319 lines
8.8 KiB
Markdown
319 lines
8.8 KiB
Markdown
# **ATP-Lactate Dynamics: The Ultimate Constraint System**
|
||
|
||
## Non modello ATP production al SOMA
|
||
|
||
### **The Dual-Source ATP System**
|
||
|
||
Neurons have **two complementary ATP production systems** that operate at different timescales:
|
||
|
||
#### **1. LOCAL ATP PRODUCTION (Presynaptic Terminal)**
|
||
|
||
**Primary Source:** **Mitochondria** within the terminal
|
||
**Capacity:** Limited (often 1-5 mitochondria per terminal)
|
||
**Timescale:** Seconds to minutes
|
||
|
||
#### **2. DISTAL ATP PRODUCTION (Soma & Axon)**
|
||
|
||
**Primary Source:** **Somatic mitochondria** (more abundant)
|
||
**Transport:** Axonal transport of ATP or ATP precursors
|
||
**Timescale:** Hours (slow transport)
|
||
|
||
You've identified the **fundamental constraint hierarchy**:
|
||
|
||
```
|
||
ATP availability ← Lactate production ← Astrocyte coordination ← Network activity
|
||
```
|
||
|
||
This creates a **energy-based modulation cascade** that constrains everything else.
|
||
|
||
## **The Energy Constraint Hierarchy**
|
||
|
||
### **Level 1: Millisecond ATP Consumption (Ground Reality)**
|
||
|
||
```
|
||
Processes consuming ATP:
|
||
- Na⁺/K⁺ pump (40-60% of ATP per AP)
|
||
- Ca²⁺ pumps (PMCA: ~1 ATP per Ca²⁺)
|
||
- Vesicle priming (SNARE assembly, NSF ATPase)
|
||
- Neurotransmitter reloading (v-ATPase)
|
||
|
||
ATP consumption per AP: ~20,000 ATP molecules
|
||
Terminal ATP store: ~10⁸ ATP molecules → ~500 APs until depletion
|
||
```
|
||
|
||
### **Level 2: Second-Minute Lactate Supply (Immediate Energy Politics)**
|
||
|
||
```
|
||
Astrocyte lactate production:
|
||
- Glutamate uptake → Na⁺ influx → Glycolysis → Lactate
|
||
- Glycogen breakdown → Lactate
|
||
- One astrocyte serves ~100,000 synapses
|
||
|
||
Lactate transport:
|
||
- MCT2 transporters on presynaptic terminal
|
||
- Conversion: Lactate → Pyruvate → ~15 ATP via TCA cycle
|
||
- Timescale: Seconds for transport, minutes for metabolism
|
||
```
|
||
|
||
### **Level 3: Minute-Hour Network Competition (Energy Economics)**
|
||
|
||
```
|
||
Shared resource problem:
|
||
- Multiple synapses compete for astrocyte lactate
|
||
- Active synapses get priority (activity-dependent coupling)
|
||
- "Energy-rich get richer" feedback
|
||
|
||
Astrocyte decision:
|
||
IF (synapse active AND lactate available) → Supply
|
||
IF (synapse inactive OR lactate limited) → Reduce supply
|
||
```
|
||
|
||
### **Level 4: Hour-Day Metabolic Adaptation (Energy Infrastructure)**
|
||
|
||
```
|
||
Long-term investments:
|
||
- More mitochondria at active synapses
|
||
- Enhanced MCT transporter expression
|
||
- Astrocyte process extension toward active synapses
|
||
```
|
||
|
||
## **ATP as the Universal Modulator**
|
||
|
||
### **ATP Availability Gates ALL Processes:**
|
||
|
||
```
|
||
IF ATP > threshold_X:
|
||
Process_Y allowed
|
||
ELSE:
|
||
Process_Y inhibited or delayed
|
||
```
|
||
|
||
### **Specific ATP Thresholds:**
|
||
|
||
```
|
||
1. High ATP (>80% of max):
|
||
- All processes operational
|
||
- Structural growth allowed
|
||
- High release probability maintained
|
||
|
||
