Henry Markram

Interneurons of the neocortical inhibitory system

Mammals adapt to a rapidly changing world because of the sophisticated cognitive functions that are supported by the neocortex. The neocortex, which forms almost 80% of the human brain, seems to have arisen from repeated duplication of a stereotypical microcircuit template with subtle specializations for different brain regions and species. The quest to unravel the blueprint of this template started more than a century ago and has revealed an immensely intricate design. The largest obstacle is the daunting variety of inhibitory interneurons that are found in the circuit. This review focuses on the organizing principles that govern the diversity of inhibitory interneurons and their circuits.

Real-time computing without stable states: a new framework for neural computation based on perturbations

A key challenge for neural modeling is to explain how a continuous stream of multimodal input from a rapidly changing environment can be processed by stereotypical recurrent circuits of integrate-and-fire neurons in real time. We propose a new computational model for real-time computing on time-varying input that provides an alternative to paradigms based on Turing machines or attractor neural networks. It does not require a task-dependent construction of neural circuits. Instead, it is based on principles of high-dimensional dynamical systems in combination with statistical learning theory and can be implemented on generic evolved or found recurrent circuitry. It is shown that the inherent transient dynamics of the high-dimensional dynamical system formed by a sufficiently large and heterogeneous neural circuit may serve as universal analog fading memory. Readout neurons can learn to extract in real time from the current state of such recurrent neural circuit information about current and past inputs that may be needed for diverse tasks. Stable internal states are not required for giving a stable output, since transient internal states can be transformed by readout neurons into stable target outputs due to the high dimensionality of the dynamical system. Our approach is based on a rigorous computational model, the liquid state machine, that, unlike Turing machines, does not require sequential transitions between well-defined discrete internal states. It is supported, as the Turing machine is, by rigorous mathematical results that predict universal computational power under idealized conditions, but for the biologically more realistic scenario of real-time processing of time-varying inputs. Our approach provides new perspectives for the interpretation of neural coding, the design of experiments and data analysis in neurophysiology, and the solution of problems in robotics and neurotechnology.

Petilla terminology: nomenclature of features of GABAergic interneurons of the cerebral cortex.

Neuroscience produces a vast amount of data from an enormous diversity of neurons. A neuronal classification system is essential to organize such data and the knowledge that is derived from them. Classification depends on the unequivocal identification of the features that distinguish one type of neuron from another. The problems inherent in this are particularly acute when studying cortical interneurons. To tackle this, we convened a representative group of researchers to agree on a set of terms to describe the anatomical, physiological and molecular features of GABAergic interneurons of the cerebral cortex. The resulting terminology might provide a stepping stone towards a future classification of these complex and heterogeneous cells. Consistent adoption will be important for the success of such an initiative, and we also encourage the active involvement of the broader scientific community in the dynamic evolution of this project.

The blue brain project

IBMs Blue Gene supercomputer allows a quantum leap in the level of detail at which the brain can be modelled. I argue that the time is right to begin assimilating the wealth of data that has been accumulated over the past century and start building biologically accurate models of the brain from first principles to aid our understanding of brain function and dysfunction.

Physiology and anatomy of synaptic connections between thick tufted pyramidal neurones in the developing rat neocortex.

