neuromatch 40 - Learning accurate path integration in ring attractors of the head direction system

Link to preprint: https://www.biorxiv.org/content/10.11...

Summary:
Head direction cells track an animal’s head direction in darkness by integrating angular velocity signals, a phenomenon called path integration [1,2]. Ring attractor models for angular path integration have recently received strong experimental support [3,4,5]. To function as integrators, head direction circuits require precisely tuned connectivity, which is costly to pass down genetically [6]. This suggests that synaptic plasticity is crucial in setting up these circuits.

We propose a network model in which a local, biologically plausible learning rule adjusts synaptic efficacies during development, guided by supervisory allothetic cues. The learning rule is inspired by layer-5 pyramidal neurons assumed to be the fundamental associative unit in the cortex, where backpropagating action potentials implement coincidence detection [7,8]. The learning rule contains an anti-Hebbian component which performs predictive coding, whereby inputs arriving in distinct compartments get associated so that inputs in one compartment can predict inputs to another. Applied to the Drosophila head direction system, where such a segregation of inputs exists [9], the model learns to path-integrate accurately for the full range of angular velocities that the fly displays, and develops a connectivity strikingly similar to the one reported in experiments [5]. The mature network is a quasi-continuous attractor (CAN), and reproduces key experiments in which optogenetic stimulation controls the internal representation of heading [4], and where the network quickly remaps to integrate with different gains akin to experiments conducted in augmented reality in rodents [10].

Our model and proposes a general framework to learn gain-1 path integration, even in architectures that lack the physical topography of a ring, like the head direction system in mammals [11]. Finally, we develop an analytically tractable reduced model that exploits symmetries present in CANs, explains how the full network solves credit assignment, and offers a rigorous mathematical framework to study the self-organization of CANs for path integration in general.


References:
1. J S Taube, R U Muller, and J B Ranck. Head-direction cells recorded from the postsubiculum in freely moving rats. i. description and quantitative analysis. The Journal of Neuroscience, 10(2):420435, February 1990, 10.1523/JNEUROSCI.10-02-00420.1990
2. A S Etienne, R Maurer, and V Séguinot. Path integration in mammals and its interaction with visual landmarks. The Journal of experimental biology, 199(1):201209, Jan 1996
3. J D Seelig and V Jayaraman. Neural dynamics for landmark orientation and angular path integration. Nature, 521(7551):186191, May 2015, 10.1038/nature14446
4. S S Kim, H Rouault, S Druckmann, V Jayaraman. Ring attractor dynamics in the Drosophila central brain. Science. 3892017 May; 356(6340):849–853, 10.1126/science.aal4835
5. D B Turner-Evans, K T Jensen, S Ali, T Paterson, A Sheridan, R P Ray, T Wolff, J S Lauritzen, G M Rubin, D D Bock, V Jayaraman. The neuroanatomical ultrastructure and function of a biological ring attractor. Neuron, 108(1):145–163.e10, 2020, 10.1016/j.neuron.2020.08.006
6. R H R Hahnloser. Emergence of neural integration in the head-direction system by visual supervision. Neuroscience, 120(3):877891, sep 2003, 10.1016/S0306-4522(03)00201-X
7. M Larkum. A cellular mechanism for cortical associations: an organizing principle for the cerebral cortex. Trends in Neurosciences, 36(3):141–151, mar 2013, 10.1016/j.tins.2012.11.006
8. R Urbanczik, W Senn. Learning by the dendritic prediction of somatic spiking. Neuron, 81(3):521–528, feb 2014, 10.1016/j.neuron.2013.11.030
9. C S Xu and et al. A connectome and analysis of the adult Drosophila central brain. bioRxiv, January 2020, 10.7554/eLife.57443
10. R P Jayakumar and et al. Recalibration of path integration in hippocampal place cells. Nature, 566(7745):533–537, February 2019, 10.1038/s41586-019-0939-3
11. R Chaudhuri, B Gerçek, B Pandey et al. The intrinsic attractor manifold and population dynamics of a canonical cognitive circuit across waking and sleep. Nat Neurosci 22, 1512–1520 (2019), 10.1038/s41593-019-0460-x