organism/neuron/appunti/behavior-PRE.md

Detailed Temporal Dynamics of Presynaptic Neurotransmitter Release

With concentration changes, calcium clearance mechanisms, and multi-timescale modifications of release probability (Pr)

---

Baseline State (Resting Condition)

Time: Continuous resting state
Presynaptic [Ca²⁺]: ~100 nM (resting cytosolic concentration)
Pr baseline: 0.3 (30% release probability per action potential)
Ready vesicles: 5-10 vesicles in readily releasable pool (RRP)
ATP/GTP levels: Normal
Phosphorylation state: Baseline kinase/phosphatase balance

---

PHASE 1: FAST TIMESCALE (0-100 ms) - SINGLE SPIKE CYCLE

0.0-0.5 ms: Action Potential Arrival & Calcium Influx

AP depolarization (+30mV) reaches terminal
    ↓
VGCCs (P/Q-type) open with ~50% probability
    ↓
Ca²⁺ enters through ~20 channels per active zone
    ↓
**Local [Ca²⁺] at release site: 100 nM → 10-50 µM** (peak in nanodomain)
    ↓
Calcium diffuses ~20-30 nm to docked vesicles
    ↓
**PROBABILISTIC DECISION POINT:**
Time window for decision: ~0.2-0.5 ms

0.5-2.0 ms: Release Decision & Execution

├── **PATH A: RELEASE OCCURS (Pr = 0.3)**
│       ↓
│   [Ca²⁺] at Synaptotagmin > 10 µM
│       ↓
│   3-5 Ca²⁺ ions bind to C2 domains of Synaptotagmin-1
│       ↓
│   Synaptotagmin inserts into membrane (Kd ~5-10 µM)
│       ↓
│   **SNARE complex completes zippering** (t ≈ 0.8 ms)
│       ↓
│   Fusion pore opens (diameter ~1 nm initially)
│       ↓
│   **~5000 glutamate molecules released** (t = 1-2 ms)
│       ↓
│   Fusion pore expands → full fusion
│       ↓
│   Vesicle membrane incorporated into plasma membrane
│
└── **PATH B: NO RELEASE (1-Pr = 0.7)**
        ↓
    [Ca²⁺] at Synaptotagmin < 5 µM (insufficient binding)
        ↓
    Calcium buffers (calbindin, parvalbumin) bind Ca²⁺
        ↓
    Vesicle remains docked/primed
        ↓
    No fusion → **silent spike**

2.0-50 ms: Calcium Clearance & Fast Recovery

**Primary clearance mechanisms:**
1. Plasma Membrane Ca²⁺ ATPase (PMCA): 
   - Kd ~100-200 nM
   - Rate: 30 Ca²⁺/sec per pump
   - **Clears 90% of Ca²⁺ in 10-20 ms**

2. Na⁺/Ca²⁺ exchanger (NCX):
   - Lower affinity (Kd ~1 µM) but higher capacity
   - Important for bulk clearance

3. Mitochondrial uptake:
   - MCU (mitochondrial Ca²⁺ uniporter)
   - Kd ~10-20 µM
   - Slower but provides long-term buffering

4. Endoplasmic reticulum uptake (SERCA):
   - Sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase
   - Kd ~0.5 µM

**Result:** [Ca²⁺] returns to ~500 nM by 50 ms
**Residual [Ca²⁺]:** ~200-300 nM persists for 100-500 ms
**Vesicle retrieval:** Clathrin-mediated endocytosis begins at ~1 sec

---

PHASE 2: SHORT-TERM PLASTICITY (100 ms - 10 sec)

Example: Spike Train at 50 Hz (20 ms interval)

Spike 1: Pr = 0.3 → Release probability
Spike 2 (20 ms later): 
   Residual [Ca²⁺] = 300 nM
   Pr increases to 0.45 (facilitation)
   Ca²⁺ influx adds to residual Ca²⁺
   
Spike 3 (40 ms):
   Residual [Ca²⁺] accumulates to 400 nM
   Pr = 0.55
   But RRP depletion begins (STD component)
   
