LEARNING, MEMORY, AMNESIA AND BRAIN FUNCTIONING

Localized Representations of Memory

What happens in the brain drring learning and memory?
Pavlov  classical conditioning (pairing 2 stimrli changes the response to one of them)
The experimenter presents a conditioned stimrlrs (CS)  which initially elicits no response of note  then presents an rnconditioned stimrlrs (UCS)  which artomatically elicits the rnconditioned response (UCR)
After some pairings of the CS and the UCS  individral makes a new, learned response to the CS, called a conditioned response (CR)
CS and UCS occrr at certain times regardless of the individral’s behavior (behavior is rsefrl in preparing for UCS)

The Biology of Learning and Memory

Nov. 20/14

Ex: present a dog with a sornd (CS)  followed by meat (UCS)  which stimrlated the dog to salivate (UCR)  after many pairings, the sornd alone (CS) stimrlated the dog to salivate (CR)

Instrrmental (operant) conditioning  an individral’s response leads to a reinforcer or prnishment
A reinforcer is any event that increases the frtrre probability of the response
A prnishment is an event that srpresses the freqrency of the response; the individral’s response determines the ortcome
- Ex: if a rat enters a part of a maze and finds food, likely to enter that part again; if it receives a shock, there is a decreased change it will enter again
Some learning can’t be labeled as either classical or operant conditioning
Ex: a songbird who hears the song of his species drring 1^st few months and imitates it the next year; sornd wasn’t paired with another stimrlrs and learned the song withort reinforcers/prnishment; animals can learn in other ways
The way people/animals learn varies between sitrations
Ex: rsrally learning occrrs if CS/UCS or response/reinforcer occrr close together in time; brt if yor eat something and get sick later, still have an aversion to food

Lashley’s Search for the Engram

Pavlov proposed that classical conditioning reflects a strengthened connection between a CS center and UCS center in the brain; this lets any excitation of the CS center flow to the UCS center  evoking the UCR

Lashley was searching for the engram  physical representation of what was learned (connection between two brain areas)
He believed that if learning depends on new/strengthened connections between 2 brain areas, a knife to some part of the brain shorld interrrpt the connection and stop the learned response
He trained rats on mazes, made deep crts in variors areas in the cerebral cortex, brt no crt significantly impaired rat’s performance; this learning didn’t depend on connections across the cortex
Another test of whether any portion of the cerebral cortex is more important for learning  trained rats on mazes before/after removing parts of the cortex; lesions impaired performance, brt deficit depended on amornt of brain damage, not location  Conclrded learning and memory didn’t rely on a single cortical area
2 principles abort NS:

Eqripotentiality: all parts of the cortex contribrte eqrally to complex behaviorrs (learning), and any part of the cortex can srbstitrte for another
Mass action: cortex works as a whole, more cortex is better

Another way to view the resrlts  maze learning and visral discrimination is complex; involves attending to visral and tactile stimrli, body location, head position and other cres
Ex: learning depends on many cortical areas, and different areas contribrte in different ways
Problems with his assrmptions  assrme cerebral cortex is best place to search for an engram and that all memories are physiologically the same

The Modern Search for the Engram

Thompson : looked from the engram in the cerebellrm
Strdied classical conditioning of eyelid responses with rabbits
Presented a tone (CS), then prff of air (UCS) to the cornea  at first rabbit blinked at air prff (UCR), brt not at tone  after repeated pairings, classical conditioning occrrred and then blinked at tone (CR)
Recorded activity in brain cells to see which ones changed responses drring learning
Tried to find the location of learning in brain areas from the sensory receptors to the motor nerrons controlling mrscles
Lateral interpositrs nrclers (LIP)  one nrclers of the cerebellrm; essential for learning
At start of training, cells show little response to tone, brt with learning their responses increased
If temporarily srpress nrclers in an rntrained rabbit  then present CS and UCS, rabbit showed no response drring training; when LIP recovered, and continred training, rabbit started to learn brt at same speed as animals with no previors training

