STDP Underlined Regulation Mechanisms Of Structure And Firing Activity Of Cortical Neural Network

Cerebral cortex networks with excitatory and inhibitory synaptic plasticity are the physiological basis for many functions.It can contribute to our understanding about the development and operations of neural circuit identifying how these two kinds of synaptic plasticity jointly regulate neural networks.Spike-timing-dependent plasticity(STDP),as a ubiquitous synaptic plasticity,regulates neural activity by changing the structure of neural networks through rewiring synaptic weights.Previous studies have shown that excitatory STDP(E-STDP)can result in continuous growth in connection weights,thus cause abnormal excitability,whereas inhibitory STDP(I-STDP)can maintain firing activity stability.Combining E-STDP and I-STDP can improve robustness of the network to stimulation and enhance neural information.The study of synchronous rhythm in cortical networks has been a focus in the field of neuroscience for more than half a century.Non-periodic synchronization(NPS),as a type of synchronous rhythm,is closely related to sensory information processing,memory formation,and other cognitive activities.Recognizing the origins of NPS enhances our understanding of the various neural states and the characteristics of neural information processing.Previous studies have shown that the generating mechanism of NPS is related to the heterogeneity of neurons,inhibitory synapses and topological structure of the neural networks.Moreover,a few computational studies have indicated that E-STDP plays a key role in transition between synchronization and desynchronization in feedforward neural networks.However,it is unclear whether the interaction between E-STDP and I-STDP can generate the non-periodic synchronization.It remains to be investigated how the combination of E-STDP and ISTDP changes the structure of the neural network such that the population synchronously and randomly fires.Here,by constructing a homogeneous cortical neural network with excitatory and inhibitory STDP,we studied the regulation of the two kinds of STDP on the evolution of neural network structure and non-periodic synchronization,thus probed generating mechanism of the NPS.Combining computational neuroscience with complex network methods,we have conducted the following researches:(1)The regulation of STDP on the structure of neural network.First,we showed the evolution process of synaptic connection weight matrix.Then,we presented the evolution of degree distribution of the neural network,and focused on the quantitative analysis of the in-degree and out-degree of excitatory neurons.Finally,we demonstrated the formation of modular structure during the simulation,and further explored the effects of E-STDP and I-STDP on the modularized structure of homogenous neural network using the cluster analysis.Our results indicate that ESTDP rapidly increases E-E synaptic weights,causing the input portion of postsynaptic neurons to become stronger than that of other neurons,furthermore,the population assembles intensively and simultaneously.I-STDP increases I-E synaptic weights following E-STDP,maintaining the balance of excitation and inhibition input at the level of both neuron and population,and contributing significantly to the cluster features of each neuron.Our results also showed that the population forms a stable modular neural network under the interaction of E-STDP and I-STDP.(2)The generation of non-periodic synchronization in neural network.First,we analyzed the firing activity characteristics formed by the interaction of E-STDP and ISTDP,which is described by several indicators such as membrane potential,firing rate,coefficient of variation of inter-spike intervals,coherence index of synchronization,respectively.Subsequently we presented the NPS phenomenon in the late stage of simulation in the following aspects,including non-periodic verification,excitatoryinhibitory balance of neural network,the activity state and frequency features of the network.The results show that the NPS of neural network reflects a competition between excitatory and inhibitory feedbacks,and the balanced degree varies dynamically.Meanwhile,the network activity trajectory fluctuates obviously,but can be described by fewer dimensions which is beneficial to neural information processing.The power spectrum of firing rate exhibits the power law distribution that further demonstrates non-periodic characteristics of the neural activities,and the synchronization occurs with higher energy in a wider frequency range.(3)The generating mechanism of non-periodic synchronization underlined by the interaction of excitatory and inhibitory STDP.First,we demonstrated that NPS cannot be generated in the homogenous structure network with the same total synaptic connection strength as that of a heterogeneous network.Next,we qualitatively analyzed the effects of the balance degree of excitation and inhibition on the generation of NPS.At last,we evaluated the impact of single neuron on the burst numbers of the network.The results show that the generation of NPS is critically dependent on the competitive interaction between E-STDP and I-STDP,and is reflected in the following aspects: 1)the heterogeneous structure plays a key role in generating NPS at the condition of equal average synaptic connection strength of the network;2)the excitatory and inhibitory feedbacks of neural network are in competition with each other and achieve tight dynamic balance;3)the different spiking of single neuron exerts impact on the firing activity of the network.To sum up,the present study shows that the interaction of E-STDP and I-STDP drives the evolution of neural network structure from homogenous to heterogeneous network.Furthermore,the non-periodic synchronous firing activity is generated at steady state that is promote neural information processing.The present study elucidates the regulation mechanisms of structure and firing activity of neural network underlined by the interaction of excitatory and inhibitory STDP,which is helpful for comprehending the development and operation mechanisms of the brain and provides some testable predictions for further physiological research.