The idea that cerebellar circuitry plays an important role in learning, the subject of this volume, is one that has robustly withstood the test of time at least since the seminal papers of Brindley and Marr in the 1960s. In recent years, there have been important transformations in our understanding of cerebellar biochemistry, anatomy, and physiology that have changed the way that we think about cerebellar mechanisms that support learning and the forms of behavior that the cerebellum can control. Researchers have revised their accounts of cerebellar learning to account for these changes, and this volume brings some of these together. As with most topics in behavioral neuroscience, these accounts need to span multiple levels of analysis, from molecules through to behavior. The organization of the volume reflects this range.
Masao Ito contributes the opening chapter to the volume with an account of the cellular mechanisms that support long-term depression as a classical model of cerebellar plasticity. Egidio D'Angelo then discusses the potential for multiple forms of plasticity to exist in the cerebellum. In the third chapter, Elisa Galliano and Chris De Zeeuw continue the narrative that challenges three long-held traditional ideas and propose ways in which these could be revised.
There follow three chapters that study cerebellar learning using relatively simple, cerebellar-dependent forms of learning in well-understood models. The benefits of these models include relatively specific questions that can be addressed and high levels of experimental control that can be achieved. In the first of these, Michael Longley and Christopher Yeo discuss the value of using the classically conditioned eyeblink and nictitating membrane responses to study cerebellar mechanisms of learning, and summarize findings from lesion and inactivation experiments. The chapter makes comparisons between learning mechanisms that support eye blink and NMR conditioning with those that support the vestibulo-ocular reflex (VOR). In the next chapter, Anders Rasmussen and Germund Hesslow discuss how work using classical eyeblink conditioning has contributed to an understanding of how learning-related feedback is itself regulated by learning-related cerebellar outputs. In the last of the three chapters that focus on si ple behaviors, Suryadeep Dash and Peter Thier discuss the roles played by specific areas of the primate cerebellar cortex in the adaptation of three kinds of eye movement behaviors (the adaptation of the VOR, saccades, and smooth pursuit).
In Chapter 7, Paul Dean and John Porrill contribute a computational account of decorrelation learning in the cerebellum, where parallel fiber synapses are weakened if they correlate positively with climbing fiber input but strengthened if they are negatively correlated. The authors discuss the application of this approach to motor control, sensory prediction, and higher cognitive function.
The evolution of the corticocerebellar system is an important subject in its own right, but in this literature little attention has been paid to it in the context of cerebellar learning. Jeroen Smaers therefore contributes a chapter that takes a comparative approach in which he discusses the evolutionary expansion of the cerebellum in the context of enhanced motor and cognitive skills.
The last two chapters focus on cerebellar interactions with frontal lobe areas and the roles that this plays in more complex forms of learning. They acknowledge the fact that such learning must involve interactions between explicit and implicit processes involving both the cerebellum and the neocortex. In the first, Jordan Taylor and Richard Ivry consider the role of the cerebellum in adaptation, strategy, and reinforcement learning. They suggest that cerebellar mechanisms are engaged in the skilled movement execution. However, the authors are skeptical about the involvement of cerebellar circuitry in skilled action selection and suggest that prefrontal and basal ganglia mechanisms are likely to play a prominent role in this. I contribute the final chapter of the volume and focus on similar issues. In contrast to the authors of the previous chapter, I suggest that the cerebellum might play an important role in the acquisition of both motor and cognitive skills and in the formation of habits through instrum ntal learning. I also suggest that just as cerebellar outputs suppress the processing of error feedback in classical eyeblink conditioning, it might also suppress higher forms of feedback such as reward error in more complex forms of learning.
The study of cerebellar learning has a very broad span and this research area is growing fast. It is not possible to do justice to all of the ideas in the field in a single volume, but the hope is that the chapters in this volume reflect current perspectives that range from molecular and cellular mechanisms of learning-related plasticity, through computational accounts and simple forms of learning in animals, to systems-level accounts of complex learning in humans. Narender Ramnani

