Metaplasticity refers to activity-dependent changes in neural functions that modulate subsequent synaptic plasticity such as long-term potentiation (LTP) and long-term depression (LTD). Simply put, it is the “plasticity of synaptic plasticity” (Abraham and Bear, 1996). Metaplasticity can be distinguished from conventional neuromodulation of plasticity, in which molecules such as other neurotransmitters (e.g., GABA or monoamines), cytokines, or hormones that are present at the time of plasticity induction regulate the degree of LTP or LTD elicited (Fig. 1). Metaplasticity, in contrast, refers to neuronal changes that are elicited at one point in time, by what is commonly called “priming” activity. By virtue of their persistence, these neuronal changes are able to regulate synaptic plasticity processes minutes, hours, or days later. At very short intervals between priming activity and plasticity induction, however, the distinction between metaplasticity and neuromodulation may become blurred. In many cases, metaplasticity is induced without producing observable changes in synaptic transmission, and it only becomes apparent during subsequent attempts to induce LTP or LTD. In practice, however, neural activity can generate metaplasticity concomitantly with synaptic plasticity.
Functionally, metaplasticity endows synapses with the capacity to integrate plasticity-relevant signals across time. It can also serve, by changing the thresholds for LTP and LTD, to keep synaptic strengths within a functional dynamic range. That is, it prevents them from becoming so strong or weak that they lead to excessive or insufficient activation, respectively, of the postsynaptic cells. Readers are directed to a recent review for a more comprehensive discussion of metaplasticity (Abraham, 2008).
Many types of metaplasticity occur homosynaptically, in that the effects of priming synaptic activity on plasticity mechanisms are confined to the primed synapses. At glutamatergic synapses, the induction of homosynaptic metaplasticity can be distinguished as being either N-methyl-D-aspartate (NMDA) receptor-dependent or metabotropic glutamate receptor (mGluR)-dependent.
NMDA receptor-activated metaplasticity
One metaplasticity consequence of activating NMDA receptors (NMDARs) is a homosynaptic facilitation of subsequent LTD (Christie and Abraham, 1992). This metaplasticity effect can be induced by relatively brief trains of low-frequency stimulation and lasts up to 90 min. The facilitation occurs sufficiently quickly that it contributes to the conventional induction of LTD by a long train of low-frequency stimulation (LFS) (Mockett et al., 2002). Thus, during LFS-induced LTD, pulses early in the stimulus train prime the development of LTD as induced by later pulses in the stimulus train.
In concert with the facilitation of LTD, NMDAR activation can homosynaptically inhibit subsequent LTP (Huang et al., 1992). This priming effect can be elicited by either LFS or high-frequency stimulation (HFS) and lasts up to 90 min. Generally, it is the induction of LTP that is impaired, but when strong tetanus protocols are used to overcome effects on LTP induction, metaplasticity may also manifest as a reduction in LTP persistence (Woo and Nguyen, 2002). The metaplastic inhibition of LTP by NMDAR activation has been shown to contribute to apparent “saturation” of LTP by repeated induction episodes. Experimentally, saturation is defined as the state when additional HFS trains produce no further LTP. However, in principle it is ambiguous whether this state represents a true saturation of potentiation mechanisms, or whether a metaplasticity effect is preventing further LTP induction.
The metaplastic inhibition of LTP by priming activation of NMDARs is mediated by activation of adenosine A2 receptors, p38 MAP kinase, and the protein phosphatases 1/2A and calcineurin (cf. Abraham, 2008). Feedback down-regulation of NMDAR receptor function itself, mediated by nitric oxide (NO) and protein kinase C (PKC), has also been postulated to contribute to the altered thresholds for LTD and LTP, but a direct link of this mechanism to metaplasticity has not yet been made.
