這篇是最新發(fā)表在Nature Reviews Neuroscience上的文章,，很值得一讀

Activity has an important role in refining synaptic connectivity during development, in part through 'Hebbian' mechanisms such as long-term potentiation and long-term depression. However, Hebbian plasticity is probably insufficient to explain activity-dependent development because it tends to destabilize the activity of neural circuits. How can complex circuits maintain stable activity states in the face of such destabilizing forces? An idea that is emerging from recent work is that average neuronal activity levels are maintained by a set of homeostatic plasticity mechanisms that dynamically adjust synaptic strengths in the correct direction to promote stability. Here we discuss evidence from a number of systems that homeostatic synaptic plasticity is crucial for processes ranging from memory storage to activity-dependent development.

Summary

Neuronal activity often leads to changes in synaptic efficacy. However, such plasticity must be accompanied by homeostatic mechanisms that prevent neural activity from being driven towards runaway activity or quiescence. One potential homeostatic mechanism is the adjustment of synaptic excitability so that firing rates remain relatively constant. At the neuromuscular junction, genetic alterations in synaptic transmission lead to compensatory changes. For example, a decrease in the number of synapses leads to a compensatory increase in quantal amplitude. Such mechanisms might normally adjust neuromuscular transmission during development to allow for changes in muscle growth or synaptic drive. Similar phenomena have been seen in cultured networks of central neurons. Blocking spontaneous activity in cortical cultures results in hyperactivity when the block is lifted. One mechanism for such adjustment is the global regulation of excitatory synapses within a given neuron. Synaptic strength can be measured by analysing miniature excitatory postsynaptic currents (mEPSCs), which result from spontaneous release of quanta of transmitter from individual vesicles. Chronic alterations in activity can increase or decrease the amplitude of mEPSCs. The amplitude seems to be scaled so that each synaptic strength is multiplied or divided by the same factor. Such multiplicative scaling should preserve the relative strengths of synapses. Synaptic strength could be regulated through changes in postsynaptic receptor numbers, presynaptic transmitter release or reuptake, or the number of functional synapses. Evidence in favour of a change in receptor number includes the increase in mEPSC amplitude and in the response to glutamate application. It is unclear whether the homeostatic regulation of receptor numbers shares a signalling pathway with the insertion of receptors into the membrane by long-term potentiation (LTP). Presynaptic changes in transmission are involved in homeostatic plasticity at the neuromuscular junction, but it is less clear whether they are involved in homeostasis in central neurons. In some circumstances, such as developing hippocampal cultures, changes in activity cause changes in the frequency of mEPSCs, as well as in their amplitude, indicating presynaptic alterations. It is unclear how homeostatic plasticity is induced. Important questions include: whether homeostatic plasticity is cell-autonomous; how changes in activity are integrated and read out; and what intracellular signalling cascades generate global changes in synaptic strength. The functioning of cortical networks requires a balance between excitatory and inhibitory inputs onto neurons. Homeostasis in recurrent networks seems to involve adjustments in the relative strengths of excitatory and inhibitory feedback. It seems that excitatory and inhibitory synapses are adjusted independently to maintain activity in the face of changes in drive. Evidence that these mechanisms are important in vivo comes from the developing visual system. For example, during development, there is an inverse relationship between mEPSC frequency and amplitude, indicating that as synaptic drive increases, synaptic strength is reduced.
Original article:

Homeostatic plasticity in the developing nervous system
G. G. TURRIGIANO AND S. B. NELSON   Nature Reviews Neuroscience 5, 97; February 2004