Portal de Periódicos da CAPES

Multiple Memory Mechanisms in the Cerebellum?

2006; Cell Press; Volume: 51; Issue: 6 Linguagem: Inglês

10.1016/j.neuron.2006.09.010

1097-4199

Ka Hung Lee, Richard F. Thompson,

Vestibular and auditory disorders

Long-term potentiation (LTP) and long-term depression (LTD) are arguably two of the most widely discussed cellular plasticity mechanisms for learning and memory. However, the extent to which they are required for behavioral plasticity and learning is not clear. In this issue of Neuron, Boyden et al. use mice lacking CaMKIV and Hansel et al. use mice lacking αCaMKII to assess the contribution of LTD to cerebellar learning. Long-term potentiation (LTP) and long-term depression (LTD) are arguably two of the most widely discussed cellular plasticity mechanisms for learning and memory. However, the extent to which they are required for behavioral plasticity and learning is not clear. In this issue of Neuron, Boyden et al. use mice lacking CaMKIV and Hansel et al. use mice lacking αCaMKII to assess the contribution of LTD to cerebellar learning. The two most widely studied and best understood forms of cerebellar-dependent learning and memory are adaptation of the vestibule-ocular reflex (VOR) and classical conditioning of eyeblink and other discrete responses (Christian and Thompson, 2003Christian K.M. Thompson R.F. Learn. Mem. 2003; 10: 427-455Crossref PubMed Scopus (488) Google Scholar, du Lac et al., 1995du Lac S. Raymond J.L. Sejnowski T.J. Lisberger S.G. Annu. Rev. Neurosci. 1995; 18: 409-441Crossref PubMed Scopus (176) Google Scholar). The VOR acts to counterbalance the effect of head movement by producing compensatory eye movements in the opposite direction of head movement, which thereby stabilizes images on the retina and prevents blurred vision. Adaptation of the VOR and eyeblink conditioning have somewhat analogous structural bases. In both cases, adaptation in initial cerebellar learning critically involves the cerebellar cortex, while the cerebellar and vestibular nuclei play a more critical role in long-term memory storage (Christian and Thompson, 2005Christian K.M. Thompson R.F. Behav. Neurosci. 2005; 119: 526-537Crossref PubMed Scopus (58) Google Scholar, du Lac et al., 1995du Lac S. Raymond J.L. Sejnowski T.J. Lisberger S.G. Annu. Rev. Neurosci. 1995; 18: 409-441Crossref PubMed Scopus (176) Google Scholar, Kleim et al., 2002Kleim J.A. Freeman Jr., J.H. Bruneau R. Nolan B.C. Cooper N.R. Zook A. Walters D. Proc. Natl. Acad. Sci. USA. 2002; 99: 13228-13231Crossref PubMed Scopus (139) Google Scholar). What are the cellular and molecular mechanisms that underlie cerebellar learning? In vitro studies have pointed to a large number of plasticity mechanisms operational within the cerebellar circuits. However, the contribution of these mechanisms to specific forms of behavioral plasticity remains less clear. Ito first proposed cerebellar LTD as the mechanism in the cerebellar flocculus for adaptation of the VOR (Ito, 1982Ito M. Annu. Rev. Neurosci. 1982; 5: 275-296Crossref PubMed Scopus (372) Google Scholar). Cerebellar long-term depression (LTD) is also widely viewed as a possible mechanism of synaptic plasticity of other forms of cerebellar-dependent learning as well (Linden and Connor, 1995Linden D.J. Connor J.A. Annu. Rev. Neurosci. 1995; 18: 319-357Crossref PubMed Scopus (352) Google Scholar). The basic issue addressed by both the Boyden et al. and Hansel et al. studies concerns whether LTD is the mechanism underlying adaptation of the VOR (and of a related form of learning called the optokinetic response, OKR) (Boyden et al., 2004Boyden E.S. Katoh A. Raymond J.L. Annu. Rev. Neurosci. 2004; 27: 581-609Crossref PubMed Scopus (320) Google Scholar, Hansel et al., 2001Hansel C. Linden D.J. D'Angelo E. Nat. Neurosci. 