In direct contrast to the NMDAR-mediated inhibition of LTP, activation of Group I metabotropic glutamate receptors (mGluRs) can facilitate subsequent LTP. This is manifest as an increase in both the induction and persistence of LTP (Cohen and Abraham, 1996), or as the removal of dependence of LTP on mGluR activation at the time of its induction (an effect described as the setting of a “molecular switch” (Bortolotto et al., 1994)). Increased induction is at least partly due to a down-regulation of the slow-afterhyperpolarization (sAHP) that is triggered by action potential firing, and that is a potent brake on LTP induction (Cohen et al., 1999). Inhibition of the sAHP is not mediated by the classical Gq signaling cascade involving phospholipase C activation, but involves a novel signaling pathway that is gated by tyrosine phosphorylation of one or more regulatory proteins. The metaplastic facilitation of LTP persistence by Group I mGluR activation is mediated by local synaptodendritic de novo synthesis of proteins, converting a rapidly decaying LTP into a more stable form (Raymond et al., 2000). While most mGluR-mediated metaplasticity has been ascribed to Group I mGluR activation, other mGluR subtypes can also generate effects. For example, activation of Group 2 mGluRs, which are negatively coupled to adenylate cyclase, can contribute to subsequent inhibition of either LTP or LTD (cf. Abraham, 2008).
Under some conditions, typically involving multiple HFS episodes, synaptic activation can lead to heterosynaptic metaplasticity that affects not only the activated synapses but also neighboring non-activated synapses (Roth-Alpermann et al., 2006; Wang and Wagner, 1999). Under these conditions, LTP is inhibited and LTD is facilitated. In CA1 slices, such effects can last in Schaffer collateral synapses for up to 90 min, but in the dentate gyrus in vivo, the effect in perforant path synapses can last for many days (Abraham et al., 2001). The latter effect is blocked by NMDAR but not protein synthesis antagonists. Postsynaptic cell firing may be sufficient to generate the change in plasticity thresholds, as postulated by the Bienenstock, Cooper and Munro (Bienenstock et al., 1982) computational model of synaptic plasticity (BCM theory), depending on the model used. Another kind of heterosynaptic interaction affecting LTP and LTD entails the de novo synthesis of proteins by activation of one set of synapses, which are then captured following weak activation of a second set of synapses to promote the persistence of LTP or LTD in the second pathway. CaMKII contributes to the capturing process through setting a synaptic “tag” (Frey and Morris, 1997), while PKMζ is likely to be one of the key newly synthesized and captured proteins that contributes to the enhanced persistence of LTP. Prior stimulation of inputs extrinsic to the hippocampal formation can also metaplastically regulate LTP in hippocampal perforant path synapses. For example, activation of the basolateral amygdala can either facilitate or inhibit subsequent dentate LTP depending on the timing interval (Akirav and Richter-Levin, 2002). Mechanisms contributing to LTP facilitation include beta-adrenergic receptor activation and protein synthesis. Several other mechanisms can contribute to heterosynaptic metaplasticity. First, the activity of a number of plasma membrane ion channels can be altered, in a process termed intrinsic plasticity. For example, both increases and decreases in the potassium channel mediating the slow afterhyperpolarization have been linked to facilitation and inhibition of LTP, respectively. A-type potassium channels are also regulated by synaptic activity and serve as potent modulators of LTP induction (Wang et al., 2003; Chen et al., 2006). Second, activity-dependent regulation of GABAergic synaptic inhibition can serve metaplasticity functions. Notably, synaptic activation of postsynaptic neurons, via Group I mGluRs and postsynaptic cell firing, can lead to retrograde release of endocannabinoids that bind to nearby GABAergic synaptic terminals and inhibit their function for either less than a minute or tens of minutes, depending on the nature of the postsynaptic activation (Carlson et al., 2002; Chevaleyre and Castillo, 2004). In either case, LTP induction is facilitated in small local bands of synapses for a period of time corresponding to the reduced GABAergic transmission.
