Abstracts

Cerebral Cortex Connectomics

The mapping of neuronal connectivity is one of the main challenges in neuroscience. Only with the knowledge of wiring diagrams is it possible to understand the computational capacities of neuronal networks, both in the sensory periphery, and especially in the mammalian cerebral cortex. Our methods for dense circuit mapping are based on 3-dimensional electron microscopy (EM) imaging of tissue, which allows imaging nerve tissue at nanometer-scale resolution across substantial volumes, extending to more than one millimeter on the side, followed by AI-based image analysis to obtain dense connectivity maps, or connectomes. With these we have recently mapped local circuitry in mouse and human cortex, determining learning-related synaptic traces, inhibitory axonal development, and discovering an expanded interneuron-to-interneuron network in the human cortex. We are currently screening cortical connectomes across age, disease states and experience to obtain a deeper understanding of their relevance for individual behavioral performance and brain pathology.

Evolution of central neural circuits: state of the art and perspectives

Neural circuits evolved over hundreds of millions of years in animals which adapted to a vast range of environments. The neural architecture of animals we see today resulted from an interplay between adaptive changes fixed by selection and developmental constraints inherited from ancestors. Thus, insight gained from evolution will help us understand the way neural circuits are configured and how they function.

Traditionally, peripheral sensory systems have been more accessible to compare across species, as opposed to studying neurons deep inside the brain. As a result, sensory systems have been proposed as a hot spot for evolutionary change, while the central brain is thought to be more conserved. However, recent work comparing closely related species of both vertebrates and invertebrates revealed changes in central circuits which underly differences in behaviour. Are central circuits more evolvable than we expected? What general principles of neural evolution can we draw from comparing closely related species?

I will start by synthesizing progress on this topic, focusing on three aspects of neural circuits that change over evolutionary time: synaptic connectivity, neuromodulation, and neurons.

Drawing from my doctoral dissertation, I will present how a developmental process, programmed cell death, appears to play a key role in changing the number and types of central neurons in the insect nervous system.

Taking on a different perspective — choosing differences in ecological niches between species as a starting point for investigation — I will talk about my current work on neural circuit changes during the evolution of host-specialisation in two fly species adapted to different food sources.

Pyramidal neurons form active, transient, multilayered circuits perturbed by autism-associated mutations 

at the inception of neocortex

Layer 5 pyramidal neurons (L5 PNs) are a major cortical output. Their connectivity approximates a recurrent neural network (RNN), with the majority of inputs coming from other L5 PNs. Understanding the formation of the L5 PN circuitry, from the time at which L5 PNs first enter cortex, would provide insight into how a biological RNN can be constructed. However, this is technically challenging because, in mice, L5 PNs first enter cortex five days before birth. Hence, we developed a method of two-photon imaging from cortical neurons, in living embryos, connected to the dam. Applying this para-uterine imaging method, we observed a structured, bimodal pattern of increased activity in embryonic L5 PNs, with a surprising early peak of activity at embryonic day (E) 14.5, only a day after neurons first migrate into the nascent cortex. Further, these neurons already respond to glutamatergic agonists, and show TTX-sensitive active conductances. Intriguingly, pairwise correlations across embryonic L5 PNs were greater than expected by chance. However, these correlations were independent of distance, and did not form waves. Throughout embryonic development, the precise temporal variation in activity in L5 PNs coincided with changes in circuit organization. This suggests that correlated activity serves a specific function. The para-uterine imaging method gives us all-optical access to characterizing this function and, thereby, to understanding this biological RNN's construction.

Emergence of organization and computation in neural circuits across scales

Many nervous systems develop and form functional circuits through a long period of development involving a myriad of mechanisms. Some of these are determined by genes and molecules, while others depend on neural activity patterns. I will present how these diverse mechanisms work together to set up neural circuits shortly after an animal is born, enabling it to gradually acquire its cognitive and behavioral capabilities. I will focus on the visual system, and demonstrate how neural circuits became established and capable of performing different computations. I will focus on some specific mechanisms such as inhibitory synaptic plasticity and the emergence of excitatory and inhibitory balance, and its role in the detection of novel stimuli. 

Multiplicative disinhibition and the logic of motion vision

 Nonlinear, multiplication-like operations carried out by individual nerve cells greatly enhance the computational power of a neural system, but our understanding of their biophysical implementation is scant. We pursue this problem in the on motion vision circuit of Drosophila melanogaster, where neural activity and connectivity are highly stereotyped. We record the membrane potentials of direction-selective T4 neurons and of each of their five columnar input elements in vivo and under identical conditions in response to visual and pharmacological stimuli. Our electrophysiological measurements and conductance-based simulations suggest a passive supralinear interaction between two distinct types of synapse on the T4 dendrite. We show that this multiplication-like operation arises from the coincidence of cholinergic excitation and release from glutamatergic inhibition. The latter depends on the expression of the glutamate-gated chloride channel GluClα in T4 neurons, which sharpens the directional tuning of the cells and shapes the optomotor behaviour of the animals. Interacting pairs of shunting inhibitory and excitatory synapses have long been postulated as an analogue approximation of a multiplication, which is integral to theories of motion detection, sound localization, and sensorimotor control. Based on information about approximately 85% of a T4 neuron's dendritic input signals, we provide both a detailed biophysical account and an intuitive understanding of how a single neuron uses multiplicative disinhibition to compute the direction of visual motion.