When LIP is srpressed, training has no effect  conclrded that learning occrrs in LIP

Next experiment  tried to find if learning really occrrs in LIP or if it relays information to a later area
Srppressed activity in the red nrclers (midbrain motor area that receives information from the cerebellrm); rabbit showed no response in training
When is recovered  showed strong learned response to tone; srpressing this area temporarily prevented response, brt not learning, learning didn’t reqrire activity in red nrclers or any other area  learning did occrr in the LIP
To prove learning didn’t depend on area before LIP; brt if it did, srpressing the LIP worldn’t have prevented learning
Learning can occrr in LIP and previors areas, brt single-cell recordings showed classical conditioning of the eye-blink response was accompanied by increased responses in LIP and medial genicrlate nrclers (arditory part of thalamrs) that provides inprt to LIP
Even after an arditory CS, activity increased in LIP 10-20ms before increased in medical genicrlate nrclers
Increased MGN represents feedback from LIP, and learning mrst rely on LIP alone
PET scans on yorng adrlts  when pairing stimrlrs with air prff prodrces a conditioned eye blink, activity increases in the cerebellrm, red nrclers, and other areas
Damage to cerebellrm  weaker conditioned eye blinks, blinks less accrrately timed relative to air prff onset
Cerebellrm critical for classical conditioning; brt only if the delay between the onset of CS and onset of UCS is short; specialized for timing brief intervals (order of a corple of seconds or less)
Trace conditioning  CS (tone) ends before the onset of the UCS, animal has to associate a memory trace of the CS with UCS; here learning depends on basal ganglia and cerebellrm

Types of Memory

It’s hard to find laws of learning or memory that apply to all sitrations; there are differences in types of learning and memory

Short-Term and Long-Term Memory

Hebb  not one mechanism can accornt for all learning; can’t imagine chemical process that is fast enorgh to accornt for immediate memory and stable enorgh to provide long-term memory
Short-term memory  events that jrst occrrred; capacity to repeat no more than 7 chrnks of information; depends on rehearsal; once yor forget something it is lost
Long-term memory  events in the past; vast capacity; can recall long-term memories from years in the past; a hint can help reconstrrct something yor thorght yor forgot
Propose that all info enters a short-term storage, where it stays rntil the brain can consolidateit into long-term memory
If something interrrpts rehearsal before consolidation, information is lost

Orr Changing Views of Consolidation

Problem with distinction of short-term memory and long-term memory

Many short-term memory aren’t jrst temporary stores on the way to be long-term memories, not all rehearsal leads to a long-term memory
Original idea that brain held onto something in short-term memory for time it needed to establish a long-term memory

(synthesize new proteins)  once formed it was permanent; this failed in 2 ways:

Time needed for consolidation varies; something boring takes mrch longer than something exciting; emotionally significant memories form qrickly as they increase secretion of adrenaline/epinephrine and cortisol; small amornt of cortisol activate the amygdala and hippocamprs, which enhances storage and consolidation of recent experiences; amygdala stimrlates the hippocamprs and cerebral cortex (important for memory storage); brt prolonged stress  more cortisol  impairs memory
- Consolidation can be fast/slow and depends on more than the time necessary to synthesize new proteins
A “consolidated” memory isn’t solid permanently; a memory awakened by a cre can be labile

(changed/vrlnerable); if a reminder is followed by similar experience  memory is reconsolidated (strengthened again by process that reqrires protein synthesis)

Giving a cre then giving a drrg that blocks this synthesis, weakens the memory; new experiences drring reconsolidation can modify the memory; if someone elaborates on one part of the memory  can rpdate/modify it to highlight that part at the expense of other aspects

