|
|
| Line 1: |
Line 1: |
| There are few things as captivating as witnessing an act of immense human skill: a gymnast deftly prancing along a balance beam; a musical virtuoso firing off notes with rapid precision; a blind man using sound to pinpoint objects with amazing precision. However, these achievements would not be possible without practice and repetition. The brain’s ability to reorganize as it integrates new experience, a process known as plasticity, makes such marvels possible. Plasticity within the cortex allows previous events shape and optimize the way our brain interprets sensory input and executes precise movements. Cortical plasticity is also what enables the brain to understand language, adapt to a physical disability, or recover after a traumatic insult. Thus, a better understanding of the mechanisms that underlie plastic changes in the cortex could lead to broad clinical applications, such as enhancing recovery after stroke, reducing cognitive decline associated with ageing, and even facilitating sensory and motor learning in healthy adults.
| |
| Although plasticity is a general term used to describe enduring changes throughout the brain, it is becoming increasingly clear that plasticity – even when confined to one area such as the cortex – is a complex, multilayered process. Indeed, several discrete forms of plasticity have been identified during development, adulthood and following injury. Underlying these subtypes is an array of mechanisms that induce plasticity through various means. Further complicating the issue is the finding that the specific mechanisms of plasticity may vary depending on brain location, cellular layer, cell type, developmental stage, and background activity. Despite this daunting complexity, common elements of cortical plasticity have emerged, including findings that cortical connections are often strengthened following sensory and motor use, and weakened when neurons carrying sensory messages are inactive due to lack of sensory input.
| |
| Perhaps the most dynamic period for cortical reorganization is during development, a time when an onslaught of sensory information directs the brain as to which features of the environment are most important and therefore should be allocated the most processing space within the brain in the future. Conversely, features that are rarely encountered during development are often minimally represented in the brain during later stages of life. Thus, changes during development can be quite extensive, and often employ several distinct mechanisms.
| |
| Some of the most extreme levels of plasticity are seen during development following removal of sensory input. Two models of sensory deprivation that have been particularly well studied are plasticity of ocular dominance columns in the visual cortex and whisker maps in the rat barrel cortex [def: ocular dominance columns, whisker map representations]. In both cases, neuronal responses to deprived inputs are reduced soon after deprivation. This is followed by a gradual increase of responses in surrounding areas where sensory input remains, ultimately leading to an expansion of the representation of intact sensory inputs. This shift in cortical responding from deprived input to spared input can have dramatic effects, such as severe deficits in verbal comprehension if individuals are not exposed to auditory features characteristic of language early in life. However, considering one of the major goals of the brain during development is to allocate resources to stimuli likely to be encountered throughout one’s lifetime, while minimizing processing of seemingly irrelevant or unnecessary inputs, such occurrences – undesirable as they may be – are understandable consequences of development.
| |
| Both the rapid loss and gradual expansion of cortical representations following sensory deprivation appear to be mediated largely by changes within the cortex itself, albeit through different mechanisms. The initial, rapid loss of responses appears to occur by decreasing the ability of “deprived” cells located at early stages of sensory processing to excite downstream cells through a process known as long-term depression (LTD), specifically NMDA receptor-dependent and CB1-mediated LTD [def: LTD]. In addition, structural plasticity (i.e., a reduction in the number of excitatory synaptic connections between cells along the sensory processing network) and a strengthening of connections between inhibitory neurons and “deprived” neurons may also play a role in response depression. On the other hand, the process behind the slower increase in neuronal responding of spared inputs is not as straight forward and may differ substantially depending on cortical region. Within the somatosensory cortex, response potentiation is likely due to strengthening of excitatory signaling between deprived neurons and neighboring cells though a process called long-term potentiation (LTP) [def: LTP]. While LTP may also play a role in response potentiation in the visual cortex, a more likely candidate is synaptic scaling via homeostatic plasticity (see below). In this scenario, reduced input (caused by sensory deprivation) triggers compensatory increases in excitability and synapse strength throughout the deprived neuron. Since the only synapses that retain relevant activity are from neighboring neurons where sensory input is spared, deprived neurons become more responsive to sensory information carried by these cells.
