Effect of motor control on sensory coding in the awake, behaving mouse

Our knowledge of the neural mechanisms of sensation comes mainly from a passive experimental paradigm where sensory stimuli are presented to the immobilised sense organ of an anaesthetised animal (e.g., spot of light projected on the retina). However, most species actively seek information by coordinating motor control of their sense organs (e.g., eyes, hands, whiskers) with sensory processing ('active sensing'). Indeed, the activity of sensory receptors can be strongly affected by sense organ movements: if eye movement is abolished, the visual image fades away. The concern that motor control might have an important impact on sensory coding was difficult to address fully in the past due to technical limitations.

Analogously to the way that humans use motor control to explore objects by movement of our hands and eyes, rodents explore their environment by rhythmically sweeping their whiskers backwards and forwards ('whisking'). The whisker sense is extremely important to these animals: it dominates the brain's somatosensory areas. The whisker sense of rats/mice offers an excellent model system to get insight into the central question of this proposal - how sensory mechanisms depend on motor control. How the whiskers move and interact with objects can now be precisely measured using high-speed video cameras. Moreover, since engineers have long been interested in how rod-like objects bend, there is a well-developed mathematical theory that can be used to estimate the mechanical forces of whisker-object interaction from video data.

The way that we perceive the world is fundamentally constrained by primary sensory neurons, which translate signals in the environment (such as bending of a whisker) into corresponding patterns of action potentials. Here, we propose to determine the tactile features that primary whisker neurons encode as a mouse actively explores objects with its whiskers. The significance of the work is that it will provide insight into the general question of how motor control (here, whisking) and sensory processing interact to constrain how we perceive the world.

To determine how sensory coding interacts with motor control, it is valuable to record the activity of primary sensory neurons in awake animals, whilst at the same time measuring the input to the (moving) sense organ. The Petersen lab recently published the first study to do this in the whisker system (Campagner et al, 2016; Elife).

To address objectives O1-O3, we will train head-fixed mice to perform a go/no-go object localisation task. The object (textured pole) will be positioned in either go or no-go location. The mouse will be rewarded for licking on go trials and punished for licking on no-go trials. High speed cameras will be used to measure both whisker movement and whisker-pole interaction. Simultaneously, we will use microelectrode arrays implanted in the trigeminal ganglion to record the activity of primary whisker neurons. Computer vision algorithms, developed by the PI, will be used to extract tactile features (e.g., whisker angle/curvature, stick-slip events) from the imaging data. Machine-learning algorithms (Generalised Linear Models) will then be used to assess how well each feature predicts neuronal activity.

To measure 3D whisker movements and 3D whisker-object interactions (O2), we will use two, calibrated high speed cameras (horizontal and coronal views). A 3D computer vision algorithm will be used to extract 3D tactile features (e.g., whisker curvature in horizontal/coronal planes).

To selectively record from PWNs that innervate Merkel cells (O3), we will use an optogenetic strategy. We have identified two lines of Cre mice (CCK-Cre and K14-Cre) where, within the whisker follicle, Cre recombinase is selectively expressed in Merkel cells. In preliminary work, we have crossed these mice with Ai32 mice that express ChannelRhodopsin-2 (ChR2) in a Cre-dependent manner, to obtain mice that express ChR2 in Merkel cells. We will identify these Merkel-PWNs by testing whether they respond to transcutaneous delivery of blue (473 nm) light.

This is a basic research proposal. In the short term, the expected outcome from this research is insight into the mechanisms of sensation in peripheral sensory nerves and, in particular, into mechanisms of sensation in the whisker system. In the long term, these advances have potential impact on (1) diseases of the peripheral nerves and (2) tactile sensory systems for robotics. Here, and in the pathways to impact document, we map out a pathway to translate outcomes from this project into impact in these two areas.

(1) Disease of peripheral nerve. In order to realise this impact, we are focussing on a major peripheral nerve disease - diabetic neuropathy. The present project can achieve impact through application of the sophisticated methods we are developing for studying sensory coding in the whisker system to animal models of diabetes. My vision is that application of our methods will lead to a better understanding of disease mechanisms, and thereby facilitate the development of improved therapies, solidly rooted in basic science. To this end, RP's group started a collaboration on diabetic neuropathy, with Manchester diabetes expert, Dr. Natalie Gardiner, to study the effect of streptozotocin-induced diabetes on whisker neuron function. An significant impact milestone was recently achieved in that we recently published the first results of our collaboration (Freeman et al, 2016). The next step is to apply for translational funding to pursue our collaboration, by applying state of the art behavioural monitoring and awake electrophysiological recording methods, developed in the present proposal, to the study of nerve function in diabetic rodents.

(2) Tactile sensory systems for robots. There is substantial commercial interest in developing robots with human-like abilities to manipulate objects through touch, with tactile sensors being developed, for example by SynTouch and Tekscan. The potential applications include surgical robots and prosthetic hands. However, current devices are incredibly crude compared to primate hands or rodent whiskers. The impact potential here is that insight into the mechanisms of active touch could stimulate the development of more sophisticated and more effective technology robotic devices. My vision is that improved understanding of mechanisms of active touch from the present project will impact first on companies developing artificial touch technology, and, as those companies' products improve, consequently on a wide range of markets. I have existing links with the UK academic roboticists who have pioneered tactile robots, including whisker-driven ones (Sheffield, Bristol), but this has yet to translate into genuine collaboration. The next step on the pathway to impact is to network in the most effective possible way with roboticist researchers and robotics companies and to develop collaborations.

Due to its multi-disciplinary character, involving not only in vivo experiment but also computational modelling, the proposed project offers exceptional training potential for the PDRA and technician.

References
OJ Freeman, MH Evans, GJS Cooper, RS Petersen, NJ Gardiner (2016) Thalamic amplification of sensory input in experimental diabetes. European Journal of Neuroscience. DOI 10.1111/ejn.13267.