2001; 4: 467-475Crossref PubMed Scopus (475) Google Scholar). Both studies address the issue using knockout mice, Boyden et al. using CaMKIV mice and Hansel et al. using the αCaMKII mice (Boyden et al., 2006Boyden E.S. Katoh A. Pyle J.L. Chatila T.A. Tsien R.W. Raymond J.L. Neuron. 2006; 51 (this issue): 823-834Abstract Full Text Full Text PDF PubMed Scopus (109) Google Scholar, Hansel et al., 2006Hansel C. de Jeu M. Belmeguenai A. Houtman S.H. Buitendijk G.H.S. Adreev D. De Zeeuw C.I. Elgersma Y. Neuron. 2006; 51 (this issue): 835-843Abstract Full Text Full Text PDF PubMed Scopus (168) Google Scholar). The roles of both αCaMKII and CaMKIV in synaptic plasticity and learning have been well characterized in the hippocampal system. CaMKIV mutant mice can learn hippocampal-dependent contextual fear conditioning normally but are impaired in long-term memory (1 and 7 days) (Wei et al., 2002Wei F. Qiu C.-S. Liauw J. Robinson D.A. Ho N. Chatila T. Zhuo M. Nature. 2002; 5: 573-579Google Scholar). Hippocampal LTP in CaMKIV mutants is impaired throughout the 45 min after the tetanus, but hippocampal LTD remains intact in slices from these animals (Ho et al., 2000Ho N. Liauw J.A. Blaeser F. Wei F. Hanissian S. Muglia L.M. Wozniak D.F. Nardi A. Arvin K.L. Holtzman D.M. et al.J. Neurosci. 2000; 20: 6459-6472Crossref PubMed Google Scholar). Similarly, αCaMKII mutant mice are impaired in the Morris water maze initially, but their performance catches up with the wild-type mice after successive training (Silva et al., 1992bSilva A.J. Paylor R. Wehner J.M. Tonegawa S. Science. 1992; 257: 206-211Crossref PubMed Scopus (1041) Google Scholar). Consistent with the hippocampus-dependent behavioral impairment, αCaMKII mutants have a hippocampal LTP deficit throughout the 60 min after the tetanus (Silva et al., 1992aSilva A.J. Stevens C.F. Tonegawa S. Wang Y. Science. 1992; 257: 201-206Crossref PubMed Scopus (1139) Google Scholar). CaMKIV has also been implicated previously in the maintenance of cerebellar LTD. In the CaMKIV KO mice, cerebellar LTD can be induced, but it is not maintained (Ho et al., 2000Ho N. Liauw J.A. Blaeser F. Wei F. Hanissian S. Muglia L.M. Wozniak D.F. Nardi A. Arvin K.L. Holtzman D.M. et al.J. Neurosci. 2000; 20: 6459-6472Crossref PubMed Google Scholar). Boyden et al. took advantage of these mice to test the model first proposed by Ito and colleagues that LTD is the mechanism for the adaptive VOR plasticity. CaMKIV is expressed in adult Purkinje cells in the cerebellum in regions implicated in VOR motor learning. Motor learning of the VOR can be induced in the laboratory by specific training protocols involving pairing rotation of the animal's head with changes in the surrounding visual stimuli. Moving the head in the opposite direction as the visual stimulus causes an adaptive increase in the amplitude (gain) of the response, while moving the head and visual stimulus in the same direction causes an adapative decrease in the VOR gain. Previous in vitro results showed that LTD induction is normal in the CaMKIV knockout mice. In line with these prior observations, Boyden et al. found that the initial acquisition for increased or decreased gain of the VOR with high-frequency stimuli is normal in the CaMKIV knockout mice, suggesting that learning in these mice is normal. However, the effects on memory retention at the behavioral level were more complicated than would be predicted by the simple model that LTD is required for VOR learning. While the 24 hr memory for increased gain was impaired, the memory for decreased gain was not impaired. Further, memory for increased gains with low-frequency stimuli was also intact. From these results, the authors argue that although LTD is a likely mechanism for one aspect of VOR plasticity (retention of an adaptive learned response to increased gain), it cannot subserve other aspects. The authors propose that plasticity mechanisms may