Files
organism/neuron/appunti/2026-01-01-pre-post-atrocyte-timing.md
T
2026-04-01 12:41:18 +02:00

816 lines
23 KiB
Markdown
Raw Blame History

This file contains ambiguous Unicode characters
This file contains Unicode characters that might be confused with other characters. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.
# Multi-Timescale Neural Component Analysis
## PRESYNAPSE
### Timescale 1: Fast (<1ms - 100ms)
**Incoming Signals:**
- Action potential depolarization (to \~+30 mV)
- Voltage-gated calcium channel (VGCC) opening
**Actions:**
- Vesicle docking and priming (SNARE complex assembly)
- Ca²⁺ influx
- Glutamate vesicle release (stochastic, probability P_r)
**Outgoing Signals:**
- Glutamate release into synaptic cleft
**Modulation:**
- **Upregulation (Facilitation):** Residual Ca²⁺ from previous spikes increases P_r for next release
- **Downregulation (Depression):** High-frequency firing depletes readily releasable vesicle pool, decreasing P_r
---
### Timescale 2: Medium (100ms - 10s)
**Incoming Signals:**
- Buildup of residual Ca²⁺
- Volume transmission neuromodulators (dopamine, acetylcholine, noradrenaline)
**Actions:**
- Accumulation/depletion of Ca²⁺ stores
- Modulation of release machinery sensitivity
**Outgoing Signals:**
- Sustained or diminished glutamate release patterns (STF/STD)
**Modulation:**
- **Short-Term Facilitation (STF):** Residual Ca²⁺ increases P_r over spike trains
- **Short-Term Depression (STD):** Vesicle pool depletion reduces P_r
- **Augmentation:** Calcium-sensing proteins (Munc13) alter release probability (1-10s range)
**Notes:**
This active clearance happens rapidly, within tens to hundreds of milliseconds. It serves two vital functions:
- **Termination of Signal:** It rapidly lowers Ca²⁺ to end the release command, ensuring neurotransmitter release is brief and precise.
- **Prevention of Toxicity:** Sustained high intracellular Ca²⁺ is cytotoxic and can trigger apoptosis (cell death). Efficient clearance is essential for neuronal health.
- **Facilitation:** If Ca²⁺ clearance is slightly slower than the arrival of the next action potential, residual Ca²⁺ accumulates near the release sites. This "leftover" Ca²⁺ adds to the influx from the next spike, making vesicle fusion more likely (increasing P<sub>r</sub>).
- **Depression:** If firing is very rapid, the pumps and exchangers cannot keep up, and Ca²⁺ levels remain elevated for longer in a more diffuse manner. This can paradoxically activate processes that inhibit release or simply outpace the recycling of vesicles, leading to depletion.
---
### Timescale 3: Slow (seconds - minutes)
**Incoming Signals:**
- Retrograde NO (nitric oxide) from postsynapse
- Retrograde BDNF (brain-derived neurotrophic factor)
- Retrograde endocannabinoids (eCBs, e.g., 2-AG)
- Astrocyte gliotransmitters (ATP, D-serine, glutamate)
**Actions:**
- Enzymatic cascade activation/suppression
- CB1 receptor activation (by eCBs)
- VGCC modulation
- Potassium channel modulation
**Outgoing Signals:**
- Modified P_r affecting subsequent releases
**Modulation:**
- **Upregulation:** NO/BDNF activates cascades that increase P_r, promote synaptic growth (facilitates LTP)
- **Downregulation:** eCBs bind CB1 receptors, inhibit VGCCs, activate K⁺ channels → profound decrease in P_r (DSE/DSI - depolarization-induced suppression)
### Timescale 4: Metabolic (minutes - hours)
**Incoming Signals:**
- Astrocyte-supplied lactate (via monocarboxylate transporters)
- Glutamine from astrocytes (glutamate-glutamine cycle)
- Metabolic state indicators (ATP levels, NAD/NADH ratio)
**Actions:**
- ATP-dependent vesicle cycling
- Glutamine→glutamate conversion (via glutaminase)
- Vesicle refilling with glutamate
- Maintenance of ion gradients
**Outgoing Signals:**
- Sustained neurotransmitter release capacity
