UNIVERSITY OF HERTFORDSHIRE COMPUTER SCIENCE RESEARCH COLLOQUIUM presents "The dependence of arithmetic operations on input location in cerebellar nucleus and cortical pyramidal neurons" Maria Psarrou Royal Society Wolfson Biocomputation Lab Centre for Computer Science & Informatics University of Hertfordhshire 25 April 2018 (Wednesday) 1 - 2 pm Hatfield, College Lane Campus Seminar Room C408 Everyone is Welcome to Attend Refreshments will be available Abstract: Neurons are constantly bombarded with numerous synaptic signals, which are integrated in order to generate output spikes. A simple way to depict neuronal computations is to plot the relationship between the neuronal input rate and the corresponding output spike rate, that is, the Input - Output relationship (I-O, or transfer function) [1]. A change in the slope or gain of the I-O curve in the presence of different cellular and synaptic mechanisms, such as synaptic noise, shunting inhibition or synaptic plasticity is an indicator of ongoing multiplicative operations [1–4]. Gain modulation is a brain-wide principle of neuronal computation, enabling nonlinear combinations of sensory and cognitive information. An essential component of gain modulation is that a modulatory input alters the sensitivity of the neuron to the original (driving) input, without changing its selectivity [5]. Different nonlinearities in the relationships between input firing rate, excitatory synaptic conductance and output firing rate have been shown to underlie gain modulation [2, 4]. In the present study, we investigate in two different types of neurons whether the dendritic location of excitatory input affects the arithmetic operation performed by different modulatory inhibitory inputs. We used two well described morphologically realistic conductance based models, a cerebellar nucleus (CN) neuron model [6] and a layer Vb pyramidal neuron model [7], and we explore various driving and modulatory input conditions. Modulatory input was provided either by distributed synaptic inhibitory input or a tonic somatic inhibitory conductance. When the driving and modulatory input were both of synaptic nature, we observed a relation between the distance of the excitatory driving input from the soma and the extent of the multiplicative gain change in the CN and the layer Vb pyramidal neurons. In the CN neuron, we found that excitatory inputs underwent additive operations when delivered in somatic and perisomatic areas, and multiplications when delivered to distal dendritic areas, with the extent of multiplication increasing systematically with the distance from the soma. In contrast, in the layer Vb pyramidal neuron excitatory driving input was always multiplied in the dendritic regions, independent of the synapse location. Interestingly, here, the change in gain was smaller than expected based on the distance from the soma. In addition, In all cases where inputs underwent multiplicative operations, the mapping between synaptic excitatory conductance and output firing rate revealed a nonlinearity, with more pronounced nonlinearities due to dendritic saturation in distal synaptic locations corresponding to larger multiplicative gain changes. To show that these non-linear mappings between input conductance and output rate were the basis of the multiplicative gain changes, we drove the two neuronal types with excitatory current injections, at the soma or different dendritic locations, in the presence of modulatory tonic somatic inhibition. In this case, the arithmetic operations performed in all distinct neuronal locations were additive shifts. Moreover, synaptic inhibition had a greater effect on neuronal output than somatic tonic inhibition. Our results indicate that the location and the nature of excitatory inputs affect in a systematic way whether the input undergoes a multiplicative or additive operation. The extent of these operations is also related to the nature of the inhibitory input. Furthermore, different neuronal types might perform different operations when the inputs are received in their perisomatic areas. This is joint work of the speaker Maria Psarrou with Maria Schilstra, Neil Davey, Benjamin Torben-Nielsen, Michael Schmuker, and Volker Steuber. References: 1. Silver RA. Neuronal arithmetic. Nat Rev Neurosci. 2010; 11:474–89. 2. Prescott S a, De Koninck Y. Gain control of firing rate by shunting inhibition: roles of synaptic noise and dendritic saturation. Proc. Natl. Acad. Sci. U. S. A. 2003; 100:2076–81. 3. Chance FS, Abbott LF, Reyes AD. Gain modulation from background synaptic input. Neuron 2002; 35:773–82. 4. Rothman JS, Cathala L, Steuber V, Silver RA. Synaptic depression enables neuronal gain control. Nature 2009; 457:1015–8. 5. Salinas E, Sejnowski TJ. Gain modulation in the central nervous system: where behavior, neurophysiology, and computation meet. Neuroscientist 2001; 7:430–40. 6. Steuber V, Schultheiss NW, Silver RA, De Schutter E, Jaeger D. Determinants of synaptic integration and heterogeneity in rebound firing explored with data-driven models of deep cerebellar nucleus cells. J. Comput. Neurosci. 2011; 30:633–58. 7. Hay E, Hill S, Schürmann F, Markram H, Segev I. Models of neocortical layer 5b pyramidal cells capturing a wide range of dendritic and perisomatic active properties. PLoS Comput. Biol. 2011; 7 --------------------------------------------------- Hertfordshire Computer Science Research Colloquium http://cs-colloq.cs.herts.ac.uk