Spike 4 (60 ms):
   RRP depleted to 60%
   Effective Pr = 0.5 × 0.6 = 0.3 (balance facilitation/depletion)
   
Spike 5-10:
   Steady-state: Pr ~0.25, RRP ~40% of baseline

Molecular Mechanisms of Short-Term Changes:

**Facilitation (0-500 ms):**
Residual Ca²⁺ (~200-500 nM) → binds to Calmodulin
    ↓
Ca²⁺-Calmodulin binds to Munc13
    ↓
Munc13 increases priming rate 3-5×
    ↓
**Pr increases for next spike**

**Depression (0-2 sec):**
Vesicle fusion → RRP depletion
    ↓
Recovery requires:
1. Vesicle recycling (endocytosis: 1-10 sec)
2. Vesicle repriming (2-30 sec)
3. Reserve pool mobilization (seconds)

---

PHASE 3: MEDIUM-TERM ADAPTATION (10 sec - 10 min)

Example: LTP Induction at 100 Hz for 1 sec (Tetanus)

**0-1 sec: High-frequency stimulation**
- 100 APs delivered
- Massive Ca²⁺ accumulation in terminal
- [Ca²⁺] builds to sustained 1-2 µM
- Complete RRP depletion
- Strong glutamate release

**1-30 sec: Retrograde signaling arrives**
Postsynaptic spine produces:
1. Nitric Oxide (NO) - diffuses in seconds
2. Brain-Derived Neurotrophic Factor (BDNF) - released in minutes
3. Endocannabinoids (eCBs) - for LTD case

NO diffuses into presynaptic terminal (t ≈ 5-10 sec)
    ↓
**NO activates soluble guanylyl cyclase (sGC)**
    ↓
sGC produces cGMP from GTP
    ↓
**cGMP increases from ~1 nM to 100 nM**
    ↓
cGMP activates Protein Kinase G (PKG)

30 sec - 5 min: PKG-Mediated Pr Enhancement

PKG phosphorylates multiple targets:

1. **VGCCs (P/Q-type):**
   - Phosphorylation at specific serine residues
   - Open probability increases from 0.5 → 0.7
   - More Ca²⁺ enters per AP

2. **Munc18:**
   - Enhanced interaction with Syntaxin
   - Vesicle priming rate increases 2×

3. **Synaptotagmin-1:**
   - Increased Ca²⁺ sensitivity (Kd decreases from 10→5 µM)
   - Faster binding kinetics

4. **RIM proteins:**
   - Enhanced vesicle tethering
   - Better VGCC-vesicle coupling

**Net effect by 5 min:**
- Pr increases from 0.3 → 0.5
- Baseline Ca²⁺ sensitivity increased
- Readily Releasable Pool size increases 30%

---

PHASE 4: SLOW CONSOLIDATION (10 min - 2 hours)

Local Protein Synthesis (Presynaptic)

**30 min - 2 hours:**
BDNF binds to TrkB receptors on presynaptic terminal
    ↓
Activation of PI3K/mTOR pathway
    ↓
**Local translation of presynaptic mRNAs:**
1. VGCC subunits (α1A, β4)
2. Synaptotagmin-1
3. Munc13-1
4. SNARE proteins

**Result by 2 hours:**
- 50% more VGCCs clustered at active zone
- 40% more Synaptotagmin molecules per vesicle
- Pr stabilizes at 0.6

Metabolic Support System

**Astrocyte coordination:**
1. Glutamate uptake → converted to glutamine
2. Glutamine exported to presynaptic terminal
3. Presynaptic mitochondria increase oxidative phosphorylation
4. ATP production increases 2× to support enhanced release

**Energy requirements:**
- Vesicle recycling: ~10,000 ATP/vesicle
- Ca²⁺ clearance: ~1 ATP/2 Ca²⁺ ions
- Protein synthesis: ~4 ATP/amino acid

---

PHASE 5: STRUCTURAL CONSOLIDATION (2 hours - 24 hours)

Nuclear Signaling & Gene Expression

**2-6 hours:**
Persistent kinase activity (PKG, PKA, MAPK)
    ↓
CREB phosphorylation in presynaptic nucleus
    ↓
**Gene expression changes:**
1. Structural proteins (Bassoon, Piccolo)
2. Active zone components
3. Vesicle cycle proteins
4. Metabolic enzymes

