Reduced learning rate and E/I imbalance drive Peripersonal Space boundaries expansion in Schizophrenia
Abnormal encoding of peripersonal space (PPS) is believed to affect bodily self disruptions in schizophrenia (SCZ). Empirical studies show that SCZ patients exhibit a narrower PPS than controls but maintain its plasticity. Computational research links this smaller PPS to increased excitation of sensory neurons and reduced feedforward synaptic density. However, it is unclear how such differences influence learning during the expansion of PPS boundaries. We hypothesise that Hebbian plasticity can account for PPS expansion after active tool use training. To explore the effect of such mechanisms on PPS plasticity, we developed a SCZ network model which was fit to behavioural data before and after tool manipulation. We found that PPS expansion occurs in spite of E/I imbalance or reduced synaptic density, but does not match the post-training PPS representation of patients. A better fit was obtained after altering plasticity by either reducing the learning rate, increasing the forgetting rate or increasing the plasticity threshold. We discuss our findings in terms of dysfunctional plasticity in SCZ and highlight the key challenges in identifying the neurobiological correlates of reduced plasticity within PPS networks. Because current empirical data supports multiple viable mechanisms, we propose experiments to distinguish between the proposed plasticity accounts and clarify mixed findings on PPS representation in SCZ. Graphical Abstract O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=123 SRC="FIGDIR/small/604515v2_ufig1.gif" ALT="Figure 1"> View larger version (34K): org.highwire.dtl.DTLVardef@14ee71forg.highwire.dtl.DTLVardef@127c3aborg.highwire.dtl.DTLVardef@19f1338org.highwire.dtl.DTLVardef@a6e437_HPS_FORMAT_FIGEXP M_FIG C_FIG HighlightsO_LIUsing a peripersonal space network model, we found that PPS expansion after tool-use occurs in spite of E/I imbalance or reduced synaptic density. C_LIO_LIReduced feedforward synaptic plasticity is required to match the post-training PPS representation of patients. C_LIO_LISuch reduction of synaptic plasticity could be achieved by either reducing the learning rate, increasing the forgetting rate or increasing the plasticity threshold relative to a healthy control model. C_LIO_LIOur model predicts that measuring PPS at intermediate time points during a longer stimulation protocol would help distinguish between these plasticity differences. C_LI