- Metabolic demand signals to astrocyte
**Modulation:**
- **Metabolic veto:** Insufficient ATP prevents vesicle release despite adequate Ca²⁺
- Lactate availability determines sustained release capacity during high activity
---
### Timescale 5: Structural (hours - days+)
**Incoming Signals:**
- Retrograde trophic factors (BDNF, sustained)
- Homeostatic scaling signals from soma
**Actions:**
- Structural growth/retraction of presynaptic bouton
- Changes in active zone size
- Alterations in vesicle pool size
**Outgoing Signals:**
- Modified synaptic strength through structural change
**Modulation:**
- Terminal size increases/decreases
- Vesicle pool capacity changes
- Active zone protein composition changes
---
## POSTSYNAPSE (Dendritic Spine)
### Timescale 1: Fast (<1ms - 100ms)
**Incoming Signals:**
- Glutamate binding to AMPA receptors (<1ms)
- Glutamate binding to NMDA receptors (Mg²⁺-blocked initially)
- Local depolarization from AMPA activation
- GABA from inhibitory interneurons
**Actions:**
- AMPA receptor opening → Na⁺ influx → local depolarization (EPSP)
- NMDA receptor Mg²⁺ unblock (requires depolarization > -40mV)
- NMDA receptor opening → Ca²⁺ influx
- AMPA receptor desensitization (if glutamate lingers)
**Outgoing Signals:**
- EPSP propagating to dendritic branch
- Local Ca²⁺ concentration changes
**Modulation:**
- **Upregulation:** Depolarization relieves NMDA Mg²⁺ block → Ca²⁺ influx amplification
- **Downregulation:** AMPA desensitization acts as low-pass filter
---
### Timescale 2: Medium (100ms - 10s)
**Incoming Signals:**
- Sustained glutamate exposure
- Metabotropic glutamate receptor (mGluR) activation
- GABA-B receptor activation (slow inhibition)
**Actions:**
- G-protein coupled signaling cascades
- Second messenger activation
- Modulation of local excitability
**Outgoing Signals:**
- Modified EPSP amplitude based on recent history
- Preparation for plasticity events
**Modulation:**
- mGluR-mediated changes in spine excitability
- GABA-B provides prolonged shunting inhibition (100ms-1s)
---
### Timescale 3: Slow (seconds - minutes)
**Incoming Signals:**
- Sustained high Ca²⁺ influx through NMDARs
- Back-propagating action potential (bAP) from soma/AIS
- D-serine co-agonist from astrocyte
**Actions:**
- CaMKII (calcium/calmodulin-dependent protein kinase II) autophosphorylation
- Synaptic tagging: Ca²⁺ creates local molecular "tag" marking synapse as recently active
- Synthesis and release of retrograde messengers:
- **NO synthesis** (from high Ca²⁺)
- **BDNF release**
- **Endocannabinoid (eCB) synthesis** (from sustained Ca²⁺)
- AMPA receptor trafficking from extrasynaptic pool into PSD (early-LTP)
**Outgoing Signals:**
- Retrograde NO (diffuses to presynapse)
- Retrograde BDNF (travels to presynapse)
- Retrograde eCBs (diffuse to presynapse)
**Modulation:**
- **Upregulation (LTP):** High Ca²⁺ (>10 μM) → CaMKII activation → spine head expansion → AMPAR insertion → increased synaptic weight
- **Downregulation (LTD):** Low/sustained Ca²⁺ (0.5-1.0 μM) → phosphatase activation → spine shrinkage → AMPAR endocytosis → decreased weight
- Tag duration: CaMKII phosphorylation state acts as \~1-2 hour memory tag
---
### Timescale 4: Metabolic (minutes - hours)
**Incoming Signals:**
- Astrocyte D-serine (NMDA co-agonist, essential for late-LTP)
- Astrocyte TNF-α, cholesterol (permissive factors)
- Metabolic substrates for local protein synthesis
**Actions:**
- Local protein synthesis in dendrites (responds to "tags")
- Structural spine remodeling
- Transition from early-LTP to late-LTP (L-LTP)
**Outgoing Signals:**
- Demand signals for continued metabolic support
**Modulation:**
- D-serine availability gates transition to L-LTP
- Metabolic state determines whether tagged synapses can undergo structural consolidation
---