**12-24 hours:**
New proteins arrive via axonal transport
    ↓
**Active zone remodeling:**
- Active zone area increases 30-50%
- More docked vesicles (RRP size doubles)
- VGCC-vesicle distance decreases to 15 nm
- **Pr stabilizes at 0.7-0.8**

---

CALCIUM HOMEOSTASIS TIMELINE SUMMARY

Time	[Ca²⁺] at Release Site	Clearance Mechanism	Residual Effect
0 ms	100 nM (baseline)	-	-
0.5 ms	10-50 µM (peak)	Diffusion only	Fusion decision
5 ms	1-5 µM	Fast buffers (calbindin)	Ca²⁺-calmodulin activation
20 ms	500 nM	PMCA pumps active	Facilitation of next spike
100 ms	300 nM	NCX contributes	Augmentation phase
1 sec	200 nM	Mitochondrial uptake	Potentiation
10 sec	150 nM	Steady-state clearance	LTP induction possible
1 min	120 nM	Full homeostasis	-
1 hour	100 nM	Normal resting state	-

---

EXAMPLE: COMPLETE LTP TIMELINE

Day 1: Induction Phase

**T=0:** 100 Hz tetanus (1 sec)
**T=5 sec:** NO arrives at presynaptic terminal
**T=30 sec:** cGMP peaks, PKG activated
**T=5 min:** Pr increases to 0.45 (phosphorylation)
**T=30 min:** Local protein synthesis begins
**T=2 hours:** Pr = 0.55, structural proteins arriving
**T=6 hours:** Active zone remodeling begins
**T=24 hours:** Pr stabilizes at 0.65, active zone enlarged 40%

Day 2-7: Maintenance

**Metabolic support ongoing:**
- Astrocyte supplies glutamine/lactate
- Mitochondrial density increases near active zone
- Vesicle recycling efficiency improves

**Structural stabilization:**
- New active zone material incorporated
- Cytoskeleton reorganizes
- Pr maintains at 0.65-0.70

---

KEY BIOLOGICAL INSIGHTS

1. Pr is multi-dimensional:
   • Fast component: Vesicle availability × Ca²⁺ sensitivity
   • Slow component: Protein composition × active zone geometry
2. Energy dependence:
   • Ca²⁺ clearance requires constant ATP
   • Without ATP, [Ca²⁺] remains elevated → toxicity
   • Metabolic veto: Low ATP → release inhibition despite Ca²⁺
3. Timescale hierarchy:
   • ms: Existing protein modification (phosphorylation)
   • min: Local protein synthesis
   • hours: Nuclear transcription
   • days: Structural reorganization
4. Homeostatic balance:
   • Enhanced Pr increases metabolic demand
   • Requires coordinated astrocyte support
   • Long-term maintenance depends on energy availability

This detailed timeline shows how a single probabilistic event (vesicle release) is embedded in a complex, multi-timescale regulatory system that balances immediate communication needs with long-term information storage and metabolic sustainability.

---

---

---

Based on the provided document, here is a specification of the Short-Term Plasticity (STP/STD) and Long-Term Plasticity (LTP/LTD) mechanisms between the Presynapse and Postsynapse, detailing both internal processes and their interactions, with explicit timescales.

Summary: Primary Plasticity Mechanisms & Timescales

Mechanism	Primary Locus	Key Internal Trigger	Key Interactive Signal	Timescale	Functional Role
Short-Term Depression (STD)	Presynaptic	Vesicle pool depletion	Reduced glutamate release	Fast (<100ms)	Filters high-frequency bursts; transient synaptic fatigue.
Short-Term Potentiation (STP)	Presynaptic	Residual Ca²⁺ buildup	Increased glutamate release probability (P_r)	Fast to Medium (<100ms to 10s)	Facilitates temporal summation; augments recent activity.
Long-Term Depression (LTD)	Postsynaptic	Moderate, sustained Ca²⁺ influx (~1-5 µM)	Retrograde endocannabinoids (eCBs)	Slow (Seconds to Minutes)	Weakens ineffective connections; homeostatic adjustment.
Long-Term Potentiation (LTP)	Postsynaptic	Strong, coincident Ca²⁺ influx (>10 µM)	Retrograde NO/BDNF	Slow (Seconds to Minutes)	Strengthens correlated pre- and postsynaptic activity.
Structural LTP/LTD	Both	Persistent molecular tags & protein synthesis	Trophic factors & homeostatic scaling	Structural (Days+)	Embeds memory persistently via physical changes.

