Information in the brain is transmitted from neuron to neuron through specialized connections called synapses. A synapse between two neurons is made up of presynaptic and postsynaptic terminals, which are separated by a synaptic cleft. The presynaptic terminal is filled with small vesicles containing chemical neurotransmitters, and the postsynaptic terminal consists of receptors specific for these neurochemicals. Neurons carry information in the form of an electrical impulse called an action potential that is initiated at the cell body and travels down the axon. At the synapse, an action potential causes the voltage-dependent release of neurotransmitter-filled vesicles, thereby converting an electrical impulse into a chemical signal. Neurotransmitters diffuse across the synaptic cleft, where they bind to receptors and generate an electrical signal in the postsynaptic neuron. The postsynaptic cell will then, in turn, fire an action potential if the sum of all its synapses reaches an electrical threshold for firing. Since a neuron can receive synapses from many different presynaptic cells, each cell is able to integrate information from varied sources before passing along the information in the form of an electrical code. The ability of neurons to modify the strength of existing synapses, as well as form new synaptic connections, is called neuroplasticity. It is believed that neuroplasticity may be the underlying cellular mechanism for the brain's ability to encode information during learning. In theory, this is how information is stored as memory.
Defined in this way, neuroplasticity includes changes in strength of mature synaptic connections, as well as the formation and elimination of synapses in adult and developing brains. This encompasses a vast field of research, and similar processes may also occur at peripheral synapses, where much of the pioneering studies on synaptic transmission first took place. In addition, neuroplasticity includes the regrowth (or sprouting) of new synaptic connections following central nervous system injury; following stroke, for example.
The notion that the brain can store information by modifying synaptic connections is not a new one. In fact, Santiago Ramon y Cajal (a founder of modern neuroscience) expressed this theory in 1894, three years before Charles Sherrington coined the term synapse to describe the connections made between neurons. In the late 1940s the neuroplasticity model was advanced by Jerzy Konorski, who used the word plasticity to describe "permanent functional transformations," and Donald Hebb, who ascribed testable physiologic characteristics to synaptic plasticity. However, experimental evidence that synapses are capable of long-lasting changes in synaptic strength did not come until the early 1970s, when Timothy Bliss and Terry Lomo described an increase in the synaptic strength of neurons in the mammalian hippocampus (a region of the brain critical for some forms of memory) following electrical stimulation. They termed this increase long-lasting potentiation, now referred to as long-term potentiation (LTP).
Changes in synaptic strength proved to be bidirectionally modifiable (they increase and decrease in strength) as Serena Dudek and Mark Bear first demonstrated in 1992 by recording activity-driven, long-term depression (LTD) in the hippocampus. The evidence that learning and memory are based on these long-lasting changes in synaptic strength is substantial, but still incomplete. However, defining the molecular constituents in the mechanistic pathway leading from synaptic activity to plasticity continues to strengthen the evidence linking neuroplasticity with learning and memory. In addition, resolving the molecular mechanisms underlying synaptic modification should lead to targets for clinical intervention in eliminating age-related memory loss or synaptic loss following brain damage by enhancing new synaptic connections.
Mechanisms of plasticity
Synaptic plasticity can occur at either the presynaptic or postsynaptic terminal. Modifications to the presynaptic terminal affect the release of neurotransmitters. As the action potential invades the presynaptic terminal, it activates voltage-gated calcium channels that conduct calcium ions into the presynaptic terminal. This rise in intracellular calcium triggers the exocytosis of vesicles (fusion with the plasma membrane) and thus the release of neurotransmitters. Each presynaptic terminal contains between 200 and 500 vesicles, though only a small proportion of these are ready to be released at any time. Vesicles in the presynaptic terminal move through a specific release cycle, including vesicle storage, priming for release, release, vesicle reformation, and reloading with neurotransmitter.
Factors that alter the presynapse resulting in either modification of the calcium channel conductance or modification of the vesicle cycle will yield changes in synaptic strength. One such factor is the cyclic nucleotide cAMP. An increase in cAMP presynaptically can enhance transmitter release by activating protein kinase A (PKA). PKA activation induces a decrease in a specific potassium channel conductance called a delayed rectifier current. Decreased delayed rectifier conductance will increase the calcium entry into the presynaptic terminal by increasing the duration of the action potential. In addition, a rise in cAMP can activate vesicular release from presynaptic terminals that were previously dormant. Such terminals are present, but do not release neurotransmitters in response to an action potential prior to a rise in cAMP. A morphologically distinct synapse that is physiologically dormant has been termed a silent synapse and can be the result of deficient presynaptic release, or a deficiency of transmitter receptors expressed postsynaptically.