### Timescale 5: Structural (hours - days+)
**Incoming Signals:**
- Persistent synaptic tags combined with eligibility signals
- Homeostatic scaling signals from soma
- Neuromodulator-driven metaplasticity signals
**Actions:**
- Spine volume changes (0.01-1.0 μm³ range)
- Receptor number changes (AMPAR density)
- Structural consolidation or elimination
- PSD protein composition changes
**Outgoing Signals:**
- Stable changes in synaptic weight
**Modulation:**
- Successful spines grow and strengthen
- Weak/unused spines shrink or are eliminated
- Homeostatic scaling adjusts all synapses proportionally
---
## DENDRITE
### Timescale 1: Fast (10-100ms)
**Incoming Signals:**
- Multiple EPSPs from spines on the branch
- Back-propagating action potential (bAP) from soma
- IPSPs from inhibitory synapses
- Neuromodulator influence on dendritic K⁺ channels
**Actions:**
- Spatial summation: EPSPs from different spines add together
- Temporal summation: EPSPs from successive spikes add together
- NMDA spike generation (local Ca²⁺ spike if threshold reached)
- Dendritic Na⁺/Ca²⁺ spike generation
**Outgoing Signals:**
- Integrated EPSP to soma
- Dendritic spike (amplified signal)
- bAP propagation modulated by local K⁺ channels
**Modulation:**
- **Upregulation:** Dendritic spikes amplify signals; coincidence of local EPSP + bAP enhances NMDA activation
- **Downregulation:** K⁺ channels limit bAP propagation; strong inhibition can veto dendritic spikes
- bAP amplitude and spread actively modulated by dendritic K⁺ channels → regulated teaching signal
---
### Timescale 2: Medium (100ms - 10s)
**Incoming Signals:**
- Patterns of dendritic spikes
- Neuromodulator tone affecting dendritic excitability
**Actions:**
- Branch-level integration over short time windows
- Modulation of dendritic spike threshold
**Outgoing Signals:**
- Pattern-classified signals to soma
**Modulation:**
- Dendritic excitability adjusted by neuromodulator context
- Short-term changes in integration properties
---
### Timescale 3: Slow (seconds - minutes)
**Incoming Signals:**
- Sustained patterns of activity
- Astrocyte gliotransmitters affecting dendritic excitability
**Actions:**
- Coincidence detection for STDP (spike-timing-dependent plasticity)
- Branch acts as pattern classifier
**Outgoing Signals:**
- Timing information (pre-post spike timing) determining LTP/LTD sign
**Modulation:**
- Timing of pre- and postsynaptic activity determines sign (LTP vs LTD) and magnitude of plastic change
- Branch-specific computation and learning rules
---
### Timescale 4: Metabolic (minutes - hours)
**Incoming Signals:**
- Metabolic support from astrocytes
- Proteins synthesized locally in dendrites
**Actions:**
- Local protein synthesis in response to activity
- Maintenance of ion gradients
- Support for sustained dendritic spike generation
**Outgoing Signals:**
- Metabolic demand signals
**Modulation:**
- Availability of metabolic substrates determines capacity for local plasticity
- Local translation enables rapid structural changes without waiting for somatic gene expression
---
### Timescale 5: Structural (hours - days+)
**Incoming Signals:**
- Homeostatic scaling signals
- Structural plasticity factors
**Actions:**
- Dendritic branch growth/retraction
- Changes in spine density
- Alterations in dendritic arbor complexity
**Outgoing Signals:**
- Modified dendritic integration capacity
**Modulation:**
- Experience-dependent dendritic remodeling
- Branch-specific structural changes based on activity history
---
## SOMA
### Timescale 1: Fast (1-100ms)
**Incoming Signals:**
- Thousands of filtered EPSPs and IPSPs from all dendritic branches
- Direct perisomatic inhibition from basket cells and chandelier cells
- HCN channel (Ih current) activity at rest
**Actions:**
- Final spatial/temporal summation of all inputs
- Voltage-gated channel activity (HCN channels stabilize membrane potential)