2. Medium ATP (30-80%):
|
||
- Core release functions maintained
|
||
- Energy-intensive processes limited
|
||
- No structural growth
|
||
|
||
3. Low ATP (<30%):
|
||
- Release probability decreases
|
||
- Ca²⁺ clearance impaired
|
||
- Vesicle recycling slows
|
||
- Emergency conservation mode
|
||
```
|
||
|
||
## **Simplified ATP-Lactate Model**
|
||
|
||
### **Variables:**
|
||
|
||
```
|
||
1. ATP(t): Energy currency at presynapse
|
||
2. Lactate_ext(t): Extracellular lactate from astrocyte
|
||
3. Activity_level(t): Recent firing rate
|
||
4. Neighbor_activity(t): Activity of nearby synapses
|
||
```
|
||
|
||
### **Dynamics:**
|
||
|
||
```
|
||
d(ATP)/dt = Production - Consumption
|
||
|
||
Production = k_prod × Lactate_ext × (1 - ATP/ATP_max)
|
||
Consumption = k_cons × Activity_level + k_baseline
|
||
|
||
d(Lactate_ext)/dt = Supply - Uptake - Diffusion
|
||
|
||
Supply = k_supply × (Activity_level + α × Neighbor_activity)
|
||
Uptake = k_uptake × ATP_deficit × Lactate_ext
|
||
Diffusion = k_diff × (Lactate_ext - Lactate_background)
|
||
```
|
||
|
||
### **The Constraint Equations:**
|
||
|
||
```
|
||
For any process X with ATP requirement R_X:
|
||
|
||
IF (ATP > R_X) THEN Process_X proceeds at normal rate
|
||
ELSE Process_X rate = normal_rate × (ATP/R_X)
|
||
```
|
||
|
||
## **The Critical Insight: Energy-Based Competition**
|
||
|
||
### **Within a Single Presynapse:**
|
||
|
||
```
|
||
Processes compete for ATP:
|
||
- Release vs Clearance vs Recycling vs Growth
|
||
|
||
Energy allocation strategy:
|
||
1. Maintenance first (pumps, basic functions)
|
||
2. Release second (core mission)
|
||
3. Recycling third (future capacity)
|
||
4. Growth last (long-term investment)
|
||
|
||
During ATP shortage: Growth → Recycling → Release → Maintenance
|
||
```
|
||
|
||
### **Between Synapses (via Astrocyte):**
|
||
|
||
```
|
||
Synapses compete for lactate:
|
||
- More active synapses → More glutamate uptake → More lactate production
|
||
- But: Astrocyte lactate production limited by glucose/glycogen
|
||
- And: Lactate diffusion favors nearby synapses
|
||
|
||
Result: Local "energy hotspots" and "energy deserts"
|
||
```
|
||
|
||
## **Modeling Recommendations**
|
||
|
||
### **Option A: Simple ATP Buffer Model**
|
||
|
||
```
|
||
ATP_level = ATP_max × (1 - exp(-t/τ_replenish)) during rest
|
||
ATP_consumed_per_AP = constant
|
||
IF ATP_level < threshold: Scale down all energy-intensive processes
|
||
```
|
||
|
||
### **Option B: Lactate-Limited Model**
|
||
|
||
```
|
||
ATP_production_rate = f(Lactate_available)
|
||
Lactate_available = g(Astrocyte_response, Neighbor_activity)
|
||
Astrocyte_response = h(Glutamate_uptake, Glycogen_level)
|
||
```
|
||
|
||
### **Option C: Full Energy Competition Model**
|
||
|
||
```
|
||
For each synapse i:
|
||
dATP_i/dt = Production_i - Consumption_i
|
||
Production_i = f(Lactate_i, Mitochondria_i)
|
||
Lactate_i = Shared_pool × (Activity_i / ΣActivity_j)
|
||
Shared_pool = Astrocyte_output - Total_uptake
|
||
```
|
||
|
||
## **Implications for Plasticity**
|
||
|
||
### **The "Energy Check" for Structural Changes:**
|
||
|
||
```
|
||
BDNF says: "Grow this synapse!"
