ocrampal/organism
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Update 2026-06-02-astrocyte-behaviors.md

Browse Source
This commit is contained in:
ocrampal
2026-06-02 16:28:43 +02:00
parent e89465437a
commit 2415641b65
1 changed files with 107 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files
elements/astrocyte/appunti/2026-06-02-astrocyte-behaviors.md
+107 -2
View File
@@ -44,11 +44,10 @@ Here is the explicit breakdown of all three modes so you can map them clearly in
---
### The Input Gating Architecture for Your Model
## The Input Gating Architecture for Your Model
When building your simulation, your input variable (Neuronal Activity) feeds into a nested logic structure:
```
IF Activity == Low-to-Moderate (1-10 Hz)
──> Engage MODE 1 (Local Housekeeping)
@@ -60,3 +59,109 @@ IF Activity == Massive/Synchronous (>100 Hz, multi-pathway)
└─► IF MODE 1 or MODE 3 persists for hours/days
──> Engage MODE 4 (Global Structural Homeostatic Scaling)
---
---
Yes. Dissoridating the two locks—displacing magnesium without D-serine, or providing D-serine without displacing magnesium—happens naturally in several physiological scenarios.
These mismatches are highly useful for your model because they represent states where the network is trying to perform a computation but the astrocyte or the post-synapse acts as a gatekeeper to alter the outcome.
---
## Situation A: Magnesium Displaced, but NO Astrocytic D-Serine
In this scenario, the electrical lock is open, but the chemical lock is closed.
* **When this happens:** **Low-Frequency Backpropagation or Artificial Depolarization.**
* Imagine the postsynaptic neuron fires an action potential from its cell body that travels *backward* up its own dendrites (a backpropagating action potential), or an experimenter forces the postsynaptic cell to depolarize using a patch-pipette, **while the presynapse is completely silent**.
* **The Biophysics:** The strong voltage shift changes the electrical charge inside the postsynaptic membrane, repelling and **displacing the $Mg^{2+}$ plug**. However, because the presynapse didn't fire, there is no glutamate to trigger the astrocyte’s microdomain calcium surge. Consequently, **no D-serine is released**.
* **The Functional Outcome:** **Silent Displacements / No Plasticity.** Even though the physical channel is technically unblocked, the NMDA receptor cannot pass calcium current because its co-agonist site is empty. The induction of LTP is blocked.
---
## Situation B: Astrocytic D-Serine Released, but Magnesium NOT Displaced
In this scenario, the chemical lock is open, but the electrical lock is closed.
* **When this happens:** **Low-Frequency Stimulation (LFS) / Long-Term Depression (LTD).**
* This happens when a presynaptic pathway fires slowly and persistently (e.g., $\sim$ 1 Hz for several minutes).
* **The Biophysics:** The slow trickle of glutamate is not strong enough to cause the summation needed to highly activate AMPA receptors. The postsynapse remains near its resting potential (around $-70\text{ mV}$), meaning **the $Mg^{2+}$ plug stays firmly inside the channel**. However, over minutes, this steady ambient glutamate leaks out of the cleft and steadily climbs into the astrocyte process, triggering slow, rhythmic calcium oscillations that cause the astrocyte to **release D-serine**.
* **The Functional Outcome:** **Long-Term Depression (LTD).** Because the $Mg^{2+}$ plug is still present, it blocks 95% of the channel. However, because D-serine is bound, whenever a tiny fluctuation in voltage slightly loosens the magnesium plug, a *minute* trickle of calcium escapes into the postsynapse.
* In neuroscience, a **massive, fast** calcium influx triggers **LTP** (learning).
* A **tiny, prolonged** calcium influx triggers **LTD** (unlearning/weakening).
* Therefore, having D-serine present *without* fully displacing magnesium is the exact biophysical state required to downscale or weaken a synapse.
---
## Summary for Your System Coding
You can think of these combinations as a matrix of states that determine the direction of your weight update rule ($\Delta W$):
| $Mg^{2+}$ Status | Astrocytic D-Serine Status | Resulting Synaptic State | Computational Result |
| --- | --- | --- | --- |
| **Blocked** (Resting V) | **Absent** (No Astro Ca) | Baseline Transmission | Stable Synapse ($W$ unchanged) |
| **Unblocked** (Depolarized) | **Absent** (No Astro Ca) | Non-Functional Opening | Gated Block ($W$ unchanged) |
| **Blocked** (Resting V) | **Present** (Slow Astro Ca) | Minimal $Ca^{2+}$ Leak | **LTD Induction** ($W$ decreases) |
| **Unblocked** (Depolarized) | **Present** (Burst Astro Ca) | Massive $Ca^{2+}$ Influx | **LTP Induction** ($W$ increases) |
By implementing these conditions, your model will accurately replicate how the timing between the neuron's electrical state and the astrocyte's chemical state determines whether a memory is created, deleted, or ignored.