| |
| Another form of reorganization that occurs during development is experience-dependent plasticity, where repeated exposure to a specific stimulus causes an increase in responding to that stimulus. In the auditory cortex, for example, tonotopic maps are often reflections of auditory experience, with highly-encountered frequencies giving rise to greater representations within the tonotopic map, presumably leading to enhanced processing of these frequencies [def: tonotopic map]. Experience-dependent strengthening of inputs such as these is predominately generated through potentiation of excitatory synapses via NMDA-dependent LTP.
| |
| For many years, cortical plasticity was believed to occur only during development*. This is understandable, as a brain in constant flux would make it difficult to retain information and skills over long periods of time without routine exposure or practice. On the other hand, a completely static adult brain would block refinement of sensory representations and motor skills, which may prove vital when immersed in new environments. Instead, retaining the ability to modify the brain in adulthood, but to a reduced degree and under limited circumstances, would allow individuals to adapt to new environments while retaining imprints from experiences past.
| |
| We now know that cortical changes can indeed take place during adulthood. However, adult reorganization is modest compared with development and is considerably harder to induce, as passive exposure to stimuli is unable to generate plastic changes in adults. Instead, only stimuli linked to reward or punishment reliably evoke plastic changes within the adult cortex, mostly in the form of increased representations of these stimuli. For example, pairing an auditory tone with electric shock causes cells in the auditory cortex to shift their receptive fields toward the frequency of the paired tone. In contrast, random, unpaired presentations of tone and shock do not alter receptive fields*.
| |
| Although the cellular mechanisms though which adult cortical plasticity takes place are largely unknown, research on skilled motor learning in the adult rat has provided some insights. For instance, following training on a forelimb reach task wherein rats learn to accurately grasp food pellets through a small aperture, LTP-like changes occur in cortical regions engaged during learning. The density of dendritic spines has also been shown to increase in neurons active during grasping movements. These findings suggest that excitatory connections between neurons controlling relevant forelimb movements are established and strengthened during learning. In addition, neuromodulators such as acetylcholine are essential for adult cortical plasticity, as animals lacking cholinergic input from the basal forebrain display impairments in both the learning and plasticity associated with forelimb reach training [def: neuromodulators].
| |
| Overall, it appears that LTP mediates the potentiation and expansion of cortical representations during both sensory and motor learning in adults. Furthermore, several lines of evidence suggest that this learning-related plasticity occurs through spike timing-dependent plasticity (STDP) [def: STDP], although the exact temporal patterns that drive LTP/LTD are unknown, and may differ depending on cell type, background network activity, and the level of neuromodulators present, among others.
| |
| Finally, a distinct form of plasticity is triggered by significant deviations from normal levels of synaptic input. This type of plasticity, known as homeostatic plasticity, uses compensatory mechanisms to return cortical activity to a set level. For example, when one eye is blocked from receiving visual input during development, the once constant stream of activity traveling from the deprived eye to the visual cortex is drastically reduced. In an attempt to restore cortical activity, manipulations of inhibitory network activity, as well as increases in the degree of excitatory input and intrinsic excitability, are seen. As mentioned earlier, the end result of such increases in cell excitability is an upregulation of input strength from active (nondeprived) neurons. Thus, in contrast to more “traditional” forms of cortical reorganization, where changes in synaptic activity and synaptic strength occur at the same location, homeostatic plasticity can alter the strength of synapses at neuronal locations distant from the site of activity modulation.
| |
|
| |
|
| Despite the wealth of knowledge accumulated over the decades, many questions remain regarding the exact mechanisms of cortical plasticity. For example, what processes direct the switch from the dynamic levels of plasticity during development, to the more tempered levels seen in adulthood? Are all cortical synapses plastic? On what timescale, and to what extent, do neuromodulators mediate plasticity? Are physiological plasticity (LTP, LTD) and structural plasticity (fluctuations in dendritic spines and axonal branching) co-regulated, or are they separable, independent entities? In the case of the latter, what are the relative contributions of each to overall plasticity?