- Integration toward or away from spike threshold
**Outgoing Signals:**
- Integrated voltage (Vm) to AIS
- Decision point: spike or no spike
**Modulation:**
- **Upregulation:** Excitatory inputs sum toward threshold; depolarization
- **Downregulation:** Perisomatic inhibition exerts powerful veto control; HCN channels act as "voltage clamp" resisting large swings
- Direct somatic inhibition can clamp voltage below threshold, overriding all excitatory input
---
### Timescale 2: Medium (100ms - seconds)
**Incoming Signals:**
- Spike afterhyperpolarization (sAHP) from recent spike
- Neuromodulator receptor activation (ACh, noradrenaline, serotonin, dopamine)
**Actions:**
- Ca²⁺-activated K⁺ (SK) channel opening (from sAHP)
- Kv7 (M-type) K⁺ channel modulation
- Adjustment of input resistance and excitability
**Outgoing Signals:**
- Modified excitability state
- Spike frequency adaptation
**Modulation:**
- **Upregulation:**
- Inactivation of Kv7 channels (by ACh or prior depolarization) → lower threshold, increased input resistance
- Reduced sAHP (by noradrenaline) → allows higher sustained firing rates
- **Downregulation:**
- sAHP produces prolonged hyperpolarization (hundreds of ms) → potently suppresses further firing → spike frequency adaptation
---
### Timescale 3: Slow (seconds - minutes)
**Incoming Signals:**
- Sustained neuromodulatory tone
- Integrated Ca²⁺ signals from dendritic activity
**Actions:**
- Modulation of global neuronal excitability
- Preparation for plasticity events
- Ca²⁺/CaMKIV signaling integration
**Outgoing Signals:**
- Somatic Ca²⁺ state influencing gene expression pathways
- Modified neuronal gain
**Modulation:**
- Neuromodulators set global "mood" of neuron
- Somatic state gates whether dendritic tags will be consolidated
---
### Timescale 4: Metabolic (minutes - hours)
**Incoming Signals:**
- Mean firing rate (F_avg) integrated over hours
- Metabolic state indicators (ATP, lactate availability)
- Astrocyte metabolic support signals
**Actions:**
- Somatic Ca²⁺/CaMKIV signaling senses mean activity
- Initiation of homeostatic responses
- Gene expression programs triggered
**Outgoing Signals:**
- Demand for metabolic support
- Signals initiating homeostatic adjustments
**Modulation:**
- Metabolic state determines capacity for sustained firing
- Energy availability gates neuronal operations
---
### Timescale 5: Structural (hours - days+)
**Incoming Signals:**
- Integrated activity history (hours-days)
- Neuromodulator-driven metaplasticity signals
- CREB, BDNF transcription activation
**Actions:**
- Global synaptic scaling: AMPA receptor transcription/trafficking changes across all synapses
- Homeostatic plasticity (typically 12-48 hours)
- Metaplasticity: changes in plasticity rules themselves
- Gene expression establishing new baseline excitability
**Outgoing Signals:**
- Homeostatic scaling factors to all synapses
- Modified intrinsic excitability parameters
- Changed plasticity thresholds
**Modulation:**
- "Corporate-wide audit": soma ensures network stability by proportionally adjusting all synaptic strengths
- Chronic high activity → global downscaling
- Chronic low activity → global upscaling
- Neuromodulators broadcast "global strategy" (e.g., "be alert and learn" vs "sleep and consolidate")
---
## AIS (Axon Initial Segment)
### Timescale 1: Fast (1-100ms)
**Incoming Signals:**
- Somatically integrated voltage (Vm) from soma
- Direct chandelier cell inhibition (targeting proximal axon)
**Actions:**
- Threshold detection (AIS has lowest spike threshold in neuron)
- Explosive opening of high-density voltage-gated Na⁺ (NaV) channels
- All-or-none action potential initiation
- Analog-to-digital conversion: graded somatic voltage → binary spike
**Outgoing Signals:**
- Action potential propagating down axon