1. Dual voltage recordings were made from pairs of adjacent, synaptically connected thick tufted layer 5 pyramidal neurones in brain slices of young rat (14‐16 days) somatosensory cortex to examine the physiological properties of unitary EPSPs. Pre‐ and postsynaptic neurones were filled with biocytin and examined in the light and electron microscope to quantify the morphology of axonal and dendritic arbors and the number and location of synaptic contacts on the target neurone. 2. In 138 synaptic connections between pairs of pyramidal neurones 96 (70%) were unidirectional and 42 (30%) were bidirectional. The probability of finding a synaptic connection in dual recordings was 0.1. Unitary EPSPs evoked by a single presynaptic action potential (AP) had a mean peak amplitude ranging from 0.15 to 5.5 mV in different connections with a mean of 1.3 +/‐ 1.1 mV, a latency of 1.7 +/‐ 0.9 ms, a 20‐80% rise time of 2.9 +/‐ 2.3 ms and a decay time constant of 40 +/‐ 18 ms at 32‐24 degrees C and ‐60 +/‐ 2 mV membrane potential. 3. Peak amplitudes of unitary EPSPs fluctuated randomly from trial to trial. The coefficient of variation (c.v.) of the unitary EPSP amplitudes ranged from 0.13 to 2.8 in different synaptic connections (mean, 0.52; median, 0.41). The percentage of failures of single APs to evoke a unitary EPSP ranged from 0 to 73% (mean, 14%; median, 7%). Both c.v. and percentage of failures decreased with increasing mean EPSP amplitude. 4. Postsynaptic glutamate receptors which mediate unitary EPSPs at ‐60 mV were predominantly of the L‐alpha‐amino‐3‐hydroxy‐5‐methyl‐4‐isoxazolepropionate (AMPA) receptor type. Receptors of the N‐methyl‐D‐aspartate (NMDA) type contributed only a small fraction (< 20%) to the voltage‐time integral of the unitary EPSP at ‐60 mV, but their contribution increased at more positive membrane potentials. 5. Branching patterns of dendrites and axon collaterals of forty‐five synaptically connected neurones, when examined in the light microscope, indicated that the axonal and dendritic anatomy of both projecting and target neurones and of uni‐ and bidirectionally connected neurones was uniform. 6. The number of potential synaptic contacts formed by a presynaptic neurone on a target neurone varied between four and eight (mean, 5.5 +/‐ 1.1 contacts; n = 19 connections). Synaptic contacts were preferentially located on basal dendrites (63%, 82 +/‐ 35 microns from the soma, n = 67) and apical oblique dendrites (27%, 145 +/‐ 59 microns, n = 29), and 35% of all contacts were located on tertiary basal dendritic branches. The mean geometric distances (from the soma) of the contacts of a connection varied between 80 and 585 microns (mean, 147 microns; median, 105 microns). The correlation between EPSP amplitude and the number of morphologically determined synaptic contacts or the mean geometric distances from the soma was only weak (correlation coefficients were 0.2 and 0.26, respectively). 7. Compartmental models constructed from camera lucida drawings of eight target neurones showed that synaptic contacts were located at mean electrotonic distances between 0.07 and 0.33 from the soma (mean, 0.13). Simulations of unitary EPSPs, assuming quantal conductance changes with fast rise time and short duration, indicated that amplitudes of quantal EPSPs at the soma were attenuated, on average, to < 10% of dendritic EPSPs and varied in amplitude up to 10‐fold depending on the dendritic location of synaptic contacts. The inferred quantal peak conductance increase varied between 1.5 and 5.5 nS (mean, 3 nS). 8. The combined physiological and morphological measurements in conjunction with EPSP simulations indicated that the 20‐fold range in efficacy of the synaptic connections between thick tufted pyramidal neurones, which have their synaptic contacts preferentially located on basal and apical oblique dendrites, was due to differences in transmitter release probability of the projecting neurones and, to a lesser extent, to differenc

Neural networks with dynamic synapses

Transmission across neocortical synapses depends on the frequency of presynaptic activity (Thomson & Deuchars, 1994). Interpyramidal synapses in layer V exhibit fast depression of synaptic transmission, while other types of synapses exhibit facilitation of transmission. To study the role of dynamic synapses in network computation, we propose a unified phenomenological model that allows computation of the postsynaptic current generated by both types of synapses when driven by an arbitrary pattern of action potential (AP) activity in a presynaptic population. Using this formalism, we analyze different regimes of synaptic transmission and demonstrate that dynamic synapses transmit different aspects of the presynaptic activity depending on the average presynaptic frequency. The model also allows for derivation of mean-field equations, which govern the activity of large, interconnected networks. We show that the dynamics of synaptic transmission results in complex sets of regular and irregular regimes of network activity.

Disynaptic Inhibition between Neocortical Pyramidal Cells Mediated by Martinotti Cells

Reliable activation of inhibitory pathways is essential for maintaining the balance between excitation and inhibition during cortical activity. Little is known, however, about the activation of these pathways at the level of the local neocortical microcircuit. We report a disynaptic inhibitory pathway among neocortical pyramidal cells (PCs). Inhibitory responses were evoked in layer 5 PCs following stimulation of individual neighboring PCs with trains of action potentials. The probability for inhibition between PCs was more than twice that of direct excitation, and inhibitory responses increased as a function of rate and duration of presynaptic discharge. Simultaneous somatic and dendritic recordings indicated that inhibition originated from PC apical and tuft dendrites. Multineuron whole-cell recordings from PCs and interneurons combined with morphological reconstructions revealed the mediating interneurons as Martinotti cells. Martinotti cells received facilitating synapses from PCs and formed reliable inhibitory synapses onto dendrites of neighboring PCs. We describe this feedback pathway and propose it as a central mechanism for regulation of cortical activity.

