Limits...
Synaptic scaling enables dynamically distinct short- and long-term memory formation.

Tetzlaff C, Kolodziejski C, Timme M, Tsodyks M, Wörgötter F - PLoS Comput. Biol. (2013)

Bottom Line: How time scale integration and synaptic differentiation is simultaneously achieved remains unclear.The interaction between plasticity and scaling provides also an explanation for an established paradox where memory consolidation critically depends on the exact order of learning and recall.These results indicate that scaling may be fundamental for stabilizing memories, providing a dynamic link between early and late memory formation processes.

View Article: PubMed Central - PubMed

Affiliation: Faculty of Physics - Biophysics, Georg August University Friedrich-Hund Platz 1, Göttingen, Germany ; Network Dynamics Group, Max Planck Institute for Dynamics and Self-Organization, Göttingen, Germany ; Bernstein Center for Computational Neuroscience, Georg-August-University Friedrich-Hund Platz 1, Göttingen, Germany.

ABSTRACT
Memory storage in the brain relies on mechanisms acting on time scales from minutes, for long-term synaptic potentiation, to days, for memory consolidation. During such processes, neural circuits distinguish synapses relevant for forming a long-term storage, which are consolidated, from synapses of short-term storage, which fade. How time scale integration and synaptic differentiation is simultaneously achieved remains unclear. Here we show that synaptic scaling - a slow process usually associated with the maintenance of activity homeostasis - combined with synaptic plasticity may simultaneously achieve both, thereby providing a natural separation of short- from long-term storage. The interaction between plasticity and scaling provides also an explanation for an established paradox where memory consolidation critically depends on the exact order of learning and recall. These results indicate that scaling may be fundamental for stabilizing memories, providing a dynamic link between early and late memory formation processes.

Show MeSH

Related in: MedlinePlus

Robustness of bifurcation structure.(A) : The  vs.  function of the fixed points of the system as already shown in Figure 3 A. For simplicity we show here only the curve without indicating the different storage domains. (B) : Changing, for instance, the desired firing rate parameter of the synaptic scaling term does not induce significant changes in the  vs.  function. The overall circuit dynamics are the same as shown in Figure 1 (see Figure S3 in Text S1). This holds for negative  values (not shown), too. (C) : Only a dramatically different  value induces changes in system's dynamic. Here, a pole emerges for small input intensities. To avoid this pole and maintain the desired dynamic the background input could be increased () to keep the system on the right side of the pole. Alternatively, other parameters could be adapted. For instance, (D1) the steepness of the neuronal output function () or (D2) the inflexion point () have to be decreased.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC3814677&req=5

pcbi-1003307-g004: Robustness of bifurcation structure.(A) : The vs. function of the fixed points of the system as already shown in Figure 3 A. For simplicity we show here only the curve without indicating the different storage domains. (B) : Changing, for instance, the desired firing rate parameter of the synaptic scaling term does not induce significant changes in the vs. function. The overall circuit dynamics are the same as shown in Figure 1 (see Figure S3 in Text S1). This holds for negative values (not shown), too. (C) : Only a dramatically different value induces changes in system's dynamic. Here, a pole emerges for small input intensities. To avoid this pole and maintain the desired dynamic the background input could be increased () to keep the system on the right side of the pole. Alternatively, other parameters could be adapted. For instance, (D1) the steepness of the neuronal output function () or (D2) the inflexion point () have to be decreased.

Mentions: Specifically, we find a saddle node bifurcation where different fixed points are reached for low as compared to high input intensities. For the particular setting displayed in Figure 3, a continuous regime of fixed points for the weights exists for firing rates below approximately (Short-Term Storage, STS; green, Figure 3 A), while above this frequency, the system jumps to a fixed point regime with substantially larger weights (Long-Term Storage, LTS; red, Figure 3 A). The gray area below STS represents the range of weights found for the randomly stimulated control neurons (targets of the yellow neurons in Figure 1 A). Note, to obtain this curve we assumed that the circuit consists of several roughly independent subnetworks. This means that in one circuit different fixed points are reached at different spatial locations. For instance, in Figure 1 C after local stimulation the (local) patch is in the LTS-regime (about in Figure 3 A) while the control units are weakly stimulated and, therefore, they are in the gray control regime (about ) with small synaptic weights. The bifurcation is essential for the dynamics discussed here. Using different parameter values for the system does not change the fixed point curve significantly (see, e.g., Figure 4 B and Figure S3 in Text S1 compared to the used setting shown in Figure 4 A and Figure 1 B,C). However, if one parameter is changed dramatically an adequate adaption of the other parameters can still guarantee the desired circuit dynamics (see Figure 4 C,D). Thereby, the range of parameters remains in a physiological regime.