- Action potential back-propagating into soma/dendrites (bAP)
**Modulation:**
- **Upregulation:** If Vm crosses AIS threshold, reliable spike initiation
- **Downregulation:**
- Absolute/relative refractory periods enforce maximum firing rate
- Chandelier cell inhibition can completely block spike generation
- High-fidelity trigger: faithfully converts somatic voltage to timed output
---
### Timescale 2: Medium (seconds - minutes)
**Incoming Signals:**
- Recent spike history
- Neuromodulator influence on AIS excitability
**Actions:**
- Refractory period enforcement
- Spike timing precision maintenance
**Outgoing Signals:**
- Precisely timed spike trains
**Modulation:**
- Refractory periods control maximum firing frequency
- Spike timing precision maintained by AIS properties
---
### Timescale 3: Slow (minutes - hours)
**Incoming Signals:**
- Activity-dependent signals
- Metabolic state information
**Actions:**
- Subtle modulation of threshold
- Preparation for structural adjustments
**Outgoing Signals:**
- Modified spike generation parameters
**Modulation:**
- Activity-dependent threshold adjustments
---
### Timescale 4: Structural (hours - days+)
**Incoming Signals:**
- Chronic activity patterns
- Homeostatic signals from soma
**Actions:**
- AIS location can be plastically adjusted (moves closer/farther from soma)
- Channel composition changes (NaV channel density/subtypes)
- Cytoskeletal matrix reorganization
**Outgoing Signals:**
- Modified intrinsic neuronal gain
**Modulation:**
- AIS repositioning effectively changes neuron's input-output function
- Chronic high activity → AIS moves away from soma (decreased excitability)
- Chronic low activity → AIS moves toward soma (increased excitability)
- Changes in AIS properties alter neuronal gain and excitability
---
## ASTROCYTE
### Timescale 1: Fast (milliseconds)
**Incoming Signals:**
- Extracellular glutamate spillover from synapses
- K⁺ efflux from neuronal firing
- Sensing via mGluRs on astrocyte processes
**Actions:**
- Rapid glutamate uptake via EAAT1/2 transporters
- K⁺ uptake via Kir4.1 channels
**Outgoing Signals:**
- Glutamate clearance (prevents excitotoxicity)
- Local K⁺ removal (prevents hyperexcitability)
**Modulation:**
- Immediate protection: prevents excitotoxicity and runaway excitation
- Maintains signal fidelity by clearing neurotransmitter
---
### Timescale 2: Medium (seconds)
**Incoming Signals:**
- Accumulated glutamate uptake
- Astrocytic internal Ca²⁺ waves (triggered by mGluR activation)
**Actions:**
- Ca²⁺ wave propagation through astrocyte syncytium
- Gliotransmitter release (ATP, D-serine, glutamate)
- K⁺ spatial redistribution via gap junctions
**Outgoing Signals:**
- Gliotransmitters to synapses (forming "tripartite synapse")
- D-serine as NMDA co-agonist
- ATP (can be converted to adenosine)
- Spatially redistributed K⁺
**Modulation:**
- Active modulation of synaptic dialogue
- Prevents local hyperexcitability through K⁺ buffering and spatial redistribution
- Can enhance or suppress synaptic transmission
---
### Timescale 3: Slow (minutes)
**Incoming Signals:**
- Sustained neuronal activity patterns
- Rising extracellular K⁺ and glutamate (sustained)
- Internal metabolic state changes (NAD/NADH ratio shifts)
**Actions:**
- Glutamate→glutamine conversion (via glutamine synthetase)
- Glucose uptake stimulation (triggered by glutamate uptake)
- Glycogenolysis (breakdown of glycogen stores)
- Glycolysis: glucose→lactate
**Outgoing Signals:**
- Glutamine export to neurons (for glutamate resynthesis)
- Lactate export to neurons (via MCTs - monocarboxylate transporters)
- Early gliotransmission adjustments
**Modulation:**
- Glutamate-glutamine cycle: "refueling loop" maintains neurotransmitter pool
- ANLS (astrocyte-neuron lactate shuttle): "turbocharger" provides rapid ATP synthesis fuel