|
||
Energy system checks:
|
||
1. Current ATP level?
|
||
2. Lactate supply reliability?
|
||
3. Competing energy demands?
|
||
|
||
Only if energy sufficient: Growth proceeds
|
||
If energy marginal: Partial growth or delay
|
||
If energy insufficient: Growth blocked
|
||
```
|
||
|
||
### **Metaplasticity Rule:**
|
||
|
||
```
|
||
Synapses that consistently receive:
|
||
- High activity + Adequate lactate → Strengthen (LTP, growth)
|
||
- High activity + Inadequate lactate → Weaken (LTD, shrinkage)
|
||
- Low activity + Adequate lactate → Maintain
|
||
- Low activity + Inadequate lactate → Weaken faster
|
||
```
|
||
|
||
## **Experimental Predictions**
|
||
|
||
### **From This Model:**
|
||
|
||
1. **Lactate deprivation** should cause:
|
||
- Faster synaptic depression during trains
|
||
- Reduced long-term potentiation
|
||
- Impaired structural plasticity
|
||
2. **Astrocyte uncoupling** should cause:
|
||
- Uncoordinated synaptic fatigue
|
||
- Winner-take-all dynamics between nearby synapses
|
||
- Impaired network synchronization
|
||
3. **Metabolic stress** should cause:
|
||
- Prioritization of inhibitory synapses (energy conservation)
|
||
- Homeostatic downscaling of excitatory synapses
|
||
- Altered metaplasticity thresholds
|
||
|
||
## **Simplest Implementation for Your Model**
|
||
|
||
### **Add These Variables:**
|
||
|
||
```python
|
||
ATP = 1.0 # Normalized ATP level (0 to 1)
|
||
lactate = 0.5 # Normalized lactate availability (0 to 1)
|
||
neighbor_activity = 0.0 # Average activity of nearby synapses
|
||
```
|
||
|
||
### **Update Rules:**
|
||
|
||
```python
|
||
def update_energy(dt, activity, neighbor_avg_activity):
|
||
# Lactate supply depends on local and neighbor activity
|
||
lactate_supply = 0.1 * (activity + 0.3 * neighbor_avg_activity)
|
||
lactate = min(1.0, lactate + lactate_supply * dt)
|
||
|
||
# ATP production from lactate
|
||
ATP_production = 0.05 * lactate * (1 - ATP)
|
||
|
||
# ATP consumption from activity
|
||
ATP_consumption = 0.02 * activity
|
||
|
||
ATP = max(0, ATP + (ATP_production - ATP_consumption) * dt)
|
||
|
||
# Constraint factor for other processes
|
||
energy_factor = ATP**2 # Square to make low ATP more restrictive
|
||
|
||
return ATP, lactate, energy_factor
|
||
```
|
||
|
||
### **Apply Energy Constraints:**
|
||
|
||
```python
|
||
def process_AP_with_energy(Ca_microdomain, RRP_count, ATP):
|
||
if Ca_microdomain > threshold and RRP_count > 0:
|
||
# Energy limits release probability
|
||
P_release_effective = P_release * ATP
|
||
|
||
# Energy limits replenishment rate
|
||
k_replenish_effective = k_replenish * ATP
|
||
|
||
# Calculate release...
|
||
```
|
||
|
||
## **The Ultimate Insight**
|
||
|
||
**ATP is the ultimate vote in the presynaptic democracy.** It says:
|
||
|
||
```
|
||
"I'm the universal currency. Every process needs me.
|
||
My availability determines what's possible.
|
||
My production depends on lactate.
|
||
Lactate depends on astrocyte coordination.
|
||
Astrocytes coordinate based on network activity.
|
||
|
||
Thus, network-level activity patterns -> Astrocyte lactate distribution ->
|
||
Presynaptic ATP levels -> Constrain all release decisions and plasticity."
|
||
```
|
||
|
||
This creates a beautiful **circular causality**:
|
||
|
||
```
|
||
Release events → Glutamate → Astrocyte activation → Lactate → ATP →
|
||
Enable more release events → ...
|
||
|
||
BUT with constraints:
|
||
- Limited astrocyte capacity
|
||
- Competition between synapses
|
||
- ATP consumption rates
|
||
- Lactate diffusion limits
|
||
```
|
||
|
||
Your model now has a **physical grounding** in energy metabolism, which explains why synapses can't just "decide" to be strong - they need the **energy infrastructure** to support that strength. This makes the model both more realistic and more constrained, which paradoxically makes it more powerful for understanding real synaptic function. |