Nature over Nurture: Functional neuronal circuits emerge in the absence of developmental activity

During development, the complex neuronal circuitry of the brain arises from limited information contained in the genome. After the genetic code instructs the birth of neurons, the emergence of brain regions, and the formation of axon tracts, it is believed that neuronal activity plays a critical role in shaping circuits for behavior. Current AI technologies are modeled after the same principle: connections in an initial weight matrix are pruned and strengthened by activity-dependent signals until the network can sufficiently generalize a set of inputs into outputs. Here, we challenge these learning-dominated assumptions by quantifying the contribution of neuronal activity to the development of visually guided swimming behavior in larval zebrafish. Intriguingly, dark-rearing zebrafish revealed that visual experience has no effect on the emergence of the optomotor response (OMR). We then raised animals under conditions where neuronal activity was pharmacologically silenced from organogenesis onward using the sodium-channel blocker tricaine. Strikingly, after washout of the anesthetic, animals performed swim bouts and responded to visual stimuli with 75% accuracy in the OMR paradigm. After shorter periods of silenced activity OMR performance stayed above 90% accuracy, calling into question the importance and impact of classical critical periods for visual development. Detailed quantification of the emergence of functional circuit properties by brain-wide imaging experiments confirmed that neuronal circuits came ‘online’ fully tuned and without the requirement for activity-dependent plasticity. Thus, contrary to what you learned on your mother's knee, complex sensory guided behaviors can be wired up innately by activity-independent developmental mechanisms.

Self-Organization in Brain Wiring

How is the brain 'genetically encoded'?  The genome contains information to grow a brain, not information that describes the brain.  Navigation through the developing brain less resembles guidance through a city grid with fixed addresses than the navigation of a city under construction, where the final address may not exist at the beginning of the journey.  This seminar will explore the question how developmental self-organization can ensure both specificity and robustness of brain wiring using the Drosophila visual system as a model.

Building bilaterian brains: innovations in molecular machinery, cell types, and circuit architecture

TBA

Interplay between homeostatic synaptic scaling and homeostatic structural plasticity

One of the extensively researched homeostatic processes is synaptic scaling, involving compensatory synaptic adjustments to maintain the neuronal firing rate in a physiological range. This is achieved through the modulation of synaptic receptors, neurotransmitters, and morphology. Nevertheless, while there is an abundance of literature on the electrophysiological aspects of homeostatic scaling, our understanding of the structural changes in synapses and dendritic trees remains relatively limited. I will be presenting research conducted in organotypic slice cultures, focusing on homeostatic functional and structural plasticity during the postnatal maturation of neural networks and following partial denervation. Through a combination of experiments in tissue cultures and computational modeling, we introduce a biphasic structural plasticity rule based on spine numbers. This rule adds a redundant and heterogeneous dimension to the existing synaptic-weight-based homeostatic synaptic scaling rule, with important implications for non-invasive brain stimulation techniques aimed at modulating neural networks in health and disease.

Homeostatic plasticity and external input shape neural network dynamics 

TBA

Panta rhei: the importance of silent synapses for vision and visual plasticity

Postsynaptic density protein 95 (PSD-95) is an important signalling scaffold of the PSD of excitatory synapses. The Schlüter and Löwel labs have previously shown that PSD-95 dependent silent synapse maturation (i.e. incorporation of AMPA receptors into nascent AMPA-receptor silent synapses) closes the critical period for ocular dominance plasticity in mouse primary visual cortex, a time window of heightened brain plasticity during which experience refines synaptic connections to achieve mature functionality: PSD-95 knock-out (KO) mice display both functional and structural hallmarks of critical period plasticity into adulthood, and synapses do not properly mature (Huang et al 2015, Favaro et al 2018, Yusifov et al 2021). Since the development of binocularity happens during the critical period, we hypothesized that PSD-95 KO mice should display compromised binocular vision. New data show that this is indeed the case. Thus proper silent synapse maturation is also essential for the development of binocular vision, underscoring the importance of silent synapses for experience-dependent refinement of cortical circuitry.

Different time scales of motor learning

Animals use afferent feedback to rapidly correct ongoing movements in the presence of a perturbation. Repeated exposure to a predictable perturbation leads to behavioural adaptation that counteracts its effects. Primary motor cortex (M1) is intimately involved in both processes, integrating inputs from various sensorimotor brain regions to update the motor output. Here, we investigate whether feedback-based motor control and motor adaptation may share a common implementation in M1 circuits. We trained a recurrent neural network to control its own output through an error feedback signal, which allowed it to recover rapidly from external perturbations. A biologically plausible plasticity rule based on this same feedback signal allowed the network to learn to counteract persistent perturbations through a trial-by-trial process, in a manner that reproduced several key aspects of human adaptation. Moreover, the resultant network activity changes were also present in neural population recordings from monkey primary motor cortex. Online movement correction and longer-term motor adaptation may thus share a common implementation in neural circuits

Frozen algorithms: how the brain’s wiring facilitates learning

Synapses and neural connectivity are plastic and shaped by experience. But to what extent does connectivity itself influence the ability of a neural circuit to learn? The Drosophila Mushroom Body exhibits a striking modularity: it is organised into 15 compartments. Each seems to learn associations independently, varying in intrinsic parameters such as optimism or learning speed. As such, overall behaviour is determined by the weighted vote of a bank of intrinsically biased, simultaneously-learning compartments. We show how, why, and when this architectural motif, which recurs qualitatively in mammalian brain areas such as the cerebellum, improves learning performance. In particular, animals need to make reasonable predictions in learning scenarios where the environment is liable to unpredictably change, and/or they have limited experience and prior knowledge. We build a theory suggesting modular architectures excel in these 'flexible learning' paradigms, at the cost of lowered best-case performance in static, data-rich conditions. Our theory predicts roles for as-yet unexplained intercompartmental wiring in the Mushroom Body across learning scenarios.