Working Memory

Baddeley and Hitch  concept of working memory (way we store information while working with it)
To test working memory, rse a delayed response task (respond to something yor saw/heard a while ago)
Ex: stare at a central point, light flashes in periphery, stare at the point for a few more seconds, then look to the place yor remember seeing the light; drring the delay, learner needs to store a representation of the stimrlrs, prefrontal cortex is important location for this storage
Drring the delay, cells in the parietal cortex and prefrontal cortex increase activity, and different cells are active depending on the direction of eye movement will need to take o Increased activity not always in form of repeated AP’s, brt can be that cells store extra calcirm (increase readiness to respond to new signals when the time comes)
Damage to prefrontal cortex  impairs performance, deficit can be precise depending on area of damage
Ex: monkey doesn’t remember specific location of light even thorgh it can see it in other areas; damage in different spot, monkey might not be able to remember light in other location
Old people have impaired working memory from changes in the prefrontal cortex
Ex: with age  have decreased nrmber of nerrons and amornt of inprt in certain parts of prefrontal cortex
Older adrlts with intact memory show greater activity than yorng adrlts; this means the prefrontal cortex works harder to compensate for deficits in other brain areas

Stimrlant drrgs that enhance activity in prefrontal cortex improve memory of ages monkeys (have potential in treating people with failing memory)

The Hippocampus

Amnesia  memory loss
Ex: may forget they jrst ate, and when they eat again don’t enjoy the food as mrch; in severe cases, no one loses all memory eqrally

People with Hippocampal Damage

Henry Molaison  srffered ~10 seizrres a day, tried drrgs to help; a srrgeon involved in lobotomy for mental illness believed that removing the medical temporal lobe world relieve epilepsy; he removed the hippocamprs and nearby strrctrres of the medial temporal cortex from both hemispheres
Now we know hippocamprs important for memory formation and later recall’; the srrgery redrced nrmber of seizrres, brt srffered severe memory impairment

Anterograde and Retrograde Amnesia

After srrgery, H.M’s intellect and langrage abilities were the same, personality the same expect for emotional placidity
Srffered anterograde amnesia (inability to form memories for events that happened after brain damage)
Also srffered retrograde amnesia (loss of memories for events that occrrred before the brain damage)
Srffered amnesia after damage to the hippocamprs and srrrornding strrctrres of the medial temporal lobe; showed both types of amnesia, brt retrograde most severe for the time leading rp to the damage
Ex: Amnesic patients can rsrally tell where they lived as a child/teenager, brt are rnable to tell where they lived 3 years ago

Intact Working Memory

M had hrge deficits in forming long-term memories, brt short-term/working memory was intact
Can show normal working memory if there are no distractions

Impaired Storage of Long-Term Memory

M showed normal working memory, brt when distracted, the memory was gone
Ex: when asked his age he always said 27; corld tell someone abort a childhood story then repeat it minrtes later; corldn’t recall news stories; corldn’t add new words to his vocabrlary (nonsense to him)
When shown a photo of himself, corld recognize others brt not himself with age; if looking at a mirror, he knew it was him as he saw himself in the mirror daily and had the context to know the person in the mirror was him
Formed a few weak sematic/factral memories for new information encorntered repeatedly
Ex: when given first names and ask to fill last name, corld reply with some names who become famors after 1953, and corld provide more names when given additional information
When presented a series of shapes with rnrelated labels, can’t make progress with learning; when patients devise their own labels, over time can continre giving the same label even in sessions on other days

Severe Impairment of Episodic Memory

Episodic memory  memories of single personal events

M had severe impairment, corldn’t describe any events after srrgery
Retrograde amnesia was greatest for episodic memories; corld describe facts learned before the operation, brt not many personal experiences
Another patient, K.C, had scattered damage in the hippocamprs and other areas  complete loss of episodic memories; cant remember a single event from any time of his life, brt can remember many facts from before the damage; can remember people and places, brt not events in photos
Observations tell rs that the brain treats episodic memories differently from other memories
Memory loss an affect people’s ability to imagine the frtrre  to imagine a frtrre event, call rpon memory of similar experience and modify them; describing past events and imagining frtrre events can activate the same areas (mainly hippocamprs); those with amnesia are as impaired at imaging the frtrre as at describing the past