| | README for project 'plasticity' |
| As seen above, the mechanisms surrounding plastic changes in the brain can vary widely. Out of this apparent chaos, however, have emerged several basic principles that aid our ongoing scientific and clinical pursuits. Future discoveries will undoubtedly bring us closer to harnessing the inherent plasticity of the brain for both clinical and everyday applications. Imagine treatments that enhance recovery following stroke by inducing the vast levels of plasticity usually confined to development, or perhaps taking a pill before your guitar lesson to boost acquisition of that day’s arpeggio lesson. While these scenarios may be a bit far off, they are certainly not far fetched. Despite the challenging road ahead of us, the brain is capable of mind-boggling achievements given immense human effort.
| |
|
| |
|
| Definitions:
| |
| Ocular dominance columns:
| |
| Whisker map: (barrel cortex)
| |
| Long-term depression:
| |
| Long-term potentiation:
| |
| Tonotopic map:
| |
| Neuromodulator:
| |
| Spike timing-dependent plasticity (STDP):
| |
|
| |
|
| | SECTION 1: MESH GENERATION |
|
| |
|
| | There are several ways to generate an acceptable polygon surface for particle diffusion. Essentially the surface is a tetrahedral mesh with vertices located in 3D space at x,y,z cartesian coordinates, such that each vertex-point is of data-type 'double' machine precision. |
|
| |
|
| | vertex-point example: |
| | x=1.859184868755403e+01 |
| | y=4.894117244518976e+01 |
| | z=-6.007788068750699e+01 |
|
| |
|
| | Each vertex also has a unique positive integer index number, that range from 1-N where N is the total number of vertices in the mesh. Note that since Python indexing starts at 0, index numbers actually range from 0-[N-1]; in Matlab indexing starts at 1, so keep this in mind when passing data between Python and Matlab. |
|
| |
|
| | Python's fenics/dolfin meshing package renders XML output for a single vertex as: |
|
| |
|
| | <vertex index="0" x="-5.000000000000000e+01" y="-4.619397662556434e+01" z="-3.086582838174552e+01" /> |
|
| |
|
| Within the cortex, most of these forms can be categorized as either experience-dependent or learning-related. [[definitions: experience-dependent, l-r]]
| | In a tetrahedral mesh each vertex-point is connected by straight lines to other vertices, forming a closed triangular mesh. Thus, to go along with a list of vertex points, the mesh also requires a triangulation connectivity list. This matrix contains the following information: |
|
| |
|
| | Each row represents a triangle or tetrahedron in the triangulation. |
| | Each index number is a unique ID for each triangle or tetrahedron. |
| | Each element is a vertex ID. |
|
| |
|
| In addition, there are many unique mechanisms through which plastic changes take place. For example, plasticity can arise from modification of existing synapses (strengthening or weakening), formation of new synapses, regulation of intrinsic excitability, alterations of inhibitory networks,…
| | Python's fenics/dolfin meshing package renders XML output for a single tetrahedron as: |
|
| |
|
| Among these is the finding that cortical plasticity during development is extensive, readily induced, and bidirectional, while adult plasticity is much more restricted. Whether during development or adulthood, exposure to sensory stimuli increases sensory representations in the cortex, primarily by strengthening connections between cortical neurons. Conversely, cortical connections are weakened when neurons carrying sensory messages are inactive due to lack of sensory input.