---

Detailed Breakdown by Timescale

1. Fast Timescale (<100 ms): STP & STD Internal Mechanisms

• Presynaptic Internal (STD): Rapid vesicle fusion and release depletes the readily releasable pool. This is a presynaptic, internal mechanism causing a transient decrease in synaptic strength.
• Presynaptic Internal (STP): Residual Ca²⁺ from a preceding action potential lingers, increasing the release probability (P_r) for the next spike. This is a presynaptic, internal facilitatory mechanism.
• Interaction (Fast Signaling): The presynapse releases glutamate (outgoing signal). The postsynapse receives it and, if sufficiently depolarized, opens NMDA receptors, allowing a Ca²⁺ influx. This Ca²⁺ transient is the postsynaptic, internal coincidence detector signal that initiates the cascade for slower plasticity.

2. Medium Timescale (100 ms – 10 s): Augmentation & Modulation

• Presynaptic Internal: Augmentation via Munc13 proteins modifies P_r based on Ca²⁺ sensing. This is a presynaptic, internal continuation of STP.
• Postsynaptic Internal: Metabotropic glutamate receptor (mGluR) activation modulates local spine excitability and prepares plasticity pathways. This is a postsynaptic, internal modulatory state.
• Interaction: Largely an extension of fast signaling, setting the stage for slower decisions. The pattern of glutamate release interacts with the postsynaptic voltage state.

3. Slow Timescale (Seconds – Minutes): LTP & LTD Decision & Expression

This is the critical window for bidirectional interaction that establishes long-term change.

• Postsynaptic Internal (The Decision):
  • LTP Trigger: High, localized Ca²⁺ (from strong NMDA activation + back-propagating AP) activates CaMKII, creating a plasticity tag.
  • LTD Trigger: Moderate, sustained Ca²⁺ (from isolated glutamate release or low-frequency stimulation) activates phosphatases (e.g., calcineurin).
  • This decision is postsynaptic and internal.
• Interaction (Retrograde Messaging):
  • For LTP: The postsynapse synthesizes and releases retrograde signals (NO, BDNF). These diffuse to the presynapse.
  • For LTD: The postsynapse releases endocannabinoids (eCBs).
  • These are interactive signals from postsynapse to presynapse.
• Presynaptic Internal (Expression):
  • For LTP: NO/BDNF activate enzymatic cascades that persistently increase the baseline P_r.
  • For LTD: eCBs bind to CB1 receptors, inhibiting VGCCs and persistently decreasing P_r (Direct Synaptic Depression, DSE).
  • This change in baseline P_r is the presynaptic, internal expression of the long-term change.

4. Metabolic Timescale (Minutes – Hours): Consolidation

• Interaction (Astrocyte Bridge): The astrocyte supplies D-serine (co-agonist for NMDAR) and lactate (energy). This external, interactive support is required for stable consolidation.
• Postsynaptic Internal: The spine initiates local protein synthesis, using the "tag" to capture newly made proteins, transitioning early-LTP/LTD to a more stable state. This is a postsynaptic, internal consolidation process.

5. Structural Timescale (Days+): Embodiment

• Presynaptic Internal: The bouton grows or retracts, changing the active zone size and vesicle pool. This is a presynaptic, internal structural change.
• Postsynaptic Internal: The spine changes its volume and number of AMPA receptor slots. This is a postsynaptic, internal structural change.
• Interaction (Global Scaling): The soma sends homeostatic scaling signals to all synapses (including this one) to maintain network stability, providing a top-down, interactive modulation that can override local weights.