---
---
The transition from temporary synaptic strengthening to permanent, astrocytically consolidated structural change (**Late-LTP**) acts as a critical filter in the brain. It ensures the brain only burns energy to store information that is deemed genuinely important.
In your computational model, whether an astrocyte decides to structurally consolidate a synapse or let it decay back to baseline is influenced by four distinct physiological factors.
---
## 1. Volume and Duration of Postsynaptic Signals (The Local Threshold)
The astrocyte does not automatically know that a synapse has opened its NMDA receptors. It relies on the postsynapse to send an active "request for consolidation" via retrograde chemical factors.
* **The Signals:** **BDNF** (Brain-Derived Neurotrophic Factor) and **Nitric Oxide (NO)**.
* **How it influences consolidation:** * **Lack of Consolidation:** If the NMDA receptors open briefly (e.g., during a single, short burst of activity), the postsynapse inserts a few temporary AMPA receptors but releases very little BDNF. The neighboring astrocyte process never detects enough BDNF to trigger its internal protein-synthesis machinery. The structural lock remains open, and the synapse decays back to baseline within an hour.
* **Successful Consolidation:** If the synapse experiences repeated, spaced bursts of high-frequency activity, the sustained calcium influx forces the postsynapse to secrete a massive wave of BDNF. This high concentration overpowers the astrocyte's receptor thresholds, initiating structural consolidation.
---
## 2. Global Astrocytic Calcium Scale (The Cooperativity Factor)
As we established in the multi-compartment architecture, the spatial scale of calcium ($Ca^{2+}$) within the astrocyte plays a massive role in dictating its long-term structural behavior.
* **How it influences consolidation:**
* **Lack of Consolidation (Single Synapse Isolation):** If only one single, isolated synapse on a dendritic branch is active, the calcium rise is confined to a tiny nanoscale microdomain ($Ca^{2+}_{\text{micro}}$). While this is enough to release D-serine for *induction*, a single microdomain often lacks the metabolic clout to trigger whole-cell protein synthesis.
* **Successful Consolidation (Synaptic Cooperativity):** If multiple neighboring synapses on the same dendritic branch fire together, their individual signals summate. This triggers a localized, high-amplitude **calcium hotspot** or a branch-specific calcium wave. This larger spatial signal activates astrocytic calcineurin and transcription factors (like NFAT), prompting the astrocyte to synthesize and secrete the structural matrix proteins (**Glypicans** and **Thrombospondins**) needed to anchor those new AMPA receptors long-term.
---
## 3. Neuromodulatory Input (The "Attention" or "Value" Switch)
The brain uses neuromodulators to stamp an event as "important." Astrocytes are highly sensitive to these chemicals because they possess vast arrays of metabotropic receptors for them.
* **The Signals:** **Norepinephrine** (released during stress/alertness), **Dopamine** (released during reward), and **Acetylcholine** (released during intense focus).
* **How it influences consolidation:**
* Even if a neuronal pathway fires perfectly to open the NMDA gate, the absence of neuromodulators often results in a failure of late-stage consolidation.
* Conversely, when you are highly alert or rewarded, a flood of norepinephrine or dopamine binds to astrocytic GPCRs (like $\alpha_1$-adrenergic or $D_1$-dopamine receptors). This triggers a massive release of calcium from the astrocyte's internal endoplasmic reticulum stores. This neuromodulator-enforced calcium surge forces the astrocyte to rapidly consolidate whatever local synapses happen to be active at that moment, transforming weak early-LTP into permanent late-LTP.
---
## 4. Extracellular Protease Activity (The Structural Eraser)
Even after the astrocyte releases matrix proteins to stabilize a synapse, consolidation can be actively undone or blocked by enzymatic degradation.
* **The Signals:** **MMPs (Matrix Metalloproteinases)** and **tPA (tissue Plasminogen Activator)**.
* **How it influences consolidation:**
* Astrocytes and neurons co-secrete these proteases into the extracellular space. MMPs act like molecular scissors that chew up the extracellular matrix and the structural anchors (like Glypicans) provided by the astrocyte.
* If a synapse is not continuously reinforced, or if a competing pathway nearby becomes highly active, local MMP activity spikes. This degrades the astrocytic scaffold, causing the clustered AMPA receptors to lose their anchor points and diffuse away. This results in a failure of consolidation or an active reversal of LTP (de-potentiation).