BEHAVIORAL RELEVANCE OF METAPLASTICITY
Metaplasticity in the developing visual cortex
The sensory environment can naturally induce metaplasticity, as exemplified by studies in the visual cortex. In the visual cortex, as in other sensory cortices, neurons progressively gain selective receptive field properties through experience-dependent plasticity during development. The BCM theory accounts for the observed developmental and use-dependent acquisition of stimulus selective properties of neurons in primary sensory cortices (Bienenstock et al., 1982). The two main tenets of this theory hold that (1) synapses are bidirectionally modifiable (e.g. capable of both LTP and LTD) and (2) the ability to modify the strength of synapses is itself modifiable (i.e., metaplasticity). The BCM theory predicted that when neuronal activity levels are low, as with visual deprivation, the LTD/LTP plasticity threshold will be lowered such that incoming activity is more likely to generate synaptic strengthening (LTP) and less likely to cause synaptic weakening (LTD). In support of this hypothesis, dark-rearing enhances the ability to strengthen layer 4 to layer 2/3 visual cortex synapses, and reduces the ability to weaken these synapses, across a range of stimulation protocols (Kirkwood et al., 1996). A change in the molecular composition of NMDA receptors appears to underlie these experience-dependent modifications in the LTD/LTP plasticity threshold. The NR1, NR2A, and NR2B NMDAR subunits predominate throughout visual cortex postnatal development. There is a progressive increase in the ratio of NR2A/NR2B subunits that is prevented by dark-rearing (Quinlan et al., 1999). Thus, a low NR2A/NR2B ratio is associated with an enhanced ability to induce LTP. The causal relationship between NMDA receptor subunit composition and experience-dependent modifications in the plasticity threshold was demonstrated in studies showing that mice lacking the NR2A subunit also lack experience-dependent metaplasticity (Philpot et al., 2007). As predicted by the BCM theory, stimulus selectivity (e.g. orientation selectivity) is attenuated in the visual cortex when metaplasticity is prevented by deletion of NR2A (Fagiolini et al., 2003). Changes in the NR2A/NR2B ratio of NMDAR subunits may be a common mechanism in the brain for altering the plasticity threshold, as sleep deprivation and olfactory learning are both manipulations that increase the induction threshold for LTP and raise the ratio of NR2A to NR2B subunits.
Experience-dependent induction of metaplasticity
In addition to the well-documented developmental and experience-dependent changes in the properties of synaptic plasticity observed in the visual cortex, it is clear that a variety of environmental stimuli can induce metaplasticity. Here we will briefly discuss three behavioral modifications that can induce metaplasticity: environmental enrichment, stress, and sleep deprivation. A caveat to these studies is that it is often difficult to distinguish true metaplastic effects from those caused by circulating hormones, neuromodulators, or neurotrophic factors. To partially overcome this limitation, many of these studies have been performed ex vivo where, in principle, brain tissue can be studied in an in vitro setting with tight regulations over the external cellular milieu.
Environmental enrichmentSensory enrichment can induce metaplasticity, depending on the duration of enrichment. While daily 6-hour exposures to a sensory-enriched environment for several weeks does not affect the magnitude of LTP that can be induced in the CA1 region of the hippocampus (Foster and Dumas, 2001), raising rodents in an enriched environment continuously for 5-8 weeks lowers the LTP threshold (Duffy et al., 2001) and raises its magnitude (Huang et al., 2006). Currently the basis for this metaplasticity remains unknown.
As discussed above for the visual system, changes in the sensory environment can induce metaplasticity in primary sensory modalities. A striking example in the somatosensory cortex highlights the idea that sensory-driven metaplasticity is likely to occur throughout the brain and that the mechanisms for metaplasticity may vary regionally. By providing mice with a single-whisker experience, induced by trimming all but one whisker, an intense somatosensory experience is driven along the neural pathway sub-serving the single whisker (Clem et al., 2008). This manipulation induces a strong NMDAR-dependent LTP at layer 4 to layer 2/3 synapses in the somatosensory cortex driven by the single-whisker experience. This LTP can occlude subsequent NMDAR-dependent LTP. However, competency for synaptic strengthening is maintained through a novel form of metaplasticity; an mGluR-dependent form of LTP emerges after the NMDAR-dependent form is eliminated.
StressThe ability to form and maintain memories is powerfully affected by stressors. Thus, it is not surprising that a variety of stressors have been shown to alter synaptic plasticity in the hippocampus (an area of the brain heavily implicated in learning and memory). In general, intensely stressful stimuli inhibit the subsequent induction of LTP and enhance the induction of LTD (Kim et al., 1996). This shift in the properties of synaptic plasticity requires activation of NMDA receptors, as the shift fails to occur if NMDA receptors are blocked during the stressful events.
Sleep deprivationSleep deprivation can impair cognitive performance and, correspondingly, sleep deprivation also alters the induction of synaptic plasticity in the hippocampus. For example, the induction of LTP is severely impaired by sleep deprivation caused by stressful stimulation (Kopp et al., 2007). Because these manipulations also raise levels of the stress hormone corticosterone (Campbell et al., 2002; McDermott et al., 2003), it is not clear whether the changes in plasticity caused by sleep deprivation are due to metaplasticity or whether they are simply a consequence of circulating corticosteroids known to alter LTP induction. However, subtly inducing sleep deprivation by giving animals a large environment to explore also produces metaplasticity, without the confound of altered corticosterone levels (Kopp et al., 2006). This sleep deprivation-induced metaplastic effect has been linked to an increase in the ratio of NR2A- to NR2B-containing NMDA receptors. Thus, the mechanism regulating metaplasticity after sleep deprivation may be similar to that observed for experience-dependent plasticity in the visual cortex.