Conclusion

• STP/STD are predominantly presynaptic, internal phenomena operating on fast to medium timescales (ms to s), governed by vesicle dynamics and residual calcium.
• LTP/LTD are initiated by a postsynaptic, internal calcium-based decision on a slow timescale (s to min). Their expression involves a critical bidirectional interaction: retrograde signals (NO/BDNF for LTP, eCBs for LTD) from the postsynapse induce presynaptic, internal changes in baseline P_r.
• These changes are then stabilized on metabolic and structural timescales through astrocytic support and physical remodeling of both synaptic components, embedding the memory trace in the physical architecture of the connection.

---

---

---

In biology, presynaptic release probability (Pr) emerges from a dynamic molecular machine, not a simple number. Changing Pr involves coordinated modifications to multiple components of the vesicle release apparatus. Here's the biological process breakdown:

The Pr Machinery: Three Core Components

Pr = f(Calcium Influx × Vesicle Readiness × Fusion Machinery Sensitivity)

1. Modulating Calcium Influx

Target: Voltage-Gated Calcium Channels (VGCCs, mainly P/Q and N-type).

• Increasing Pr: Phosphorylation of VGCCs by kinases (PKA, PKC) enhances their open probability or prolongs open time.
• Decreasing Pr:
  • Direct inhibition by G-proteins (e.g., via activated CB1 receptors during LTD).
  • Physical relocation of channels away from release sites.
  • Dephosphorylation by phosphatases (calcineurin).

Biological Process: A retrograde messenger (e.g., NO) activates a kinase cascade in the presynaptic terminal, leading to VGCC phosphorylation → more Ca²⁺ enters per action potential → higher Pr.

2. Modulating Vesicle Readiness (Docking/Priming)

Targets: Docking proteins (Syntaxin, SNAP-25), priming proteins (Munc13, Munc18), and the vesicle pool itself.

• Increasing Pr:
  • Munc13 activation: Residual Ca²⁺ binds to calmodulin, which binds to Munc13, increasing its priming activity. This is the main mechanism for short-term facilitation.
  • Phosphorylation of priming proteins by PKC/CaMKII makes them more active.
  • Increased expression or recruitment of vesicles to the "readily releasable pool" (RRP).
• Decreasing Pr:
  • Dephosphorylation of priming proteins.
  • Physical depletion of RRP during high-frequency firing (STD).
  • Ubiquitination and degradation of priming proteins.

3. Modulating Fusion Machinery Sensitivity (Ca²⁺ Sensor)

Target: The primary Ca²⁺ sensor Synaptotagmin and the SNARE complex (Syntaxin, Synaptobrevin, SNAP-25).

• Increasing Pr:
  • Phosphorylation of Synaptotagmin increases its Ca²⁺ affinity.
  • Phosphorylation of SNARE proteins (e.g., SNAP-25 by PKC) enhances fusion kinetics.
  • Assembly of more SNARE complexes.
• Decreasing Pr:
  • Cleavage of SNARE proteins by toxins (e.g., botulinum).
  • Increased binding of inhibitory proteins like Complexins.

Specific Biological Pathways for Pr Changes

Fast Pr Increase (Facilitation, <100ms)

Process: Action potential → Ca²⁺ influx → residual Ca²⁺ binds to calmodulin → Ca²⁺-calmodulin binds to Munc13 → Munc13 increases vesicle priming rate → more vesicles become release-ready for the next spike.

Biological signature: Transient, activity-dependent, decays with Ca²⁺ clearance.

Slow Pr Increase (LTP Expression, Minutes+)

Process: Retrograde NO diffuses into presynaptic terminal → activates soluble guanylyl cyclase → produces cGMP → activates Protein Kinase G (PKG) → PKG phosphorylates multiple targets:

1. VGCCs → increased Ca²⁺ influx
2. Munc18 → enhanced vesicle priming
3. Synaptotagmin → increased Ca²⁺ sensitivity Plus: Local protein synthesis of new vesicle proteins.

Biological signature: Persistent, requires gene expression for maintenance, structurally embedded.

Slow Pr Decrease (LTD Expression, Minutes+)

Process: Retrograde eCBs bind presynaptic CB1 receptors → activates Gi/o proteins → βγ subunits directly inhibit VGCCs → reduced Ca²⁺ influx → decreased Pr. Additionally: Gi/o inhibits adenylyl cyclase → reduced cAMP → reduced PKA activity → decreased phosphorylation of release machinery.