be used in a task-selective fashion. A different approach to this same question of whether LTD is the universal plasticity mechanism is to determine to what extent LTD can also serve as a mechanism for other kinds of cerebellar memory. In our lab, we have trained the same CaMKIV KO mice in eyeblink conditioning, and the results are clear. Similar to the results in Boyden et al., we have shown that animals learn the conditioned response normally, comparable to the wild-type controls, but their long-term memory for the learned response is markedly impaired (K. Lee, N.Q. Truong, T.A. Chatila, R.A. Ram, and R.F. Thompson, 2004, Soc. Neurosci., abstract). So LTD can serve as one of the mechanisms for these two quite different forms of cerebellar-dependent memory. Previous studies had also assessed motor learning in animals deficient in LTD induction. As an example, Shibuki et al., 1996Shibuki K. Gomi H. Chen L. Bao S. Kim J.J. Wakatsuki H. Fujisaki T. Fujimoto K. Katoh A. Ikeda T. et al.Neuron. 1996; 16: 587-599Abstract Full Text Full Text PDF PubMed Scopus (291) Google Scholar showed that mutants with complete absence of GFAP in glia show no cerebellar cortical LTD and are markedly impaired in eyeblink conditioning. (Incidentally, this is an intriguing example of a critical role glia may play in neuronal plasticity and memory.) In a recent study, Shutoh et al., 2006Shutoh F. Ohki M. Kitazawa H. Itohara S. Nagao S. Neuroscience. 2006; 139: 767-777Crossref PubMed Scopus (143) Google Scholar analyzed the mechanisms of memory storage of adaptation of the horizontal optokinetic response (OKR), a form of learning analogous to the VOR. They found that the flocculus was essential for initial learning but the long-term trace (1 week) appeared to be established in the medial vestibular nucleus. They also found that both day-long and week-long adaptations were depressed when neural nitric oxide synthase was pharmacologically or genetically disrupted, thus supporting LTD as a mechanism for both short-term and long-term plasticity. It would be interesting to see the effects of blocking neural nitric oxide synthase in the flocculus in the controls and CamKIV knockouts used in the Boyden et al. study. In a related article, Hansel et al. (Hansel et al., 2006Hansel C. de Jeu M. Belmeguenai A. Houtman S.H. Buitendijk G.H.S. Adreev D. De Zeeuw C.I. Elgersma Y. Neuron. 2006; 51 (this issue): 835-843Abstract Full Text Full Text PDF PubMed Scopus (168) Google Scholar) explore related issues in the CamKII mutant mice. Like CaMKIV, αCaMKII is expressed by Purkinje cells in the cerebellum. Activation of CaMKII by calcium influx has been proposed to be a common requirement for LTP induction at cortical and hippocampal excitatory synapses, but whether CaMKII plays a similar role in cerebellar plasticity has not yet been addressed. Here, Hansel et al. report that cerebellar cortical LTP (i.e., the Purkinje neuron response to repeated parallel fiber stimulation) is normal in the αCaMKII mutant. Conversely, LTD (i.e., the joint activation of parallel and climbing fibers) is decreased in juvenile αCaMKII mutant animals and becomes potentiated in adult αCaMKII. Interestingly, these results would appear to be opposite of what is observed in the hippocampus at the CA3-CA1 synapse, where αCaMKII KO is required for LTP but not LTD. Behaviorally, the αCaMKII KO mutants showed impaired gain-increase adaptation of both the VOR and the OKR. While 24 hr retention of adaptation was not examined, the mutants did show adaptation in the gain-decrease paradigm, although less than wild-type mice. The authors noted that while cerebellar morphology appears grossly normal in the adult animals at the EM light microscopy level, climbing fiber elimination is delayed in the αCaMKII mice. To address whether this delay in climbing fiber elimination could contribute to the