| | <tetrahedron index="0" v0="224" v1="325" v2="1334" v3="1576" /> |
|
| |
|
| | Here's a more complete example of the XML file generated using fenics/dolfin: |
|
| |
|
| Experience-dependent plasticity
| | <?xml version="1.0"?> |
| Experience-dependent plasticity refers to
| | <dolfin xmlns:dolfin="http://fenicsproject.org"> |
| | <mesh celltype="tetrahedron" dim="3"> |
| | <vertices size="1870"> |
| | <vertex index="0" x="-5.00000e+01" y="-4.61939e+01" z="-3.08658e+01" /> |
| | <vertex index="1" x="-5.00000e+01" y="-4.04219e+01" z="-2.08183e+01" /> |
| | <vertex index="2" x="-5.00000e+01" y="-3.25491e+01" z="-1.23416e+01" /> |
| | ... |
| | <vertex index="1867" x="1.35696e+02" y="3.91776e+01" z="-5.20005e+01" /> |
| | <vertex index="1868" x="2.11946e+02" y="4.14023e+01" z="-2.20125e+01" /> |
| | <vertex index="1869" x="6.98190e+00" y="1.10172e+01" z="-7.31615e+01" /> |
| | </vertices> |
| | <cells size="7759"> |
| | <tetrahedron index="0" v0="224" v1="325" v2="1334" v3="1576" /> |
| | <tetrahedron index="1" v0="1111" v1="1201" v2="1265" v3="1427" /> |
| | <tetrahedron index="2" v0="92" v1="129" v2="329" v3="1306" /> |
| | ... |
| | <tetrahedron index="7756" v0="1132" v1="1501" v2="1640" v3="1869" /> |
| | <tetrahedron index="7757" v0="1132" v1="1230" v2="1640" v3="1869" /> |
| | <tetrahedron index="7758" v0="1230" v1="1501" v2="1640" v3="1869" /> |
| | </cells> |
| | </mesh> |
| | </dolfin> |
|
| |
|
|
| |
|
| Learning-related plasticity
| |
| The defining feature of L-R plasticity is the requirement of cognitive engagement on the part of the individual.
| |
|
| |
|
| Plasticity is an elementary property of the brain whose methods and effects vary from region to region. Within the cortex,
| |
| it is responsible for modification of sensory representations, Thus, distinction made when discussing
| |
| As our comprehension advances, it is becoming increasingly clear that there exist several functionally-distinct classes of plasticity.
| |
| The term plasticity
| |
|
| |
|
| In recent years, considerable progress has been made in discovering these mechanisms using different levels of analysis (cortical maps, single unit receptive fields, neurophysiology, synaptic reconstruction, insertion/removal of receptors, phosphorylation states, etc). As the review by Feldman in this issue points out, although “plasticity” is a general term used to refer to enduring changes in the brain, it is becoming increasingly clear that there exist many unique paths to plasticity, with each being employed under select circumstances. For example, alterations during development, adulthood and following injury appear to utilize distinct mechanisms when reorganizing the brain. Therefore, it is necessary to distinguish between these different pathways when discussing specific processes of plasticity.
| |
|
| |
|
|
| |
|
| | ############################ SHELL PATH ############################ |
|
| |
|
| What are the mechanisms of plasticity?
| | ###### APPLE PYTHON 2.7 ###### |
| Before launching into the various subtypes of plasticity and their particular processes, a
| | export PATH="$PATH:/Library/Frameworks/Python.framework/Versions/2.7/bin" |
|
| |
|
| general overview of these processes is in order.
| | ###### FENICS ###### |
| | source "/Applications/FEniCS.app/Contents/Resources/share/fenics/fenics.conf" |
|
| |
|
| The term plasticity is ubiquitously used to refer to enduring changes in the brain, be it during development, adulthood, or following injury. However, it is becoming increasingly clear that there are many distinct paths to plasticity, each being employed under select circumstances. For example,
| | ###### CANOPY PYTHON (ENABLES DOLPHIN & IPYTHON) ###### |
| In addition, there exist many measures of plasticity, including alterations of cortical maps, neuronal receptive fields, axonal branching, synaptic contacts, physiological properties, etc. Although it is generally assumed that such measures are all reflections of one another, one must take care not to assume
| | export PATH="$PATH:/Users/bradleymonk/Library/Enthought/Canopy_64bit/User/bin" |
|
| |
|
|
| |
|
| Within the brain, there exist many measures of plasticity. These include alterations of cortical maps, neuronal receptive fields, axonal branching, synaptic contacts, and physiological responses. In addition, plastic changes are seen throughout all stages of life, including development, adulthood, and following injury. Given such diversity, it is of little surprise that many distinct components of plasticity have been identified. However, within the cortex, most forms of plasticity can be categorized as either experience-dependent or learning-related.