Synaptic scaling enables dynamically distinct short- and long-term memory formation.

Tetzlaff C, Kolodziejski C, Timme M, Tsodyks M, Wörgötter F - PLoS Comput. Biol. (2013)

Robustness of bifurcation structure.(A) : The  vs.  function of the fixed points of the system as already shown in Figure 3 A. For simplicity we show here only the curve without indicating the different storage domains. (B) : Changing, for instance, the desired firing rate parameter of the synaptic scaling term does not induce significant changes in the  vs.  function. The overall circuit dynamics are the same as shown in Figure 1 (see Figure S3 in Text S1). This holds for negative  values (not shown), too. (C) : Only a dramatically different  value induces changes in system's dynamic. Here, a pole emerges for small input intensities. To avoid this pole and maintain the desired dynamic the background input could be increased () to keep the system on the right side of the pole. Alternatively, other parameters could be adapted. For instance, (D1) the steepness of the neuronal output function () or (D2) the inflexion point () have to be decreased.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC3814677&req=5

pcbi-1003307-g004: Robustness of bifurcation structure.(A) : The vs. function of the fixed points of the system as already shown in Figure 3 A. For simplicity we show here only the curve without indicating the different storage domains. (B) : Changing, for instance, the desired firing rate parameter of the synaptic scaling term does not induce significant changes in the vs. function. The overall circuit dynamics are the same as shown in Figure 1 (see Figure S3 in Text S1). This holds for negative values (not shown), too. (C) : Only a dramatically different value induces changes in system's dynamic. Here, a pole emerges for small input intensities. To avoid this pole and maintain the desired dynamic the background input could be increased () to keep the system on the right side of the pole. Alternatively, other parameters could be adapted. For instance, (D1) the steepness of the neuronal output function () or (D2) the inflexion point () have to be decreased.
Mentions: Specifically, we find a saddle node bifurcation where different fixed points are reached for low as compared to high input intensities. For the particular setting displayed in Figure 3, a continuous regime of fixed points for the weights exists for firing rates below approximately (Short-Term Storage, STS; green, Figure 3 A), while above this frequency, the system jumps to a fixed point regime with substantially larger weights (Long-Term Storage, LTS; red, Figure 3 A). The gray area below STS represents the range of weights found for the randomly stimulated control neurons (targets of the yellow neurons in Figure 1 A). Note, to obtain this curve we assumed that the circuit consists of several roughly independent subnetworks. This means that in one circuit different fixed points are reached at different spatial locations. For instance, in Figure 1 C after local stimulation the (local) patch is in the LTS-regime (about in Figure 3 A) while the control units are weakly stimulated and, therefore, they are in the gray control regime (about ) with small synaptic weights. The bifurcation is essential for the dynamics discussed here. Using different parameter values for the system does not change the fixed point curve significantly (see, e.g., Figure 4 B and Figure S3 in Text S1 compared to the used setting shown in Figure 4 A and Figure 1 B,C). However, if one parameter is changed dramatically an adequate adaption of the other parameters can still guarantee the desired circuit dynamics (see Figure 4 C,D). Thereby, the range of parameters remains in a physiological regime.

Bottom Line: How time scale integration and synaptic differentiation is simultaneously achieved remains unclear.The interaction between plasticity and scaling provides also an explanation for an established paradox where memory consolidation critically depends on the exact order of learning and recall.These results indicate that scaling may be fundamental for stabilizing memories, providing a dynamic link between early and late memory formation processes.

View Article: PubMed Central - PubMed

Affiliation: Faculty of Physics - Biophysics, Georg August University Friedrich-Hund Platz 1, Göttingen, Germany ; Network Dynamics Group, Max Planck Institute for Dynamics and Self-Organization, Göttingen, Germany ; Bernstein Center for Computational Neuroscience, Georg-August-University Friedrich-Hund Platz 1, Göttingen, Germany.

ABSTRACT
Memory storage in the brain relies on mechanisms acting on time scales from minutes, for long-term synaptic potentiation, to days, for memory consolidation. During such processes, neural circuits distinguish synapses relevant for forming a long-term storage, which are consolidated, from synapses of short-term storage, which fade. How time scale integration and synaptic differentiation is simultaneously achieved remains unclear. Here we show that synaptic scaling - a slow process usually associated with the maintenance of activity homeostasis - combined with synaptic plasticity may simultaneously achieve both, thereby providing a natural separation of short- from long-term storage. The interaction between plasticity and scaling provides also an explanation for an established paradox where memory consolidation critically depends on the exact order of learning and recall. These results indicate that scaling may be fundamental for stabilizing memories, providing a dynamic link between early and late memory formation processes.

Show MeSH
Related in: MedlinePlus