Spines Flashcards
(22 cards)
Experiments to see diffusion of spines + results
FRAP
Spines - slower, incomplete recovery
~50% retained (50% = immobile fraction)
Slower diffusion coefficient (initial rate of recover)
Soma - faster (larger diffusion coefficient), more complete recovery ~10% immobile fraction
Stubby spines compared to spines:
-Faster rate of recovery (AMPARs being replaced more rapidly)
Mushroom = slower rate of recover
Immobile fraction the same - same number of PSD-95 molecules???
How can we monitor actin polymerisation in dendritic spines?
FRET
CFP-tag G-actin monomers; YFP-tag other actin-monomers
CFP emission excites YFP = FRET
Hippocampal acute slices
HFS stimulation = increase FRET - increase in actin polymerisation
LFS stimulation = decrease FRET - decrease in actin polymerisation
Spine enlargement/shrinkage during LTP/LTD/basal
cLTP = spine growth
STED microscopy - show that LTP causes spines to get fatter + shorter
Why fatter? Fatter - less plastic compared to narrow neck - functional consequences are unclear
cLTD = spine shrinkage - decrease is less dramatic compared to LTP-induced increase
Basal:
- Pruning/deletion of spines = developmental = independent from LTP
- More spines are lost than gained -spine loss reduces with age
Spine-induced enlargements/shrinkages are a reversibly-induced phenomenon
Spine enlargement lags behind functional LTP (increase in EPSPs)
Actin
G-actin = monomers
F-actin = filamentous acting; polymer
Actin = ATPase
ATP-bound actin = increased affinity for being in a filament; ADP-bound acting = decrease affinity - dissociate from the complex
Requires a nucleus of 2/3 monomers to begin polymerisation (Arp2/3 = actin nucleator)
Monomers added to + end; dissociate from - end (depolymerise)
Treadmiling - + end beneath the membrane - polymerisation occurs, generating a physical force + pushing against membrane = enlargement
Capping factors = stabilise + end of the polymer
Severing factors ie. cofillin = promotes depolymerisation (polymer cleavage)
LTP spine enlargement = requires actin polymerisation
LTD spine shrinkage = required actin depolymerisation
Actin polymerisation also occurs with endocytosis - generate physical force to push the newly formed vesicle away from the PM; Arp2/3 independent actin polymerisation because PICK1 inhibits Arp2/3
Actin regulators
Activate = Arp2/3 (actin nucleator) - promote polymerisation
Stimulate depolymerisation/cleaves filaments = cofillin (severing protein)
Arp2/3 + experiments
Actin nucleator protein - promotes branched actin polymerisation
Activation - conformational change induces heterodimerisation of Arp2 + Arp3
Co-localises with PSD-95 = ensures that activation of actin occurs to cause spine enlargement
Activated by: WAVE-1
Inhibited by: PICK-1
Arp2/3 deletion with Cre-LoxP - dissociated neurones:
GFP-tagged actin, perform FRAP experiments - slower recovery of actin filaments in Arp2/3 deletion
BUT - do not know for sure that actin polymerisation is occuring (need FRET) - could just be exchanging actin monomers!
Arp2/3 deletion
Glutamate uncaging w/ HFS laser (LTP) = no spine enlargement - require Arp2/3 for spine increase
Loss of Arp2/3 prevents the maintenance of activity-induced spine enlargement!
cLTD (NMDA) = no change in spine reduction in the presence/absence of Arp2/3
**Contradicts other results (PICK1 K/O) = expected Arp2/3 deletion (therefore inhibited) to occlude LTD - PICK1 K/O (inhibits Arp2/3) prevents spine-induced shrinkage
Leads to a gradual loss of mushroom spines - more filapodia-like thin spines = over time ~ 8 weeks!
Arp2/3 is required for long-term spinal stability!
Need Arp2/3 to maintain the bulbous shape of the spine - required for long-term spine stability!