The postsynaptic terminal can also be modified to produce changes in synaptic efficacy. Signaling molecules in the postsynaptic compartment such as protein kinase A (PKA) and the alpha subunit of calcium/calmodulin-dependent kinase II (α-CaMKII) are thought to play major roles in synaptic plasticity. For example, when a mouse is genetically altered to express a version of α-CaMKII incapable of activation, LTP and learning are disrupted. While α-CaMKII can directly phosphorylate neurotransmitter receptors leading to an increase in conductance, it is likely to play additional roles in synaptic plasticity as well. Neurotransmitter receptors can cycle in and out of the postsynaptic membrane (in a process not unlike the presynaptic vesicles), and α-CaMKII phosphorylation of an as yet unidentified substrate could lead to the rapid insertion of more receptors. This would result in LTP of an active synapse and the unsilencing of a synapse that was not previously expressing these receptors in its membrane. As stated above, there is substantial evidence implicating long-lasting changes in synaptic strength with the formation of memory. It should be noted that synapses do not act in isolation. The neural circuits to which they belong are a result of the many thousands of synapses contained therein. Although the cellular coding of information may be encoded at synapses, memory itself is likely dependent upon the circuit(s) in which they are contained.
Plasticity, memory, and aging
As humans age, an impairment of memory occurs that is not associated with neurological damage or disease. The age of onset for this decline varies, but it is clear that this is a selective deficit and not a generalized decrease in cognitive skills. Moreover, the deficit is also apparent in animal models of aging and is manifest as a greater number of trials required to memorize a task and a decrease in memory retention that begins approximately twenty-four hours post-training. Interestingly, LTP also changes with age, typically requiring a more robust stimulus to induce and yielding a synaptic potentiation that decays more rapidly. Since aging animals and humans both maintain the ability to store memory, the fundamental mechanisms that underlie information storage may remain essentially intact. The deficit may not be a lack of ability, but rather a decline in the efficiency of storage—or an inability to maintain the neural plasticity induced during learning. Since the formation of memory is dependent on new protein synthesis, one way to address the decreased stability of memory is to identify proteins made during learning. Consistent with this, synaptic plasticity has at least two temporally distinct components: transient changes that do not require new protein synthesis, and enduring modifications (e.g., LTP and LTD) that require the production of new proteins. Identification of newly formed proteins, their site of action, and the molecular basis for their role in neural plasticity may provide insights into the maintenance of memory, and thus indicate clinical targets for the amelioration of age-related memory decline.
David G. Wells
See also Brain; Learning; Memory; Neurochemistry; Neurodegenative Diseases.
Bliss, T. V. P., and Lomo, T. "Long-Lasting Potentiation of Synaptic Transmission in the Dentate Area of the Anaesthetized Rabbit Following Stimulation of the Perforant Path." Journal of Physiology (London) 232 (1973): 331–356.
Cowen, W. M., and Kandel, E. R. "A Brief History of Synapses and Synaptic Transmission." In Synapses. Edited by W. M. Cowen, T. C. Sudhof and C. F. Stevens, Baltimore, Md.: The Johns Hopkins University Press, 2001. Pages 1–88.
Davis, H. P., and Squire, L. R. "Protein Synthesis and Memory: A Review." Psychology Bulletin 96 (1984): 518–559.
Dudek, S. M., and Bear, M. F. "Homosynaptic Long-Term Depression in Area CA1 of Hipocampus and Effects of N-methyl-D-aspartate Receptor Blockade." Proceedings of the National Academy of Science 89 (1992): 4363–4367.
Foster, T. C. "Involvement of Hippocampal Synaptic Plasticity in Age-Related Memory Decline." Brain Research Review 30 (1999): 236–249.
Giese, K. P.; Fedorov, N. B.; Filipkowski, R. K.; and Silva, A. J. "Autophosphorylation at Thr286 of the Alpha Calcium-Calmodulin Kinase II in LTP and Learning." Science 279 (1998): 870–873.
Hayashi, Y.; Shi, S.-H.; Esteban, J. A.; Piccini, A.; Poncer, J. C.; and Malinow, R. "Driving AMPA Receptors into Synapses by LTP and CaMKII: Requirements for GluR1 and PDZ Domain Interactions." Science 287 (2000): 2262–2267.
Ma, L.; Zablow, L.; Kandel, E. R.; and Siegelbaum, S. A. "Cyclic AMP Induces Functional Presynaptic Boutons in Hippocampal CA3-CA1 Neuronal Cultures." National Neuroscience 2 (1999): 24–30.
Tong, G.; Malenka, R. C.; and Nicoll, R. A. "Long-Term Potentiation in Cultures of Single Hippocampal Granule Cells: A Presynaptic Form of Plasticity." Neuron 16 (1996): 1147–1157.