Transcript coding space – head direction cells

PART 4: BEHAVIORAL PLASTICITY
#25: SPATIAL NAVIGATION IN RATS II
 spatial learning
 cells that code for space
 synaptic plasticity in the hippocampus
 experiments that are knockouts
 summary
PART 4: BEHAVIORAL PLASTICITY
#26: SPATIAL NAVIGATION IN RATS II
 spatial learning
 cells that code for space
 synaptic plasticity in the hippocampus
 experiments that are knockouts
 summary
CODING SPACE – HIPPOCAMPAL PLACE CELLS
 place cells encode more than simple space
 T-maze, trained (fruit loops) to alternate L & R turns
 subset of place cells showed interesting pattern
 e.g., activity (sector 3) anticipating right turns only
 suggests hippocampal network represents episodic
memories, cells are small segments of an episode
 link of cells with overlapping episodes  memories
CODING SPACE – HIPPOCAMPAL PLACE CELLS
 spatial dreaming
 large # space cells
 only ~ 15% active in any 1 environ.
 some silent in one environ., active in others
 time- & labor-intensive to get larger picture
 device to measure 150 cells at once
 population or ensemble code
 code predicts rat behavior in maze
 many environments & codes
 overlapping, not interfering
 used to study plasticity...
CODING SPACE – HIPPOCAMPAL PLACE CELLS
 spatial dreaming
 plasticity
 strengthening of code  learning
 accompanied by reduced inhibitory activity
 does code relate to consolidated (permanent) memory
 trained rats in spatial task
 measured code during
 training
 sleeping before training
 sleeping after training
 dreaming replay of events  memory consolidation
CODING SPACE – HEAD DIRECTION CELLS
 navigation requires knowledge of
 place
 direction... another class of cells...
 in another structure... postsubiculum
 cells fire ~ head position
CODING SPACE – HEAD DIRECTION CELLS
 basic features of head direction cells
 retain direction preference in novel environments
 ~ 90° arc around preferred direction
 populations of cells with different preferences
 not ~ rat position in environment
 ~ independent of rat’s own behavior
CODING SPACE – HEAD DIRECTION CELLS
 common features of head direction cells & place cells
 influenced by salient external cues
 direction cells also fire after cues (light) removed
  capable of deduced reckoning
 using ideothetic cues
 informed by vestibular and visual input
 direction cells do not remap in a novel environments
CODING SPACE – HEAD DIRECTION CELLS
 navigation involves computation by the brain
 temporal process (~ video vs photograph)
 memory of past events
 prediction of future events
 processed by sub-populations of head direction cells
 2 areas measured in behaving rats
 postsubicular cortex (PSC)
 anterodorsal nucleus (ADN) of thalamus
CODING SPACE – HEAD DIRECTION CELLS
 navigation involves computation by the brain
 analyzed firing pattern relative to
 momentary head direction
 both cell types have preferred direction
CODING SPACE – HEAD DIRECTION CELLS
 navigation involves computation by the brain
 analyzed firing pattern relative to
 angular velocity
 PSC retain preference
 ADN shift preference  future position
CODING SPACE – HEAD DIRECTION CELLS
 navigation involves computation by the brain
 ADN shift preference  predict future position
 e.g., if a cell (of many) prefers 180° it may fire @
 160° when  180°
 200° when 180° 
 180° when @ 180°
(future = present)
CODING SPACE – HEAD DIRECTION CELLS
 why bother with all of this?... in theory...
 deductive reckoning circuit
 direction cells work by integrating internal cues
 ADN cells combine information about
 current head direction
 head movement (turning)
 proposed that PSC & ADN cells...
 constitute a looping circuit, compute direction by
 integrating motion/time
 but... how is “time” measured?