AND transgenic mice w/ Arp2/3 = show behavioural abnormalities (SZ-like)
WAVE-1
Activates Arp2/3
Requires 2x WAVE to activate Arp2/3
WAVE activation induces conformational changes in Arp2/3 (heterodimerisation of Arp2 + Arp3) = brings first 2 monomers to Arp2/3 complex to promote the formation of the daughter filament
Arp2/3 overcomes energy barrier to allow for the formation of new actin branch filaments
WAVE-KO
Enhanced LTP = unexpected - expected to inhibit LTD!
Maybe smaller spines;
-stronger/more effective Ca signal
-more room for spine enlargement by alternative mediators
WAVE K/O = enhanced NMDA:AMPA ratio
- increased NMDARs = favours LTP over LTD
- demonstrates that WAVE therefore also impacts molecular mechanisms of synaptic plasticity (not just actin dynamics)
Occludes LTD - Arp2/3 is inhibited in LTD
Inhibit actin polymerisation to decrease membrane tension in order to allow endocytosis to occur!
**Contradicts Arp2/3 deletion results which show that Arp2/3 deletion does not effect LTD!!!
PICK1-KO
LTD:
- Inhibits functional LTD (decrease in EPSCs)
- Inhibits spine shrinkage
Basal
-Increases basal spine size
PICK1 required:
- LTD to occur (GluA2 internalisation)
- Spine to shrink (Arp2/3 inhibition)
- Restrict spine size under basal conditions
CONTRADICT Arp2/3 deletion = expect to occlude LTD shrinkage but does not change
Kalirin
Rac GEF
Localised to the PSD via binding to a PDZ domain on PSD-95 - therefore enlargement (actin polymerisation) occurs at the PSD!
Mutant which is unable to bind PDZ domains = induces local formations of aberrant filapodia neurites
Activation of GIT
GIT = Arf GAP
GIT not activated = GTP-Arf = inhibits PICK1 = spine stability
GIT activated (NMDAR unknown activation)= GDP-Arf = disinhibition of PICK1 PICK1 inhibits Arp2/3 = spine shrinkage PICK1 = GluA2-internalisation
Evidence that GTPases are in spines + responsible for down-stream spine morphological changes
FRET:
CFP-Rac
YFP-PAK
Rac = activated, FRET increases
Glutamate uncaging (HFS) = FRET increases, spine enlargement With CamKII blocker = no FRET, no increase in spine enlargement With AP5 = no FRET, no increase in spine enlargement
Rac activation = CamKII + NMDA-dependent!!!!
Hippocampal memory task
Morris water maze = spatial memory
LTP-dependent
Disrupt GluA2-AP2 = prevent natural memory loss
Disrupt GluA2-NSF = speed up memory loss
Long-term memory maintenance reliant on stable GluA2-AMPARs
Entorhinal memory task
Novel object recognition
Rac
Rac GEFs = kaliriin
Rac K/O
- no LTP (LTP requires actin polymerisation)
- no spine enlargement
Hippocampus
-no LTP = deficit in spatial learning memory = Morris water maze
K/O = impaired in locating the platform
Therefore - GTP-bound Rac required for the mechanism of spatial memory learning in the hippocampus
How does membrane trafficking affect spine size?
Constitutive trafficking important - recycling endosomal pathway involved in the maintenance of spine size under basal conditions!
Over-express recycling endosomal markers = Rab11, Syntaxin13
- promote recycling process
- increase number of spines!!! and increase spine growth!
Mutants of Rab11/Synatxin13
- lose dendritic spines + shrink existing spines!
- also block LTP -no delivery of AMPARs!!!
Membrane trafficking affects spine size - during exocytosis, the vesicle membrane fuses with the plasma membrane = increase surface area (therefore increase capacitance)
Membrane trafficking from recycling endosomes is required to maintain spines!