Contributors v

Preface vii

CHAPTER 1 Long-Term Depression as a Model of Cerebellar Plasticity 1

1.A Historical Overview of LTD Studies 2

      1.1.Leading Theories 2

      1.2.Experimental Approach to LTD and Motor Learning 3

      1.3.Memory Mechanisms in the Cerebellum 3

      1.4.Recent Issues 4

2.Molecular Mechanisms of LTD 4

      2.1.Signal Transduction Underlying LTD 4

      2.2.Constitutive Trafficking of AMPA Receptors 6

      2.3.Receptor Recycling in LTP 7

      2.4.Receptor Recycling in LTD 8

3.Roles of LTD in Motor Learning 9

      3.1.Adaptation of HOKR 9

      3.2.Adaptation of HVOR 11

      3.3.LTD in Fast HOKR Adaptation 11

      3.4.Memory Transfer in Slow HOKR Adaptation 12

      3.5.LTD and Memory Transfer 13

      3.6.Saccade and Other Reflexes 13

      3.7.Arm Movement 14

4.Significance of LTD in Cerebellar Neural Network 14

      4.1.Role of Plasticity: Memory Formation Versus Signal Enhancement 14

      4.2.Induction Mechanism: Climbing-Fiber Specific Changes Versus Nonspecific Changes 16

      4.3.Learning Principles: Supervised Learning Versus Unsupervised Learning 17

      4.4.Ideal Location for Memory: Purkinje Cells Versus Other Cells 17

      4.5.Memory Formation Process: Acquisition Versus Consolidation 18

      4.6.Time at Impairment: Mature Brain Versus Developing Brain 18

5.LTD Versus Learning Mismatch 19

      5.1.Experimental Gaps in Testing of LTD and Motor Learning 19

      5.2.Compensation Might Restore Motor Learning 20

      5.3.Limitations in Testing LTD in Slices 20

6.Perspectives 21

Acknowledgments 22

References 22

CHAPTER 2 The Organization of Plasticity in the Cerebellar Cortex: From Synapses to Control 31

1.Introduction 32

2.Plasticity in the Granular Layer 34

3.Mossy Fiber-Granule Cell LTP and LTD 34

      3.1.Plasticity in the Granular Layer Inhibitory Circuit 35

      3.2.Plasticity in the Granular Layer In Vivo 35

      3.3.The Consequences of Granular Layer Plasticity: Geometry, Timing, and Coding 35

      3.4.Theoretical Implications 36

4.Plasticity in the Molecular Layer 37

      4.1.Mechanisms of Postsynaptic Parallel Fiber LTP and LTD 37

      4.2.Mechanisms of Presynaptic Parallel Fiber LTP and LTD 39

      4.3.Mechanisms of Climbing Fiber LTD 40

      4.4.Plasticity of Purkinje Cell Intrinsic Excitability 40

      4.5.Plasticity at Molecular Layer Inhibitory Synapses 40

      4.6.The Neurophysiological Consequences of Molecular Layer Plasticity 41

      4.7.The Behavioral Consequences of Molecular Layer Plasticity 42

5.An Integrated View of Cerebellar Cortical Plasticity 42

      5.1.Potentiation of Transmission Channels and Signal-to-Noise Ratio in the Mossy Fiber Pathway 44

      5.2.Contrast Enhancement and Geometrical Organization of Plasticity 45

      5.3.Coordination of Plasticity During Patterned Circuit Activity 45

      5.4.Gating of Plasticity by Neuromodulatory Systems 46

6.Cerebellar Cortical Plasticity and Timing 46

7.Integration of Plasticity in the Cerebellar Cortex and Nuclei 47

      7.1.Plasticity in the Deep Cerebellar Nuclei 48

      7.2.The Effect of Integrated Plasticity in the Cerebellar Cortex and Nuclei 48

8.Cerebellar Plasticity in Learning and Control 49

9.Conclusions 50

Acknowledgments 51

References 51

CHAPTER 3 Questioning the Cerebellar Doctrine 59

1.The Cerebellar Doctrine and Its Three Pillars 59

2.The First Pillar: The Sole Cerebellar Function Is to Control Motor Behavior 61

3.The Second Pillar: Inputs Converge Only at the Level of Purkinje Cells 68

4.The Third Pillar: Depression at the Parallel Fiber to Purkinje Cell Synapse Is the Molecular Substrate of Cerebellar Learning 70

5.Concluding Remarks 72

References 73

CHAPTER 4 Distribution of Neural Plasticity in Cerebellum-Dependent Motor Learning 79

1.Introduction 79

2.Cerebellum-Dependent Learning—Eyeblink and NMR Conditioning as Behavioral Models for Analysis 81

3.Lesion Studies Reveal that NMR Conditioning Depends upon Cerebellar Compartments with C1 and C3 Cortical Zones in Lobule HVI 83

4.Inactivation Experiments Reveal Essential Roles for the Cerebellar Nuclei and Inferior Olive in the Acquisition of NMR and Eyeblink Conditioning 87

5.Inferior Olive Function in NMR and Eyeblink Conditioning 89

6.Cerebellar Cortex Function in NMR and Eyeblink Conditioning 91

7.Distribution of Plasticity at Cerebellar Cortical and Cerebellar Nuclear, or Brainstem, Levels 92