SYNATPTIC PLASTICITY IN THE HIPPOCAMPUS
 how do place cells and head directions cells
 learn to change their preferences?
 maintain their preferences over time?
 clues from electrophysiology experiments...
 brief, high-frequency stimulation of trisynaptic circuit...
 all 3 pathways
SYNATPTIC PLASTICITY IN THE HIPPOCAMPUS
  increased excitatory postsynaptic potentials (EPSPs)
in postsynaptic hippocampal neurons
 synaptic facilitation
 increase lasts for hours
 3 sites, 3 patterns, CA1 
 measured in brain
“slices”
 phenomenon called long-term potentiation (LTP)
 a very big deal in mammalian cell.-phys. of learning
 but... difficult to demonstrate relevance for behavior
SYNATPTIC PLASTICITY – LTP IN CA1
 3 properties of LTP in hippocampus CA1 neurons
 cooperativity: a minimum # of CA1 fibers must be
activated together (1 weak, 2 bottom strong)
SYNATPTIC PLASTICITY – LTP IN CA1
 3 properties of LTP in hippocampus CA1 neurons
 associativity: a weak tetanus paired with a strong
will gain - by association - value of strong
 measured in response after “training” (3 top)
 features ~ behavior, associative learning
SYNATPTIC PLASTICITY – LTP IN CA1
 3 properties of LTP in hippocampus CA1 neurons
 specificity: LTP can be restricted to single activated
pathway (2 bottom), others unchanged (2 top)
 localized to
 regions of hippocampus
 inputs regions on single cells (2)
SYNATPTIC PLASTICITY – LTP IN POSTSYNAPTIC CELLS
 CA1 pyramidal neurons
 LTP in CA1 is dependent on pyramidal neurons (PNs)
 inhibition of PN activity blocks LTP in CA1
 hyperpolarize PN membrane blocks LTP in CA1
 blocked inhibition of PN facilitates LTP in CA1
 depolarize PN membrane
 facilitates LTP in CA1 during weak tetanus
 not on its own (i.e., effect is associative)
 the postsynaptic cell must be depolarized for LTP to
occur in the presynaptic cell
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 glutamate (GLU), main excitatory transmitter (brain)
 N-methyl-D-aspartate (NMDA) 1 (of many) receptors
 LTP requires depolarization to open NMDA channel
 doubly gated channel, by.. GLU (receptor) & voltage
(sensor)
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 evidence for NMDA involvement in LTP
 NMDA blockers, e.g. aminophosphnovalerate (APV)
 blocks NMDA activity
 blocks LTP
 cooperativity: GLU from
 weak input  depolarize postsynaptic cell
 strong input  depolarizes postsynaptic cell
 associativity: GLU from
 strong input  depolarizes postsynaptic cell
 weak input (paired)  opens NMDA channels*
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 evidence for NMDA involvement in LTP
 Hebb’s Rule: synapses are strengthened if a
presynaptic cell repeatedly participates in driving
spikes in a postsynaptic cell
 GLU & NMDA receptor satisfies the rule
 have coincident activity of cells
 presynaptic release of GLU  receptors
 postsynaptic depolarization by non-NMDA
receptors
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 Ca++ influx into the postsynaptic cell is required for LTP
 block calcium (buffer)
 blocks LTP
 calcium influx through NMDA receptor/channel
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 evidence for NMDA involvement in LTP
 specificity: dendritic spines
 NMDA receptors on dendritic spine heads
  Ca++ entry restricted by necks
  anatomical subdivisions
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 evidence for NMDA involvement in LTP
 specificity: dendritic spines
 NMDA receptors on dendritic spine heads
  Ca++ entry restricted by necks
  anatomical subdivisions
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 Ca++ influx into the postsynaptic cell is required for LTP
 Ca++  LTP mediated by 2nd messenger signaling
 Ca++/calmodulin kinase (CaMKII)
 protein kinase C (PKC)
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 2 types of LTP described in CA1 neurons
 early-phase LTP (E-LTP)
13h
 cAMP & protein synthesis-independent
 late-phase LTP (L-LTP)
 10 h +
 cAMP & protein synthesis-dependent
 LTP in rats ~
 long-term synaptic facilitation in Aplysia
 long-term memory in Drosophila
SYNATPTIC PLASTICITY – LTP & NMDA RECEPTORS
 2 types of LTP described in CA1 neurons
 early-phase LTP (E-LTP)
13h
 cAMP & protein synthesis-independent
 late-phase LTP (L-LTP)
 10 h +
 cAMP & protein synthesis-dependent
 LTP in rats ~
 long-term synaptic facilitation in Aplysia
 long-term memory in Drosophila
SYNATPTIC PLASTICITY – LTP & SPATIAL LEARNING
 does LTP have anything to do with learning?... difficult
 spatial learning & memory in the water maze
 block LTP with AP5
 block memory
 ask the 3 Qs...
 correlation?
 necessity?
 sufficiency?