RE trafficking during LTP
MyosinVb binds Ca = folded-to-extended/elongated conformational change (sedimentation assays) -expose binding site for RabII (RE marker)
Recuits REs into the dendritic spine
RE receptor
Transferrin - good marker for recycling endosome exocytosis
pH-luorin-tagged (GFP which fluoresces at neutral pH) transferrin receptor = fluorescence increases w/ LTP due to more receptors being inserted into the membrane
Show that fusion of recycling endosomes occurs following LTP
Performed with TIRF microscopy (Total Internal Reflection Microscopy)
LTP + Fusion
LTP = increased recycling of endosomes, increase fusion of vesicles, increased spine surface area, increased AMPAR insertion, development of new spines/enlarged spines
Techniques: TIRF of pH-luorin tagged transferrin receptors; capacitance studies to indirectly measure the increase in surface area = confirm that the vesicle membrane does fuse with the plasma membrane (as opposed to kiss+run events)
LTD = not been shown whether endocytosis of vesicles (membrane) is required in order for spine shrinkage to occur
Spine morphology + AMPARs
FOR:
Found via chimeric DNA constructs - NTD of GluA2 is responsible for spine morphology
- N-cadherin binds (adhesion molecules; have links to the actin cytoskeleton)
- co-immunoprecipitation studies = GluA2 + cadherin interact
- cadherins can modulate the activity of Arp2/3 = impact spine formation + stability
-DNA chimeric constructs =identified NTD responsible for spine enlargement
-2-photon excitation of caged glutamate caging w/ high spatiotemporal resolution to map glutamate receptors = measure glutamate sensitivities of different dendrites
-positive + strong correlation = size of spine head + glutamate sensitivity (no. of AMPARs)
-weak + neg. correlation = length of neck + glutamate sensitivity (filapodia = weak; mushroom = strong)
(identified NMDAR currents = removing Mg from solution; slower current)
- non-stationary fluctuation analysis = used to count the number of functional AMPARs at a dendrite (measurement of the single-channel currents composing a macroscopic current) ~46
- the expression of functional AMPARs has a strong correlation with spine-head volume
-hypothesis: silent synapses = ‘filapodia-like’ = small head + long neck
-explains why spine size changes lag behind expression of LTP/LTD
LTP = insertion of GluA1 -replaced later by GluA2
Spine morphology + AMPARs
AGAINST: =
Against:
-silent synapses shouldn’t ‘exist’
-insulin-induced endocytosis occludes NMDA-induced endo (same pool of AMPARs are internalised) BUT no spine shrinkage occurs
-LTD + spine shrinkage occur co-independently of eachother
Block LTD (GluA2-AP2) + LFS = still got spine shrinkage
Block spine shrinkage w/ peptide which blocks de-P cofilin (stabilise spine)
-basal recycling does not cause transient increases/decreases in spines
Block basal exocytosis with tetanus toxin = decrease in EPSPs but no decrease in spine size
-trafficking of AMPARs to the extra-synaptic site (reducing synaptic GluA2) does not cause spine shrinkage
Same study also found that actin depolymerisation is required for both processes!!!
Why for LTD/LTP????
-mediate AMPAR trafficking
-permit trafficking of key signalling mediators within LTP/LTD
-the actin polymerisation occurring is morphologically silent (no spine shrinkage/growth is observed)
Spines + behaviour
Barrel somatosensory cortex
- whiskers = segregation in somatosensory cortex
- GFP-labelled spines = image in vivo using a cranial window + 2-photon microscopy
- trim whiskers, lose synaptic input = increase spine turnover (more new spines + lost spines)
- external stimuli can affect dendritic spines
- novel sensory stimuli can drive the stabilisation of new spines within cortical neurones
Repetitive learning
- GFP-labelled spines, imaged in vivo via a cranial window
- repetitive, sustained training = form a cluster of new spines
- different tasks = different patterns of spine formation!!!!
- spatial clustering of motor actions = similar to striatum in PD = different motor tasks form different ‘clusters’ of SPNs firing in the striatum
Learning + eradicating a motor task
- modified GFP-optogenetic-Rac = acted as a probe to identify + label newly generated spines (potentiated spine marker)
- modified Rac = targeted to potentiated spines; can visualise using a cranial window + 2-photon microscopy in vivo
- shine blue light (prolonged Rac activation = induce spine shrinkage) = reverse spine enlargement = eradicates motor learning!!!
- therefore = spine structural plasticity is required for learning!!!!
Also required for learning = lactate produced by astrocytes = inhibit lactate production, inhibit learning + memory