Metaplasticity in learning
Just as LTD and LTP appear to be important for learning, it is likely that metaplasticity also contributes to the processes that integrate new memories into stable neural networks while preserving existing memories. One form of learning-related metaplasticity entails altering the intrinsic membrane properties of neurons (termed intrinsic plasticity). A number of learning paradigms reduce the size of the slow after-hyperpolarization (AHP) that occurs following an action potential (Cohen-Matsliah et al., 2007; Moyer et al., 1996; Oh et al., 2003), allowing action potentials to fire in more rapid succession. In eye-blink conditioning, temporally pairing a conditioned stimulus (e.g. a tone) with an unconditioned stimulus (e.g. air puff to the eye) can lead to a learned eye-blink response to the conditioned stimulus that is associated with a decrease in the size of the slow AHP observed in hippocampal pyramidal cells (Moyer et al., 1996). These changes are long-lasting (5 days) and are not observed when the conditioned and unconditioned stimuli are presented randomly. Similar reductions in the size of the AHP have been observed in the CA1 region of the hippocampus following water maze learning (Oh et al., 2003) and in pyramidal cells of the piriform cortex after olfactory-discrimination task learning (Cohen-Matsliah et al., 2007). What is the consequence on synaptic plasticity of having a learning-induced reduction in the AHP? It appears that changes in the AHP are associated with metaplasticity, although the valence of the shift appears to vary with the region of the brain and the type of the task. For example, at the same time olfactory learning reduces the AHP in neurons within the piriform cortex, there is also a reduction in the ability to induce LTP and an enhancement in the ability to induce LTD (Lebel et al., 2001; Quinlan et al., 2004). In contrast to what is observed in piriform cortex, a learning-induced reduction in the AHP in CA1 pyramidal neurons appears to lower the threshold for inducing LTP and improve performance in new learning tasks (Disterhoft and Oh, 2006). Notably, the learning-induced reduction in the AHP appears to be specific to the learning process, as pseudo-trained mice fail to exhibit a change in AHP (Zelcer et al., 2006). In the CA1 region, olfactory learning caused a reduction in the AHP before there was evidence that the olfactory discrimination rule had been learned. During this time when the AHP was reduced, the ability to learn a different hippocampal-dependent task was enhanced (Zelcer et al., 2006). Thus, metaplasticity may be important for priming the ability to learn.
IMPORTANCE OF METAPLASTICITY
Metaplasticity helps to prevent synapses from reaching a point of saturation or extinction, by maintaining synapses within a dynamic range of plasticity. The information storage capacity of a neuronal network would be severely limited in the absence of metaplasticity, because synapses would quickly reach an upper or lower limit of plasticity. Consider the example of a learning experience where a subset of synapses become strongly potentiated. Because these synapses are strengthened, they become more likely to drive postsynaptic firing of action potentials and to potentiate again with subsequent neural activity. These synapses would eventually reach a point of saturation if their ability to potentiate were left unchecked. Metaplasticity acts to reset the plasticity threshold to help prevent this saturation, and therefore maintains synapses within a dynamic range capable of both potentiating and weakening. Thus, metaplasticity endows synapses and networks with an ongoing ability to respond to an ever-changing environment. This allows neural networks to develop properly in an experience-driven manner and provides neural networks with a continued capacity to “learn.”
In addition to supporting synaptic learning in networks, metaplasticity may itself be a type of memory trace. That is, metaplasticity represents a trace of the recent history of neural activity. For example, in an environment of reduced neural activity, metaplasticity lowers the LTP induction threshold. Thus a low induction threshold for LTP is an indicator that the recent history of neural activity has been low, while a high LTP induction threshold would indicate the converse. Through the ability either to directly encode information regarding the activation history of a network or to indirectly endow a neural network with the ability to do so, metaplasticity is likely to have an important role in learning.
