Effect of Adult-Born Immature Granule Cells on Pattern Separation in The Hippocampal Dentate Gyrus

Abstract

Young immature granule cells (imGCs) appear via adult neurogenesis in the hippocampal dentate gyrus (DG). In comparison to mature GCs (mGCs) (born during development), the imGCs exhibit two competing distinct properties such as high excitability and low excitatory innervation. We develop a spiking neural network for the DG, incorporating the imGCs, and investigate their effect on pattern separation (i.e., a process of transforming similar input patterns into less similar output patterns). We first consider the effect of high excitability. The imGCs become very highly active due to their low firing threshold. Then, because of high activation, strong pattern correlation occurs, which results in pattern integration (i.e., making association between events). On the other hand, the mGCs exhibit very sparse firing activity due to strongly increased feedback inhibition (caused by the high activation of the imGCs). As a result of high sparsity, the pattern separation efficacy (PSE) of the mGCs becomes very high. Thus, the whole population of GCs becomes a heterogeneous one, composed of a (major) subpopulation of mGCs (i.e., pattern separators) with very low activation degreeDa(m)and a (minor) subpopulation of imGCs (i.e., pattern integrators) with very high activation degreeDa(im). In the whole heterogeneous population, the overall activation degreeDa(w)of all the GCs is a little reduced in comparison to the activation degreeDa(out)in the presence of only mGCs without imGCs. However, no pattern separation occurs, due to heterogeneous sparsity, in contrast to the usual intuitive thought that sparsity could improve PSE. Next, we consider the effect of low excitatory innervation for the imGCs, counteracting the effect of their high excitability. With decreasing the connection probability of excitatory inputs to the imGCs,Da(im)decreases so rapidly, and their effect becomes weaker. Then, the feedback inhibition to the mGCs is also decreased, leading to increase inDa(m)of the mGCs. Accordingly,Da(w)of the whole GCs also increases. In this case of low excitatory connectivity, the imGCs perform pattern integration. On the other hand, due to increase inDa(m), the PSE of the mGCs decreases from a high value to a limit value. In the whole population of all the GCs, when the excitatory connection probability decreases through a threshold, pattern separation starts, the overall PSE increases and approaches that of the mGCs. However, due to heterogeneity caused by the imGCs, the overall PSE becomes deteriorated, in comparison with that in the presence of only mGCs.

Effect of Adult Neurogenesis on Sparsely Synchronized Rhythms of The Granule Cells in The Hippocampal Dentate Gyrus

“…In this way, the imGC may have a lower firing threshold [52–55], which is well shown in Fig. 2 for the f – I (i.e., firing rate-current) curves of the mGC (red curve) and the imGC (blue curve) in [57].…”

“…Based on the anatomical and the physiological properties described in [16,17,21], we developed the DG spiking neural networks in the works for the winner-take-all competition [45], the SSR [58], and the pattern separation [30]. In our DG spiking neural network for the adult neurogenesis [57], the young adult-born imGCs and the mGCs are incorporated, and more synaptic connections with a high degree of anatomical and physiological realism are also included [60,61].…”

“…The parameter values for the synaptic strength per synapse, the synaptic rising time constant, the synaptic decay time constant, the synaptic latency time constant, and the synaptic reversal potential for the synaptic currents into the GCs, the BCs, the MCs, and the HIPP cells are given in Tables I-III in [57]. These parameter values are also based on the physiological properties of the relevant cells [21,[73][74][75][76][77][78][79][80].…”

“…studied [57]. In our DG network, both high excitability and low excitatory innervation for the imGCs are considered; approximately no inhibition is provided to the imGCs.…”

“…For 0 ≤ x ≤ 1, the mGCs and the whole GCs were found to exhibit pattern separation (i.e., a process of transforming similar input patterns into less similar output patterns), while the imGCs were found to show pattern integration (making association between patterns); efficacy of pattern separation and pattern integration was characterized in terms of pattern separation degree ( X = m and w ) and pattern integration degree ( X = im ), respectively [30, 57]. Here, quantitative relationship between and of the SSRs and and of pattern separation and integration is also discussed.…”