Skip to main content

Cellular and circuit determinants of dendritic computation

Final Report Summary - DENDRITE (Cellular and circuit determinants of dendritic computation)

This project aimed to address a simple question: what is the fundamental unit of computation in the brain? Answering this question is crucial not only for understanding how the brain works, but also if we are to build accurate models of brain function, which require abstraction based on identification of the crucial elements for carrying out computations relevant to behaviour. The project aims specifically to investigate whether single dendrites, or subsets of the dendritic tree, can be used as independent computational units capable of higher-order signal processing, and to relate this to the computations performed by the entire neuron when it is solving a problem during a behavioural task. These aims were addressed using a combination of experimental and theoretical approaches in the mouse cerebellum and neocortex. We have made the following technical and conceptual breakthroughs in the course of this project:

1. Development of new techniques for patterned 2-photon glutamate uncaging onto single dendrites, which have shown that single pyramidal cell dendrites, as well as multiple dendrites, can discriminate spatiotemporal sequences of inputs (Branco et al. Science 2010).

2. Patterned dendritic glutamate uncaging was used to reveal that there exists a gradient of synaptic integration along single pyramidal cell dendrites: proximal inputs tend to sum linearly and require precise temporal coincidence for effective summation, whereas distal inputs are amplified with high gain and integrated over broader time windows (Branco et al., Neuron 2011).

3. Development of a set of algorithms to capture how the detailed structure of the dendritic tree depends on the conservation of cytoplasm and conduction time. These algorithms allow us to replicate the structure of a range of different dendritic trees, which is crucial for creating realistic synthetic neural networks (Cuntz et al. PLoS Computational Biology 2010). These algorithms are freely available on-line ( as an integrated set of Matlab routines (Cuntz et al 2011).

4. Development of a new general quantitative theory relating the total length of dendritic wiring to the number of branch points and synapses (Cuntz et al. PNAS 2012). This theory is consistent with data from a wide variety of neurons across many different species, and should therefore provide important constraints for our understanding of the design principles governing dendritic structure and function.

5. The first direct patch-clamp recordings from cerebellar Purkinje cell dendrites in vivo, combined with 2-photon calcium imaging, allowing the systematic study of dendritic excitability of Purkinje cell dendrites in the intact brain, as well as to map the spatiotemporal pattern of dendritic calcium signals triggered by sensory synaptic input (Kitamura and Häusser, J. Neurosci. 2011).

6. The first direct measurements from dendrites of pyramidal neurons in visual cortex in vivo, revealing that dendritic spikes that are triggered by visual input contribute to a fundamental cortical computation: enhancing orientation selectivity in the visual cortex. This demonstrates that dendritic excitability is an essential component of behaviourally relevant computations in neurons (Smith et al. Nature 2013).

7. The first patch-clamp recordings from grid cells in medial entorhinal cortex in head-fixed mice navigating in a virtual reality environment, allowing us to probe how synaptic integration of dendritic inputs leads to grid cell firing patterns (Schmidt-Hieber & Häusser, Nature Neuroscience 2013).

8. Introduction of a new approach for reading out the spatiotemporal pattern of inputs to a neuron in vivo by imaging presynaptic axons using 2-photon microscopy (Wilms & Häusser, Nature Communications 2015).

9. Development of a novel technique for ‘all-optical’ interrogation of neural circuits at cellular resolution, which combines two-photon imaging and two-photon optogenetics (Packer et al., Nature Methods 2015).