Abraham, W.C. (2008). Metaplasticity: tuning synapses and networks for plasticity. Nat Rev Neurosci 9, 387.
Abraham, W.C., and Bear, M.F. (1996). Metaplasticity: the plasticity of synaptic plasticity. Trends Neurosci 19, 126-130.
Abraham, W.C., Mason-Parker, S.E., Bear, M.F., Webb, S., and Tate, W.P. (2001). Heterosynaptic metaplasticity in the hippocampus in vivo: a BCM-like modifiable threshold for LTP. Proc Natl Acad Sci U S A 98, 10924-10929.
Akirav, I., and Richter-Levin, G. (2002). Mechanisms of amygdala modulation of hippocampal plasticity. J Neurosci 22, 9912-9921.
Bienenstock, E.L., Cooper, L.N., and Munro, P.W. (1982). Theory for the development of neuron selectivity: Orientation specificity and binocular interaction in visual cortex. J. Neurosci. 2, 32-48.
Bortolotto, Z.A., Bashir, Z.I., Davies, C.H., and Collingridge, G.L. (1994). Metabotropic glutamate receptors activate a molecular switch which regulates the induction of long-term potentiation. Nature 368, 740-743.
Campbell, I.G., Guinan, M.J., and Horowitz, J.M. (2002). Sleep deprivation impairs long-term potentiation in rat hippocampal slices. J Neurophysiol 88, 1073-1076.
Carlson, G., Wang, Y., and Alger, B.E. (2002). Endocannabinoids facilitate the induction of LTP in the hippocampus. Nat Neurosci 5, 723-724.
Chen, X., Yuan, L.L., Zhao, C., Birnbaum, S.G., Frick, A., Jung, W.E., Schwarz, T.L., Sweatt, J.D., and Johnston, D. (2006) Deletion of Kv4.2 gene eliminates dendritic A-type K+ current and enhances induction of long-term potentiation in hippocampal CA1 pyramidal neurons. J Neurosci 26,12143-51.
Chevaleyre, V., and Castillo, P.E. (2004). Endocannabinoid-mediated metaplasticity in the hippocampus. Neuron 43, 871-881.
Christie, B.R., and Abraham, W.C. (1992). Priming of associative long-term depression in the dentate gyrus by theta frequency synaptic activity. Neuron 9, 79-84.
Clem, R.L., Celikel, T., and Barth, A.L. (2008). Ongoing in vivo experience triggers synaptic metaplasticity in the neocortex. Science 319, 101-104.
Cohen-Matsliah, S.I., Brosh, I., Rosenblum, K., and Barkai, E. (2007). A novel role for extracellular signal-regulated kinase in maintaining long-term memory-relevant excitability changes. J Neurosci 27, 12584-12589.
Cohen, A.S., and Abraham, W.C. (1996). Facilitation of long-term potentiation by prior activation of metabotropic glutamate receptors. J. Neurophysiol. 76, 953-962.
Cohen, A. S., Coussens, C. M., Raymond, C. R., and Abraham, W. C. (1999) Long-lasting increase in cellular excitability associated with the priming of LTP induction in rat hippocampus. J Neurophysiol 82, 3139–3148.
Disterhoft, J.F., and Oh, M.M. (2006). Learning, aging and intrinsic neuronal plasticity. Trends Neurosci 29, 587-599.
Duffy, S.N., Craddock, K.J., Abel, T., and Nguyen, P.V. (2001). Environmental enrichment modifies the PKA-dependence of hippocampal LTP and improves hippocampus-dependent memory. Learn Mem 8, 26-34.
Fagiolini, M., Katagiri, H., Miyamoto, H., Mori, H., Grant, S.G., Mishina, M., and Hensch, T.K. (2003). Separable features of visual cortical plasticity revealed by N-methyl-D-aspartate receptor 2A signaling. Proc Natl Acad Sci U S A 100, 2854-2859.
Foster, T.C., and Dumas, T.C. (2001). Mechanism for increased hippocampal synaptic strength following differential experience. J Neurophysiol 85, 1377-1383.
Frey, U., and Morris, R.G. (1997). Synaptic tagging and long-term potentiation. Nature 385, 533-536.
Huang, F.L., Huang, K.P., Wu, J., and Boucheron, C. (2006). Environmental enrichment enhances neurogranin expression and hippocampal learning and memory but fails to rescue the impairments of neurogranin null mutant mice. J Neurosci 26, 6230-6237.
