Brain orchestra under spontaneous conditions: Identifying communication modules from the functional architecture of area V1.

We used two-photon imaging to record from granular and supragranular layers in mouse primary visual cortex (V1) under spontaneous conditions and applied an extension of the spike time tiling coefficient (STTC; introduced by Cutts and Eglen) to map functional connectivity architecture within and across layers. We made several observations: Approximately, 19-34% of neuronal pairs within 300 µm of each other exhibit statistically significant functional connections, compared to ~10% at distances of 1mm or more. As expected, neuronal pairs with similar tuning functions exhibit a significant, though relatively small, increase in the fraction of functional inter-neuronal correlations. In contrast, internal state as reflected by pupillary diameter or aggregate neuronal activity appears to play a much stronger role in determining inter-neuronal correlation distributions and topography. Overall, inter-neuronal correlations appear to be slightly more prominent in L4. The first-order functionally connected (i.e., direct) neighbors of neurons determine the hub structure of the V1 microcircuit. L4 exhibits a nearly flat degree of connectivity distribution, extending to higher values than seen in supragranular layers, whose distribution drops exponentially. In all layers, functional connectivity exhibits small-world characteristics and network robustness. The probability of firing of L2/3 pyramidal neurons can be predicted as a function of the aggregate activity in their first-order functionally connected partners within L4, which represent their putative input group. The functional form of this prediction conforms well to a ReLU function, reaching up to firing probability one in some neurons. Interestingly, the properties of L2/3 pyramidal neurons differ based on the size of their L4 functional connectivity group. Specifically, L2/3 neurons with small layer-4 degrees of connectivity appear to be more sensitive to the firing of their L4 functional connectivity partners, suggesting they may be more effective at transmitting synchronous activity downstream from L4. They also appear to fire largely independently from each other, compared to neurons with high layer-4 degrees of connectivity, and are less modulated by changes in pupil size and aggregate population dynamics. Information transmission is best viewed as occurring from neuronal ensembles in L4 to neuronal ensembles in L2/3. Under spontaneous conditions, we were able to identify such candidate neuronal ensembles, which exhibit high sensitivity, precision, and specificity for L4 to L2/3 information transmission. In sum, functional connectivity analysis under spontaneous activity conditions reveals a modular neuronal ensemble architecture within and across granular and supragranular layers of mouse primary visual cortex. Furthermore, modules with different degrees of connectivity appear to obey different rules of engagement and communication across the V1 columnar circuit.

Inhibition of Elevated Ras-MAPK Signaling Normalizes Enhanced Motor Learning and Excessive Clustered Dendritic Spine Stabilization in the MECP2-Duplication Syndrome Mouse Model of Autism.

The inflexible repetitive behaviors and "insistence on sameness" seen in autism imply a defect in neural processes controlling the balance between stability and plasticity of synaptic connections in the brain. It has been proposed that abnormalities in the Ras-ERK/MAPK pathway, a key plasticity-related cell signaling pathway known to drive consolidation of clustered synaptic connections, underlie altered learning phenotypes in autism. However, a link between altered Ras-ERK signaling and clustered dendritic spine plasticity has yet to be explored in an autism animal model in vivo The formation and stabilization of dendritic spine clusters is abnormally increased in the MECP2-duplication syndrome mouse model of syndromic autism, suggesting that ERK signaling may be increased. Here, we show that the Ras-ERK pathway is indeed hyperactive following motor training in MECP2-duplication mouse motor cortex. Pharmacological inhibition of ERK signaling normalizes the excessive clustered spine stabilization and enhanced motor learning behavior in MECP2-duplication mice. We conclude that hyperactive ERK signaling may contribute to abnormal clustered dendritic spine consolidation and motor learning in this model of syndromic autism.

Excessive Formation and Stabilization of Dendritic Spine Clusters in the MECP2-Duplication Syndrome Mouse Model of Autism.

Autism-associated genetic mutations may perturb the balance between stability and plasticity of synaptic connections in the brain. Here, we report an increase in the formation and stabilization of dendritic spines in the cerebral cortex of the mouse model of MECP2-duplication syndrome, a high-penetrance form of syndromic autism. Increased stabilization is mediated entirely by spines that form cooperatively in 10-µm clusters and is observable across multiple cortical areas both spontaneously and following motor training. Excessive stability of dendritic spine clusters could contribute to behavioral rigidity and other phenotypes in syndromic autism.

Dictionary Learning for Spontaneous Neural Activity Modeling.

Modeling the activity of an ensemble of neurons can provide critical insights into the workings of the brain. In this work we examine if learning based signal modeling can contribute to a high quality modeling of neuronal signal data. To that end, we employ the sparse coding and dictionary learning schemes for capturing the behavior of neuronal responses into a small number of representative prototypical signals. Performance is measured by the reconstruction quality of clean and noisy test signals, which serves as an indicator of the generalization and discrimination capabilities of the learned dictionaries. To validate the merits of the proposed approach, a novel dataset of the actual recordings from 183 neurons from the primary visual cortex of a mouse in early postnatal development was developed and investigated. The results demonstrate that high quality modeling of testing data can be achieved from a small number of training examples and that the learned dictionaries exhibit significant specificity when introducing noise.