SYNATPTIC PLASTICITY – LTP & SPATIAL LEARNING
 does LTP have anything to do with learning?... difficult
 spatial learning & memory in the circular platform maze
 aging  LTP ~
 aging  memory
 ask the 3 Qs...
 correlation?
 necessity?
 sufficiency?
EXPERIMENTS THAT ARE KNOCKOUTS (MOUSE)
 genetic engineering - e.g. already with Drosophila
 transgenic “knockouts” (also “knockins”)
 single gene manipulations  LTP & spatial learning
 fyn gene knockout are tyrosine kinase– and...
 knockouts of CaMKII–
  LTP in CA1 cells
  spatial learning
 ask the 3 Qs...
 correlation?
 necessity?
 sufficiency?
EXPERIMENTS THAT ARE KNOCKOUTS (MOUSE)
 CaMKII knockouts - enzyme cannot be Ca++ modulated
 LTP impaired (in “functional” range)
 place cells
 fewer
  specificity
 focus  stable
 platform maze
  spatial learning
 ask the 3 Qs...
EXPERIMENTS THAT ARE KNOCKOUTS (MOUSE)
 NMDA receptor knockouts
 LTP severely impaired
 place cells (multi-elect.)
  specificity
  coordinated firing
 NMDA-receptor-mediated synaptic plasticity
required for proper representation of space in
CA1 region of hippocampus
EXPERIMENTS THAT ARE KNOCKOUTS (MOUSE)
 NMDA receptor knockouts
 water maze
  spatial learning
 ask the 3 Qs...
 arguments more compelling with each experiment
 spatial & temporal targeting of knockout,
correlation of lesion, LTP, behavior remains
SUMMARY
 spatial navigation uses 2 types of cues
 external (landmarks)
 internal (ideothetic)  deductive reckoning (memory)
 spatial navigation studied in rats using
 radial arm maze
 T-maze
 water maze
 circular platform maze
SUMMARY
 tasks are designated as
 spatial (using distal cues)
 cued (or non-spatial, using proximal cues)
 lesion studies, hippocampus  for spatial learning
 if lesions precede learning
 working & reference memory tasks are impaired
 cued tasks are not impaired
 if learning precedes lesions
 time between events important
 usually older memories are less affected
SUMMARY
 two classes of neurons encode space
 place cells, CA1 hippocampus
 firing field
 stability ~ weeks, memory
 influenced by
 external cues (landmarks)
 internal cues (vestibular, visual ~ motion)
 field in dark ~ active
 can be event-related, predictive (e.g., turning)
 work together  ensemble code
 replay in sleep... consolidation?... dreaming?
SUMMARY
 two classes of neurons encode space
 head direction cells, CA1 hippocampus
 fire ~ head direction
 similarly influenced by
 external cues (landmarks)
 internal cues (vestibular, visual ~ motion)
 2 types of cells
 PSC cells encode current direction
 ADN cells encode future direction
SUMMARY
 LTP is a prominent form of hippocampal synaptic
plasticity, with the following properties:
 cooperativity
 associativity
 specificity
 LTP in CA1 neurons ~ NMDA receptor, 2 requirements:
 depolarization of the postsynaptic cell
 binding of glutamate with the NMDA receptor
 allows channel opening, Na+ & Ca++ influx
 Ca++ influx is required for induction of LTP
SUMMARY
 NMDA receptor  mechanism for Hebb’s Rule
 Evidence that LTP underlies (or is involved with)
mechanisms for learning
 drugs blocking LTP also block spatial learning
 aging affects LTP and spatial learning
 mice knockouts for “LTP genes” show deficits in
 LTP
 place cell properties
 spatial learning