8.Cortical Plasticity in NMR and Eyeblink Conditioning 94

9.Conclusions 95

Acknowledgment 95

References 95

CHAPTER 5 Feedback Control of Learning by the Cerebello-Olivary Pathway 103

1.Feedback is Essential for Learning 104

2.Anticipating Consequences 104

3.Classical Conditioning 105

4.The Cerebellar Microcomplex 106

5.Classical Conditioning Requires the Cerebellum 108

6.The Nucleo-Olivary Pathway and Negative Feedback 108

7.Reaching Equilibrium 111

8.Back to Behavior 114

9.Feedback, Anticipation, and Nucleo-Olivary Inhibition 114

10.Broadening the Perspective 115

References 117

CHAPTER 6 Cerebellum-Dependent Motor Learning: Lessons from Adaptation of Eye Movements in Primates 121

1.Introduction 122

2.Adaptation of the VOR 122

3.Short-Term Saccadic Adaptation 124

4.Smooth Pursuit Adaptation 126

5.Oculomotor Cerebellum—An Overview 128

6.Floccular Complex 130

      6.1.Anatomical Considerations 130

      6.2.Role of FC in VOR and VOR Adaptation 130

      6.3.Role of FC in SPEMs and SPA 132

7.Oculomotor Vermis 134

      7.1.Anatomical Considerations 134

      7.2.Role of the OMV in Saccades, STSA, and Saccadic Resilience 135

      7.3.Role of the OMV in SPEMs, SPA, and SPEM Resilience 142

8.Complex Spike Activity During STSA and SPA 144

9.Conclusions 149

References 149

CHAPTER 7 Decorrelation Learning in the Cerebellum: Computational Analysis and Experimental Questions 157

1.Introduction 158

2.Implementation of Learning Rule 160

      2.1.STDP Version 160

      2.2.Experimental Questions 162

3.Properties of Learning Rule 164

4.Sensory Prediction 165

      4.1.The Reafference Problem 165

      4.2.General Sensory Prediction 173

5.Motor Control 173

      5.1.Signaled-Avoidance Learning 173

      5.2.Gaze Stabilization 177

      5.3.General Motor Control 180

6.Future Directions 181

      6.1.Coordination 181

      6.2.Cognitive Tasks 182

Appendix Derivation of Supervised-Leaning Rule 184

Acknowledgments 185

References 185

CHAPTER 8 Modeling the Evolution of the Cerebellum: From Macroevolution to Function 193

1.Cerebellum, Learning, and Human Evolution 193

2.Evolutionary Neuroscience and Its Adoption of the Cerebellum 194

3.Comparative Studies of Cerebellar Connectivity 197

4.Macroevolutionary Studies of the Cerebellum 198

      4.1.(Phylogenetic) Scaling: Who Has the Biggest Cerebellum? 198

      4.2.Phylogenetic Correlations: Toward Patterns of Evolutionary Connectivity 203

      4.3.Phylogenetic Mapping: Inferring Detailed Patterns of Evolutionary History 204

5.How Can Macroevolutionary Studies Contribute to Our Understanding of Cerebellar Function? 207

6.Summary 210

Acknowledgment 211

References 211

CHAPTER 9 Cerebellar and Prefrontal Cortex Contributions to Adaptation, Strategies, and Reinforcement Learning 217

1.Introduction 217

2.The Cerebellum and Error-Based Learning 218

3.Computational Models of Sensorimotor Adaptation 222

4.Multiple Learning Mechanisms in Sensorimotor Adaptation 224

5.Strategy Use During Sensorimotor Adaptation 229

6.Cerebellar and Neocortical Contributions to Sensorimotor Adaptation 234

7.Systems Interaction in Sensorimotor Learning 238

8.Cerebellum and Sensorimotor Learning: Beyond Adaptation 243

9.Conclusions 245

Acknowledgments 246

References 246

CHAPTER 10 Automatic and Controlled Processing in the Corticocerebellar System 255

1.Dual Systems, Skills, and Habits 256

2.Control Theory 257

3.The Cerebellum and Forward Models 259

4.Cognitive Habits 265

      4.1.Cerebellar Cortex and Higher Level Feedback 269

      4.2.Could Climbing Fibers Convey Errors Related to Cognitive Processing? 271

      4.3.Cerebellar Communication with the Dopamine System 275

5.Conclusion 277

Acknowledgment 278

References 278

Index 287

Other volumes in PROGRESS IN BRAIN RESEARCH 295

中国医科院医学信息研究所