Huang, Y.Y., Colino, A., Selig, D.K., and Malenka, R.C. (1992). The influence of prior synaptic activity on the induction of long-term potentiation. Science 255, 730-733.
Kim, J.J., Foy, M.R., and Thompson, R.F. (1996). Behavioral stress modifies hippocampal plasticity through N-methyl-D-aspartate receptor activation. Proc Natl Acad Sci U S A 93, 4750-4753.
Kirkwood, A., Rioult, M.G., and Bear, M.F. (1996). Experience-dependent modification of synaptic plasticity in visual cortex. Nature 381, 526-528.
Kopp, C., Longordo, F., and Luthi, A. (2007). Experience-dependent changes in NMDA receptor composition at mature central synapses. Neuropharmacology 53, 1-9.
Kopp, C., Longordo, F., Nicholson, J.R., and Luthi, A. (2006). Insufficient sleep reversibly alters bidirectional synaptic plasticity and NMDA receptor function. J Neurosci 26, 12456-12465.
Lebel, D., Grossman, Y., and Barkai, E. (2001). Olfactory learning modifies predisposition for long-term potentiation and long-term depression induction in the rat piriform (olfactory) cortex. Cereb Cortex 11, 485-489.
McDermott, C.M., LaHoste, G.J., Chen, C., Musto, A., Bazan, N.G., and Magee, J.C. (2003). Sleep deprivation causes behavioral, synaptic, and membrane excitability alterations in hippocampal neurons. J Neurosci 23, 9687-9695.
Mockett, B., Coussens, C., and Abraham, W.C. (2002). NMDA receptor-mediated metaplasticity during the induction of long-term depression by low-frequency stimulation. Eur J Neurosci 15, 1819-1826.
Moyer, J.R., Jr., Thompson, L.T., and Disterhoft, J.F. (1996). Trace eyeblink conditioning increases CA1 excitability in a transient and learning-specific manner. J Neurosci 16, 5536-5546.
Oh, M.M., Kuo, A.G., Wu, W.W., Sametsky, E.A., and Disterhoft, J.F. (2003). Watermaze learning enhances excitability of CA1 pyramidal neurons. J Neurophysiol 90, 2171-2179.
Philpot, B.D., Cho, K.K., and Bear, M.F. (2007). Obligatory role of NR2A for metaplasticity in visual cortex. Neuron 53, 495-502.
Quinlan, E.M., Lebel, D., Brosh, I., and Barkai, E. (2004). A molecular mechanism for stabilization of learning-induced synaptic modifications. Neuron 41, 185-192.
Quinlan, E.M., Olstein, D.H., and Bear, M.F. (1999). Bidirectional, experience-dependent regulation of N-methyl-D-aspartate receptor subunit composition in the rat visual cortex during postnatal development. Proc. Natl. Acad. Sci. USA 96, 12876-12880.
Raymond, C.R., Thompson, V.L., Tate, W.P., and Abraham, W.C. (2000). Metabotropic glutamate receptors trigger homosynaptic protein synthesis to prolong long-term potentiation. J Neurosci 20, 969-976.
Roth-Alpermann, C., Morris, R.G., Korte, M., and Bonhoeffer, T. (2006). Homeostatic shutdown of long-term potentiation in the adult hippocampus. Proc Natl Acad Sci U S A 103, 11039-11044.
Wang, H., and Wagner, J.J. (1999). Priming-induced shift in synaptic plasticity in the rat hippocampus. J Neurophysiol 82, 2024-2028.
Wang, Z. R., Xu, N. L., Wu, C. P., Duan, S. M., and Poo, M. M. (2003) Bidirectional changes in spatial dendritic integration accompanying long-term synaptic modifications. Neuron 37, 463–472.
Woo, N.H., and Nguyen, P.V. (2002). "Silent" metaplasticity of the late phase of long-term potentiation requires protein phosphatases. Learn Mem 9, 202-213.
Zelcer, I., Cohen, H., Richter-Levin, G., Lebiosn, T., Grossberger, T., and Barkai, E. (2006). A cellular correlate of learning-induced metaplasticity in the hippocampus. Cereb Cortex 16, 460-468.
- Robert H. Cudmore and Niraj S. Desai (2008) Intrinsic plasticity. Scholarpedia, 3(2):1363.