Dendritic calcium transients evoked by single back-propagating action potentials in rat neocortical pyramidal neurons

1. Dendrites of rat neocortical layer V pyramidal neurons were loaded with the Ca2+ indicator dye Calcium Green‐1 (CG‐1) or fluo‐3, and the mechanisms which govern action potential (AP)‐evoked transient changes in dendritic cytosolic Ca2+ concentration ([Ca2+]i) were examined. APs were initiated either by synaptic stimulation or by depolarizing the soma or dendrite by current injection, and changes in fluorescence of the indicator dye were measured in the proximal 170 microns of the apical dendrite. 2. Simultaneous two‐pipette recordings of APs from the soma and apical dendrite, and dendritic fluorescence imaging indicated that a single AP propagating from the soma into the apical dendrite evokes a rapid transient increase in fluorescence indicating a transient increase in [Ca2+]i. At 35‐37 degrees C the decay time constant of the fluorescence transient following an AP was around 80 ms. 3. Voltage‐activated Ca2+ channels (VACCs) of several subtypes mediated the AP‐evoked fluorescence transient in the proximal (100‐170 microns) apical dendrite. The AP‐evoked fluorescence transient resulted from Ca2+ entry through L‐type (nifedipine sensitive; 25%), N‐type (omega‐conotoxin GVIA sensitive; 28%) and P‐type (omega‐agatoxin IVA sensitive; 10%) Ca2+ channels and through Ca2+ channels (R‐type) not sensitive to L‐, N‐ and P‐type Ca2+ channel blockers (cadmium ion sensitive; 37%). 4. The decay time course of the dendritic fluorescence transient was prolonged by the blockers of endoplasmic reticulum (ER) Ca(2+)‐ATPase, cyclopiazonic acid and thapsigargin, suggesting that uptake of Ca2+ into the ER in dendrites governs clearance of dendritic Ca2+. 5. The decay time course of the fluorescence transient was slightly prolonged by benzamil, a blocker of plasma membrane Na(+)‐Ca2+ exchange and by calmidazolium, a blocker of the calmodulin‐dependent plasma membrane Ca(2+)‐ATPase, suggesting that these pathways are less important for dendrite Ca2+ clearance following a single AP. Neither the mitochondrial uncoupler carbonyl cyanide p‐(trifluoromethoxy)phenylhydrazone (FCCP) nor the blocker of Ca2+ uptake into mitochondria, Ruthenium Red, had any measurable effect on the decay time course of the fluorescence transient. 6. Dendritic fluorescence transients measured during trains of dendritic APs began to summate at impulse frequencies of 5 APs s‐1. At higher frequencies APs caused a concerted and maintained elevation of dendritic fluorescence during the train.(ABSTRACT TRUNCATED AT 400 WORDS)

A synaptic organizing principle for cortical neuronal groups

Neuronal circuitry is often considered a clean slate that can be dynamically and arbitrarily molded by experience. However, when we investigated synaptic connectivity in groups of pyramidal neurons in the neocortex, we found that both connectivity and synaptic weights were surprisingly predictable. Synaptic weights follow very closely the number of connections in a group of neurons, saturating after only 20% of possible connections are formed between neurons in a group. When we examined the network topology of connectivity between neurons, we found that the neurons cluster into small world networks that are not scale-free, with less than 2 degrees of separation. We found a simple clustering rule where connectivity is directly proportional to the number of common neighbors, which accounts for these small world networks and accurately predicts the connection probability between any two neurons. This pyramidal neuron network clusters into multiple groups of a few dozen neurons each. The neurons composing each group are surprisingly distributed, typically more than 100 μm apart, allowing for multiple groups to be interlaced in the same space. In summary, we discovered a synaptic organizing principle that groups neurons in a manner that is common across animals and hence, independent of individual experiences. We speculate that these elementary neuronal groups are prescribed Lego-like building blocks of perception and that acquired memory relies more on combining these elementary assemblies into higher-order constructs.

Reconstruction and Simulation of Neocortical Microcircuitry

UNLABELLED We present a first-draft digital reconstruction of the microcircuitry of somatosensory cortex of juvenile rat. The reconstruction uses cellular and synaptic organizing principles to algorithmically reconstruct detailed anatomy and physiology from sparse experimental data. An objective anatomical method defines a neocortical volume of 0.29 ± 0.01 mm(3) containing ~31,000 neurons, and patch-clamp studies identify 55 layer-specific morphological and 207 morpho-electrical neuron subtypes. When digitally reconstructed neurons are positioned in the volume and synapse formation is restricted to biological bouton densities and numbers of synapses per connection, their overlapping arbors form ~8 million connections with ~37 million synapses. Simulations reproduce an array of in vitro and in vivo experiments without parameter tuning. Additionally, we find a spectrum of network states with a sharp transition from synchronous to asynchronous activity, modulated by physiological mechanisms. The spectrum of network states, dynamically reconfigured around this transition, supports diverse information processing strategies. PAPERCLIP VIDEO ABSTRACT.