Biological signature: Reversible initially, can become structural with prolonged activation.

The Complete Biological Cascade for LTP-Induced Pr Increase

Retrograde NO/BDNF
      ↓
Presynaptic receptor activation
      ↓
Second messenger cascade (cGMP/cAMP)
      ↓
Kinase activation (PKG/PKA/PKC)
      ↓
Phosphorylation of multiple targets:
1. VGCCs → ↑ Ca²⁺ influx per AP
2. Munc13/18 → ↑ vesicle priming rate
3. Synaptotagmin → ↑ Ca²⁺ sensitivity
4. SNARE proteins → ↑ fusion efficiency
      ↓
Increased Pr from baseline 0.2 → 0.4
      ↓
More reliable vesicle release per AP
      ↓
Structural reinforcement (Days+):
- More active zone material
- More docked vesicles
- Closer VGCC-vesicle coupling

Simulation vs. Biological Reality

In Simulation	In Biology
Pr = 0.3	Pr emerges from: 5 docked vesicles × 40% channel open probability × 60% fusion probability
Update rule: Pr += 0.1 if LTP	LTP: NO → cGMP → PKG → phosphorylates 3 proteins → increases Ca²⁺ influx 30% & priming rate 40%
Single variable	Distributed property of ~50 protein species
Instant change	Change occurs over: 100ms (fast), 1min (slow), 24hr (structural)

Key Biological Insights

1. Pr is multi-parametric: Not one number but the product of multiple probabilities.
2. Redundancy: Multiple pathways converge on Pr (kinases, protein synthesis, structural changes).
3. Timescale hierarchy:
   • Fast (ms): Modify existing proteins (phosphorylation)
   • Slow (min): Synthesize new proteins
   • Structural (days): Change physical architecture
4. Metabolic dependence: All phosphorylation requires ATP. During energy crisis, kinases fail → Pr decreases even with "LTP signals."
5. Cross-talk: The same kinase (e.g., PKC) can phosphorylate both VGCCs and SNARE proteins, creating coordinated enhancement.

In essence, Pr is the presynaptic terminal's "readiness state"—a complex, dynamic equilibrium of dozens of interacting proteins that the neuron tunes through phosphorylation, gene expression, and structural remodeling to implement learning and adaptation.

---

---

---

The presynapse does NOT release neurotransmitter every time a presynaptic spike arrives. This is a fundamental principle of synaptic transmission: release is probabilistic, not deterministic.

What Pr Actually Influences

Pr (Release Probability) determines the likelihood that a given action potential will trigger vesicle fusion and neurotransmitter release. It's essentially the synapse's "coin toss" probability for transmission.

Biological Reality:

For a typical central synapse:

• Pr ≈ 0.2-0.5 (meaning only 20-50% of spikes cause release)
• Some specialized synapses (like the neuromuscular junction) have Pr ≈ 0.9
• Some cortical synapses have Pr as low as 0.1

What Happens at a Presynaptic Terminal When a Spike Arrives:

Action Potential arrives at terminal
    ↓
VGCCs open → Ca²⁺ enters
    ↓
**PROBABILISTIC DECISION POINT**
    ↓
├── **With probability Pr (e.g., 0.3):**
│       ↓
│   Ca²⁺ binds to Synaptotagmin on docked vesicle
│       ↓
│   SNARE complex zippers completely
│       ↓
│   Vesicle membrane fuses with presynaptic membrane
│       ↓
│   **GLUTAMATE RELEASED** → Postsynaptic response
│
└── **With probability 1-Pr (e.g., 0.7):**
        ↓
    Ca²⁺ enters but insufficient to trigger fusion
        ↓
    Vesicle remains docked but unfused
        ↓
    **NO RELEASE** → No postsynaptic response

