Potassium channels in the auditory brainstem

A series of projects examined how different voltage-gated K+ channels enable a fast inverting relay neuron in the auditory brainstem to fulfil its specialised role in sound source localisation. Key papers from this work are here and here. The image to the left shows Kv3.1 staining in the MNTB, note the puncta in the axons where the nodes of ranvier are labelled.
Screen Shot 2018-05-05 at 10.38.45

Potassium channel physiology in the MNTB

Olfactory bulb microcircuit

This project explored the intrinsic properties of olfactory bulb output neurons. We showed that input from the olfactory nerve activates Ca2+channels located in the primary dendrites and these contribute to dendritic glutamate release. The resulting depolarisation and boosted glutamate release synchronises the activity of multiple output neurons belonging to the same glomerulus. The papers can be found here and here


Rapid mapping of receptive fields


Spatial and temporal receptive field of a retinal ganglion cell measured using FBP

And its application to multi-neuronal electrophysiology and imaging

Neurons in sensory systems have receptive fields. In the somatosensory system, this might be an area of the skin to which the neuron is responsive, whereas in the visual system the receptive field is an area on the retina. Several properties of the receptive field are of interest. The spatial extent of the receptive field, e.g. the area on the retina to which the neuron is responsive, and the temporal characteristics, e.g. when will the neuron respond after applying a stimulus. Measuring the properties of a neuron’s receptive field is an important step towards understanding its function.



The spatiotemporal receptive field of an OFF retinal ganglion cell. The square in the bottom left corner indicates the time of stimulation. Field of view is 900 µm x900 µm. Data is played at 1/20th real time 


This project developed a new method that allows the rapid mapping of receptive fields from multiple neurons simultaneously. The procedure is illustrated to the right and the full details are available in the paper hereA series of bars are flashed at different locations across the retina and this is repeated with the bars rotated to cover at least 5 evenly spaced angles. The receptive fields can then be easily recovered using an algorithm applied in CAT scans, the filtered back projection. The code to implement this method is available here.
We showed that the FBP method can recover both the spatial and temporal components of the receptive field. The FBP method had several advantages over the frequently used “spike-triggered average”  approach. The FBP could recover receptive fields significantly faster, with higher signal to noise and the resolution of the temporal impulse response was superior.

Summary of the method

We also demonstrated how this method is suitable for functional imaging of neural activity. Imaging data is often noisy and clear unitary events such as spikes are not readily discernible. As our method does not rely on detection of events it readily lends itself to mapping receptive fields of neurons measured in imaging experiments. We demonstrated this by mapping the receptive fields of an array of retinal bipolar cell synapses expressing SyGCaMP6 shown below.

Motion Anticipation in the retina

General Features of the retinal connectome determine the computation of motion anticipation


A Retinal Ganglion Cell filled with Alexa 488 to recover its morphology

Light is converted into electrical signals by specialized cells in the retina called photoreceptors. This conversion process, termed phototransduction, is relatively slow, taking around a tenth of a second. Although this might not sound like a long time, it is enough for a tennis ball to have travelled 3-4 meters when served by a professional.
Despite this, we do not experience moving objects being delayed, which indicates that moving objects are somehow processed more rapidly. An example of how moving objects are processed differently to static objects can be seen in the flash-lag illusion. This apparent faster processing of moving objects begins in the eye.
Screen Shot 2016-05-11 at 08.14.52

Selective disruption of inhibition in a single retinal ganglion cell

Retinal ganglion cells (pictured), which send signals from the retina to the rest of the brain, respond earlier than expected for moving objects, overcoming the slow process of phototransduction. This phenomenon has been termed “motion anticipation”. This project revealed how motion anticipation arises from the circuitry of the retina.
Using a combination of electrical recordings and computer simulations, we showed that inhibitory signals play a key role in motion anticipation. In particular, we found that an excess of inhibitory inputs onto retinal ganglion cells enables motion anticipation for objects moving in any direction across the retina. 
Screen Shot 2016-05-11 at 08.17.56

Computational model showing the effect of inhibition for a moving stimulus

The project described how the non-linear interaction of excitatory and inhibitory synapses, within ganglion cell dendrites, enables encoding of the actual position of a moving object instead of its delayed representation. The fact that retinal ganglion cells receive an excess of inhibitory inputs is a long-observed feature of retinal wiring, and a role for this can now be understood in terms of providing the mechanism of motion anticipation. The full paper can be read here

Orientation & Dynamic Predictive Coding in the Retina

Sensory systems must reduce the transmission of redundant information to function efficiently. One strategy is to continuously adjust the sensitivity of neurons to suppress responses to common features of the input while enhancing responses to new ones.

In this project we used 2-photon imaging of the synaptic transmitter glutamate. Taking advantage of the genetically encoded fluoresecent reporter of glutamate iGluSnFR we were able to image the synaptic inputs impinging upon retinal ganglion cells as well as the output of the same ganglion cells at their axon terminals.


We show that the synaptic terminals of retinal bipolar cells compute the orientation of edges and that the retina signals the orientation of edges by two distinct means. One population of retinal ganglion cells are statically tuned, receiving inputs from retinal bipolar cells that are all tuned to a single orientation. A second population of retinal ganglion cells receive a mix of inputs from retinal bipolar cells tuned to different orientations. These ganglion cells are able to respond transiently to a change in orientation yet mainatin sensitivity to any further change in orientation, a computation termed dynamic predictive coding.

We also revealed key roles for inhibitory neurons within this circuit. Lateral inhibition is required for the computation of orientation by retinal bipolar cell synapses (see diagram). Whereas feedforward inhibition is required to generate a high-pass filter that enables ganglion cells carrying a dynamic predictive code to only transmits the initial activation of their different  retinal bipolar cell inputs, thus removing redundancy. These results demonstrate how a dynamic predictive code can be implemented by circuit motifs common to many parts of the brain. The full paper is available here.


A Retinal Circuit Generating a Dynamic Predictive Code for Oriented Features