- Metabolic buffering for burst neuronal activity
---
### Timescale 4: Metabolic (minutes - hours)
**Incoming Signals:**
- Chemical sensors: sustained high K⁺ and glutamate
- Internal Ca²⁺ waves (integrated over time)
- Metabolic redox state (NAD/NADH ratio)
- Energy depletion: falling glucose and glycogen levels
**Actions:**
- Sustained glucose uptake and glycolysis
- Glycogen replenishment
- Vasomodulation: astrocyte endfeet release vasoactive signals (prostaglandins, epoxyeicosatrienoic acids)
- pH buffering via bicarbonate transporters
- Volume regulation via Aquaporin-4 (AQP4) water channels
- ATP metabolism producing adenosine
- D-serine production for late-LTP support
- Release of metabolic substrates and factors (TNF-α, cholesterol) for local dendritic protein synthesis
**Outgoing Signals:**
- Sustained lactate supply (metabolic fuel)
- Glutamine for neurotransmitter recycling
- Vasoactive signals → local blood vessel dilation (neurovascular coupling)
- Adenosine accumulation (sleep pressure signal)
- pH regulation maintaining optimal enzyme function
- Modulated extracellular space volume and tortuosity (affects neurotransmitter diffusion)
- D-serine for late-LTP
- Permissive factors for local protein synthesis
**Modulation:**
- **Resource manager:** Maintains neurotransmitter pools and delivers emergency fuel
- **Environmental steward:** Homeostasis of ions (K⁺), pH, water balance
- **Systemic regulator:** Matches blood flow to metabolic demand; builds sleep pressure via adenosine
- **Plasticity enabler:** Provides D-serine and metabolic support to transition early-LTP to late-LTP
- Astrocyte microdomain (\~100,000 synapses) functions as local metabolic unit
- Prevents metabolic collapse during high-speed signaling
---
### Timescale 5: Structural (hours - days+)
**Incoming Signals:**
- Chronic activity patterns in local network
- Sleep-wake cycle signals
- Long-term metabolic demand patterns
**Actions:**
- Structural remodeling of astrocyte processes
- Glycogen storage capacity changes
- Glymphatic system clearance (during slow-wave sleep)
- Aquaporin-4 channel facilitation of CSF influx
- Metabolic support for neuronal gene expression programs
- Support for epigenetic modifications
**Outgoing Signals:**
- Changed coverage of synapses (physical enwrapment)
- Waste clearance (amyloid-β, tau) via glymphatic system
- Long-term metabolic support for structural plasticity
- Support for systems-level consolidation
**Modulation:**
- Astrocyte morphology adapts to network activity history
- Nightly glymphatic clearance prevents toxic protein accumulation
- Essential for transitioning labile memory traces to stable long-term form
- Supports neuronal structural rewiring and homeostatic adjustments
- Enables sustainable high-speed computation over lifetime
---
## SYNTHESIS
**Key Principles:**
1. **Traces propagate upward:** Fast processes leave traces (residual Ca²⁺, tags, metabolic demand) that persist into slower timescales
2. **Context flows downward:** Slower processes create context that reinterprets fast events (metabolic state determines if Ca²⁺ can trigger release; homeostatic scaling changes synaptic weights)
3. **The astrocyte is the multi-scale bridge:** Operates at every timescale from milliseconds (glutamate clearance) to lifetime (metabolic support for epigenetics)
4. **No component operates in isolation:** Each receives inputs from multiple timescales and sends outputs that affect multiple timescales
5. **Modulation is contextual:** The "same" signal (e.g., Ca²⁺ influx) has different effects depending on metabolic state, recent history, neuromodulatory tone, and structural configuration
6. **Metabolic veto is real:** Components can refuse to execute operations if metabolic resources are insufficient—this is not a bug but a feature of biological computation