The Biological Basis of This Stochasticity

1. Calcium Nanodomain Stochasticity

• Ca²⁺ channels are ~20-30 nm from vesicle release sites
• When a channel opens, only ~100-300 Ca²⁺ ions enter
• These ions form a brief, localized "nanodomain"
• Random diffusion and buffering mean the Ca²⁺ concentration at the sensor varies randomly
• If <5-10 µM at the sensor → no fusion; if >10 µM → fusion

2. Molecular Stochasticity

• Synaptotagmin (Ca²⁺ sensor) requires 3-5 Ca²⁺ ions to bind
• SNARE complex zippering is thermally driven and probabilistic
• These molecular processes have intrinsic randomness at nanoscale

3. Vesicle State Stochasticity

• Vesicles exist in different "readiness states":
  • Docked & primed (immediately releasable)
  • Docked but unprimed
  • Reserve pool (undocked)
• Transition between states is probabilistic

Why Evolution Selected Probabilistic Release

Advantages of Low/Moderate Pr:

1. Dynamic Range: Allows both facilitation (Pr increases with repeated spikes) and depression (Pr decreases with vesicle depletion)
2. Noise Filtering: Random failures filter out isolated, potentially noisy spikes while allowing bursts to get through (due to facilitation)
3. Energy Efficiency: Release machinery (vesicle fusion/recycling) is energetically expensive. Why waste energy on every spike?
4. Information Capacity: Probabilistic release allows synapses to encode information in release probability as well as spike rate
5. Plasticity Substrate: The stochastic nature provides a substrate for learning through Hebbian mechanisms

What Actually Gets Released When Release Occurs

Even when release occurs, it's not always "all or nothing":

Quantal Release:

• Miniature EPSCs (mEPSCs): Single vesicle release (~1000-5000 glutamate molecules)
• Univesicular Release: Most CNS synapses release 0 or 1 vesicle per active zone per spike
• Multivesicular Release: Some synapses can release 2-3 vesicles simultaneously

Key Measurement:

EPSC amplitude = Q × n × Pr

Where:

• Q = quantal size (response to one vesicle)
• n = number of release sites (active zones)
• Pr = release probability per site

How This Plays Out in Neural Computation

Example: Cortical Pyramidal Neuron Synapse

Spike train arrives: [1, 1, 1, 1, 1, 1] (6 spikes at 50 Hz)

With Pr = 0.3:
Actual releases might be: [0, 1, 0, 0, 1, 1]

Postsynaptic effect:
- Spike 1: No response
- Spike 2: EPSP
- Spike 3: No response
- Spike 4: No response
- Spike 5: EPSP
- Spike 6: EPSP

But due to facilitation:
Pr increases to 0.5 by spike 6

Simulation Implications

In simulations, you would implement this as:

import numpy as np

class Synapse:
    def __init__(self, Pr=0.3):
        self.Pr = Pr  # Release probability
        self.vesicles_available = True
        
    def process_spike(self):
        if not self.vesicles_available:
            return 0  # No release due to depletion
            
        # Probabilistic release decision
        if np.random.random() < self.Pr:
            # Release occurs
            self.vesicles_available = False  # Temporary depletion
            self.trigger_recycle_timer()  # Schedule vesicle recycling
            return 1  # 1 vesicle released
        else:
            # Release fails probabilistically
            return 0

Biological Determinants of Pr

High Pr synapses tend to have:

• Denser VGCC clustering near release sites
• Higher Ca²⁺ sensitivity of Synaptotagmin
• More efficient priming machinery (Munc13)
• Larger readily releasable pool

Low Pr synapses tend to have:

• Sparse VGCC distribution
• Lower Ca²⁺ sensitivity
• Less efficient priming
• Smaller readily releasable pool

The Big Picture

Pr is the synapse's "transmission reliability setting" that:

1. Filters spike trains probabilistically
2. Encodes short-term plasticity through dynamic changes
3. Embeds long-term memory through stable modifications
4. Conserves energy by not releasing on every spike

This stochasticity isn't a bug—it's a feature that allows synapses to be temporal filters, memory elements, and computational units all at once. The fact that transmission isn't guaranteed is precisely what gives synapses their rich computational properties, enabling phenomena like paired-pulse facilitation, synaptic filtering of high-frequency bursts, and providing a substrate for Hebbian learning through coincident detection failures.