| |
|
| |
|
|
| |
|
| Line 70: |
Line 78: |
|
| |
|
|
| |
|
| Although I have yet to see a monkey earn a 10 on the balance beam routine, the level of sensory discrimination attainable in these animals is remarkable.
| |
|
| |
|
| LTD is reduction of strength…Thus, LTD is good candidate for mech of response depression following input deprivation. Indeed, shown to be.
| |
|
| |
|
| {{C}}
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | |
| | [[Category:Plasticity]] |
README for project 'plasticity'
SECTION 1: MESH GENERATION
There are several ways to generate an acceptable polygon surface for particle diffusion. Essentially the surface is a tetrahedral mesh with vertices located in 3D space at x,y,z cartesian coordinates, such that each vertex-point is of data-type 'double' machine precision.
vertex-point example:
x=1.859184868755403e+01
y=4.894117244518976e+01
z=-6.007788068750699e+01
Each vertex also has a unique positive integer index number, that range from 1-N where N is the total number of vertices in the mesh. Note that since Python indexing starts at 0, index numbers actually range from 0-[N-1]; in Matlab indexing starts at 1, so keep this in mind when passing data between Python and Matlab.
Python's fenics/dolfin meshing package renders XML output for a single vertex as:
<vertex index="0" x="-5.000000000000000e+01" y="-4.619397662556434e+01" z="-3.086582838174552e+01" />
In a tetrahedral mesh each vertex-point is connected by straight lines to other vertices, forming a closed triangular mesh. Thus, to go along with a list of vertex points, the mesh also requires a triangulation connectivity list. This matrix contains the following information:
Each row represents a triangle or tetrahedron in the triangulation.
Each index number is a unique ID for each triangle or tetrahedron.
Each element is a vertex ID.
Python's fenics/dolfin meshing package renders XML output for a single tetrahedron as:
<tetrahedron index="0" v0="224" v1="325" v2="1334" v3="1576" />
Here's a more complete example of the XML file generated using fenics/dolfin:
<?xml version="1.0"?>
<dolfin xmlns:dolfin="http://fenicsproject.org">
<mesh celltype="tetrahedron" dim="3">
<vertices size="1870">
<vertex index="0" x="-5.00000e+01" y="-4.61939e+01" z="-3.08658e+01" />
<vertex index="1" x="-5.00000e+01" y="-4.04219e+01" z="-2.08183e+01" />
<vertex index="2" x="-5.00000e+01" y="-3.25491e+01" z="-1.23416e+01" />
...
<vertex index="1867" x="1.35696e+02" y="3.91776e+01" z="-5.20005e+01" />
<vertex index="1868" x="2.11946e+02" y="4.14023e+01" z="-2.20125e+01" />
<vertex index="1869" x="6.98190e+00" y="1.10172e+01" z="-7.31615e+01" />
</vertices>
<cells size="7759">
<tetrahedron index="0" v0="224" v1="325" v2="1334" v3="1576" />
<tetrahedron index="1" v0="1111" v1="1201" v2="1265" v3="1427" />
<tetrahedron index="2" v0="92" v1="129" v2="329" v3="1306" />
...
<tetrahedron index="7756" v0="1132" v1="1501" v2="1640" v3="1869" />
<tetrahedron index="7757" v0="1132" v1="1230" v2="1640" v3="1869" />
<tetrahedron index="7758" v0="1230" v1="1501" v2="1640" v3="1869" />
</cells>
</mesh>
</dolfin>
- SHELL PATH ############################
- APPLE PYTHON 2.7 ######
export PATH="$PATH:/Library/Frameworks/Python.framework/Versions/2.7/bin"
- FENICS ######
source "/Applications/FEniCS.app/Contents/Resources/share/fenics/fenics.conf"
- CANOPY PYTHON (ENABLES DOLPHIN & IPYTHON) ######
export PATH="$PATH:/Users/bradleymonk/Library/Enthought/Canopy_64bit/User/bin"