decreased LTD seen in the juvenile animals, the authors make use of a selective CaMKII inhibitor, KN-93. Application of KN-93 to wild-type slices from juvenile animals prevented the induction of LTD and in fact actually led to a potentiative response, arguing that αCaMKII plays a direct role in parallel fiber LTD and that the unstable LTD observed in the αCaMKII mice is not due to a secondary effect resulting from the delayed climbing fiber elimination. Interestingly, Chen et al., 1995Chen C. Kano M. Abeliovich A. Chen L. Bao S. Kim J.J. Hashimoto K. Thompson R.F. Tonegawa S. Cell. 1995; 83: 1233-1242Abstract Full Text PDF PubMed Scopus (262) Google Scholar showed that PKC γ KO mice show persistent multiple climbing fiber innervation of Purkinje neurons. These mice display normal LTD but significantly enhanced learning of the conditioned eyeblink response, supporting a general role for climbing fibers as the unconditioned stimulus teaching input in this paradigm. Both the Boyden et al. and Hansel et al. studies provide support for Ito's LTD hypothesis (Ito, 1982Ito M. Annu. Rev. Neurosci. 1982; 5: 275-296Crossref PubMed Scopus (372) Google Scholar) for adaptation of the VOR, at least for increased gain with higher-frequency stimuli, but argue that other mechanisms of neuronal plasticity must also be involved in other aspects of cerebellar learning (e.g., learning in response to decreased gain and with low-frequency stimulation). As the authors acknowledge in each case, one caveat is that, in both studies, the evidence is basically correlational: the mutants exhibit impaired cerebellar LTD and alterations in adaptation of the VOR (and OKR) in some conditions but not others. With correlations it is always possible that other factors could result in both effects. One such possibility, noted by Boyden et al., is the occurrence of altered patterns of spiking activity important for induction of plasticity at several sites in the circuit (e.g., Smith and Otis, 2003Smith S.L. Otis T.S. J. Neurosci. 2003; 23: 367-372PubMed Google Scholar). Boyden et al. obviate this possibility because the original induction of LTD and adaptation of the VOR were normal, only retention was impaired. While the studies of Boyden et al. and Hansel et al. therefore provide us with important insights into the signal transduction mechanisms that are important for cerebellar LTD, further work will be required to firmly establish the causal role of these, and other, factors in various forms of cerebellar plasticity and learning. αCaMKII Is Essential for Cerebellar LTD and Motor LearningHansel et al.NeuronSeptember 21, 2006In BriefActivation of postsynaptic α-calcium/calmodulin-dependent protein kinase II (αCaMKII) by calcium influx is a prerequisite for the induction of long-term potentiation (LTP) at most excitatory synapses in the hippocampus and cortex. Here we show that postsynaptic LTP is unaffected at parallel fiber-Purkinje cell synapses in the cerebellum of αCaMKII−/− mice. In contrast, a long-term depression (LTD) protocol resulted in only transient depression in juvenile αCaMKII−/− mutants and in robust potentiation in adult mutants. Full-Text PDF Open ArchiveSelective Engagement of Plasticity Mechanisms for Motor Memory StorageBoyden et al.NeuronSeptember 21, 2006In BriefThe number and diversity of plasticity mechanisms in the brain raises a central question: does a neural circuit store all memories by stereotyped application of the available plasticity mechanisms, or can subsets of these mechanisms be selectively engaged for specific memories? The uniform architecture of the cerebellum has inspired the idea that plasticity mechanisms like cerebellar long-term depression (LTD) contribute universally to memory storage. To test this idea, we investigated a set of closely related, cerebellum-dependent motor memories. Full-Text PDF Open Archive