of PN groups peaked in rate at an intermediate threshold λthr (Figure 7—figure supplement 1F) for all three conditions. When processed at the corresponding , cascade sizes under normal conditions distributed according to a power law that was robust to changes in λthr (Figure 7I, left). As expected, the power law was destroyed when spiking activity was shuffled (Figure 7I, right). As previously shown for LFP-based analysis (Plenz, 2012), AP-based cluster size distributions became strongly bimodal during pharmacological disinhibition and slightly bimodal during disfacilitation (Figure 7J, PTX and DNQX/AP5, respectively; Figure 7—figure supplement 2). YC2.60, while being sensitive to single APs, tends to saturate for very strong spike bursts (Yamada et al., 2011). In contrast, the GECI GCaMP3 (Tian et al., 2009) naturally has a higher threshold for AP detection (>3 APs) but reports even strong bursts linearly (Yamada et al., 2011) (Figure 7—figure supplement 3A,B). In line with our expectation of threshold invariance for LFP-based avalanches in the AW monkey (Petermann et al., 2009) and our YC2.60 measurements, we found that AP bursts measured with GCaMP3 were irregular at the single neuron level (Figure 7—figure supplement 3C–H), while AP cascades formed a clear power law (Figure 7—figure supplement 3I,J; n = 9 cultures). These in vitro results demonstrate neuronal avalanches to describe the spatiotemporal spike activity in L2/3 PN groups, that is, sensitive to the balance of excitation and inhibition and can be detected using high-threshold GECIs.10.7554/eLife.07224.014Figure 7.Spatiotemporal clustering in ongoing spiking activity recorded from groups of L2/3 PNs in vitro.

Bottom Line: As the animal transitions from the anesthetized to awake state, spontaneous single neuron firing increases in irregularity and assembles into scale-invariant avalanches at the group level.In vitro spike avalanches emerged naturally yet required balanced excitation and inhibition.This demonstrates that neuronal avalanches are linked to the global physiological state of wakefulness and that cortical resting activity organizes as avalanches from firing of local PN groups to global population activity.

View Article: PubMed Central - PubMed

Affiliation: Section on Critical Brain Dynamics, National Institute of Mental Health, Bethesda, United States.

ABSTRACT
Spontaneous fluctuations in neuronal activity emerge at many spatial and temporal scales in cortex. Population measures found these fluctuations to organize as scale-invariant neuronal avalanches, suggesting cortical dynamics to be critical. Macroscopic dynamics, though, depend on physiological states and are ambiguous as to their cellular composition, spatiotemporal origin, and contributions from synaptic input or action potential (AP) output. Here, we study spontaneous firing in pyramidal neurons (PNs) from rat superficial cortical layers in vivo and in vitro using 2-photon imaging. As the animal transitions from the anesthetized to awake state, spontaneous single neuron firing increases in irregularity and assembles into scale-invariant avalanches at the group level. In vitro spike avalanches emerged naturally yet required balanced excitation and inhibition. This demonstrates that neuronal avalanches are linked to the global physiological state of wakefulness and that cortical resting activity organizes as avalanches from firing of local PN groups to global population activity.

No MeSH data available.


Related in: MedlinePlus