A model of dendritic spine Ca 2 + concentration exploring possible bases for a sliding synaptic modification threshold ( N-methyl-D-aspartate receptor / synaptic plasticity / long-term potentiation )
J. Gold,Mark F. BEARt
Abstract:We used a biophysical model of an isolated dendritic spine to assess quantitatively the impact of changes in spine geometry, Ca2+ buffer concentration, and channel kinetics on Ca2+ dynamics following high-frequency activation of N-methyl-D-aspartate receptors. We found that varying the buffer concentration in the postsynaptic density from 50 to 500 ,uM can result in an 8-fold difference in the peak Ca2+ concentration following three pulses -at 100 Hz. Similarly, varying the spine neck diameter from 0.1 to 0.55 jAm can result in a 15-fold difference in the peak Ca2+ concentration. The amplification of peak Ca2+ concentration also depended on temporal summation of N-methyl-D-aspartate-mediated excitatory postsynaptic currents. Variation of the current duration on the order of 100 msec can significanly affect summation at a given stimulation frequency, resulting in a 10-fold difference in the peak Ca2+ concentration at 100 Hz. It is suggested that activity-dependent modifications of these parameters may be important for the regulation of synaptic plasticity in the brain. "Hebbian" synaptic modification, in which an excitatory connection is strengthened whenever preand postsynaptic elements are synchronously active, has been shown by computational models to be useful to account for aspects of learning, memory, and neural development (for reviews, see refs. 1-3). Success of these models usually requires, in addition, mechanisms to constrain synaptic weights so that the network of modifiable synapses reaches a stable equilibrium. Thus, two questions of neurobiological interest emerge from this work: (i) is there a neurobiological basis for Hebbian synaptic modification and (ii) how are Hebbian synaptic modifications regulated? The answer to the first question is clearly yes. Long-term potentiation (LTP), a form of synaptic plasticity first described in the hippocampus (4, 5), appears to have the characteristics expected of a Hebbian modification (for reviews, see refs. 6 and 7). Induction of LTP in area CA1 of hippocampus (8-10), in dentate gyrus (11), and in some regions of neocortex (12-14) occurs following simultaneous presynaptic input activity and strong postsynaptic depolarization. There are a number oflines ofevidence that this form of LTP in hippocampus is triggered by a brief surge in intracellular Ca2+ during a tetanus (15-19). In addition, activation of N-methyl-D-aspartate (NMDA) receptors is a requisite step for the induction of this form of plasticity (20). Thus, the discovery that a substantial, voltage-dependent Ca2+ conductance is activated by agonists acting at the NMDA channel paired with significant postsynaptic activity (21, 22) appeared to account for all the properties of Hebbian LTP (23). Recently, however, this simple model has been brought into question. Two separate theoretical analyses of the Ca2+ dynamics that might result from NMDA receptor activation on dendritic spines have led to the same conclusion; namely, the "nonlinear" properties of the NMDA receptor-mediated Ca2+ flux are not sufficient to account for the "nonlinear" stimulation requirements for induction of LTP (24, 25). Both groups suggested that an additional nonlinear factor, such as a Ca2+ diffusion barrier at the spine neck or the existence of a saturable Ca2+ buffer in the spine head, could account for the properties of LTP (making the critical assumption, of course, that the relation between the Ca2+ concentration and the amount or probability of LTP is monotonic). We were struck by the possibility that regulation of these same parameters might offer an answer to the second question posed above, concerning how LTP is regulated. The theoretical underpinnings for this supposition are provided by the work of Cooper and colleagues (26-30). This work has provided an analysis of a theoretical form of synaptic plasticity in which the occurrence of Hebbian modification is contingent upon the activation of the postsynaptic neuron beyond a critical value termed the modification threshold. The key property that imparts stability in these models is the proposition that the value of the modification threshold (set for the entire cell) is some function of recent postsynaptic activity. Based on the models of Ca2+ dynamics in spines reviewed above, the proposal was made that regulation of expression of Ca2+-binding proteins as a function of cell activity might be a physical basis for such a "sliding" modification threshold (31). The aim of the present work was to use a computational model of Ca2+ dynamics in dendritic spines to examine the feasibility of such a proposal. We elaborate on previous models to explore in more detail how parameters such as buffers, spine geometry, and the time course of NMDA receptor-mediated excitatory postsynaptic currents (EPSCs) affect the rise in Ca2+ concentration that might accompany a tetanus. The results show that as little as a 2-fold change in spine neck diameter, buffer concentration, and/or NMDAmediated EPSC decay time will substantially alter the peak rise in Ca2+ and may therefore significantly affect the ability to induce LTP. MATERIALS AND METHODS The simulations performed were based on the idealized gedankenexperiment presented in Fig. 1A and concerned the dynamics of Ca2+ following tetanic stimulation of a single axospinous synapse. The experiment is highly idealized: stimulation is at a single synapse which contains a full complement of NMDA receptors; NMDA receptor activaAbbreviations: EPSC, excitatory postsynaptic current; LTP, longterm potentiation; NMDA, N-methyl-D-aspartate; Vm, membrane potential. *Present address: Department ofNeurobiology, Stanford University Medical Center, Stanford University, Stanford, CA 94305-5401. tTo whom reprint requests should be addressed. 3941 The publication costs of this article were defrayed in part by page charge payment. This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. §1734 solely to indicate this fact. 3942 Neurobiology: Gold and Bear