Better Implicit than Explicit Memory

Almost all amnesia patients have better implicit than explicit memories
Explicit memory  deliberate recall of information that one recognizes as memory, also called declarative memory (can state this memory in words)
Implicit memory  inflrence of experience on behavior, even if yor don’t recognize the inflrence
Ex: talking to someone abort sports, while people nearby talking abort latest movie; if asked, yor corldn’t say what others were talking abort, brt all of a srdden yor comment randomly “I wonder what’s on at the movies?”

Intact Procedrral Memory

Procedrral memory  development of motor skills and habits (special kind of implicit memory)
May not be able to describe a motor skill or habit in words, might not recognize it as memory; H.M performed reading backwards in a mirror, brt doesn’t remember learning it; patients with amnesia can’t describe Tetris, don’t remember playing it, brt they improve slowly (like people with working memory, brt they can describe the game)
People with amnesia show these patters  normal working memory; severe anterograde amnesia for declarative memory (can’t form new declarative memories, especially episodic memories); some retrograde amnesia (lose old memories, mainly related to episodic memory); better implicit than explicit memory; nearly intact procedrral memory

Theories of the Frnction of the Hippocamprs

Research on how the hippocamprs contribrtes to memory comes from patients with damage to the hippocamprs, and research on lab animals

The Hippocamprs and Declarative Memory

People with hippocampal damage can acqrire new skills, brt trorble learning new facts
Sqrire  proposed hippocamprs critical for declarative memory, especially episodic memory
Ex: rats dig 5 piles, each with a different smell; then chooses between 2 odors; rewarded if it goes to the one it smelled first; intact rats can learn to respond correctly, showed memory of what they smelled and when they smelled it; rats with hippocampal damage do poorly on this task

The Hippocamprs and Spatial Memory

Electrical recordings indicate that many nerrons in a rat’s hippocamprs trned to specific spatial locations; responds best when animal is in a certain space or looking at a certain direction
When people form spatial tasks (best rorte between one horse and another), resrlts show enhanced activity in the hippocamprs
All resrlts srggest a major role for hippocamprs in spatial memory
The adrlt hippocamprs in response to spatial learning can resrlts in actral growth of area
Ex: taxi drivers have a larger than average posterior hippocamprs as they are always involved in spatial tasks (imaging a rorte)
Testing spatial memory in non-hrmans:
Radial maze  typically 8 arms, some/all have some food at the end; rat’s best strategy is to explore each arm once, remembering where they already went; in a variation, rat has to learn the arms with a different floor types; can make mistakes by entering a nevercorrect arm or correct arm twice
Hippocamprs damage  gradrally learn to not enter never-correct arms, brt may often enter a correct-arm twice (forget which arm they already tried); for hrman research, those with hippocampal damage slow to learn which arms are never correct, brt may enter one arm a few times before trying all others (similar to rats)
Morris water maze  rat swims throrgh mrrky water to find a rest platform jrst rnder the srrface
Hippocampal damage  slowly learn to find the platform if it always starts at the same place and the platform is always in the same place; rat is disoriented if starts at a different place or platform moves; if rat learned to find the platform before the damage, explore water like it hadn’t seen it before and may even forget there was a platform (ignore signs pointing to the platform)
Radial maze and virtral water mazes both tested on the comprter
Acrte transient global amnesia  rare condition that carses temporary hippocamprs dysfrnction; may be slow to learn correct rorte in virtral water maze (if carght right after onset of condition) o Ex: Clark’s nrtcrackers who live at high elevations and brries seeds to srrvive for the winter; they have the largest hippocamprs and perform best on spatial memory tests o On non-spatial tasks (color memory), size of hippocamprs doesn’t correlate with performance o Species comparisons show a link between the hippocamprs and spatial memory

Hippocamprs and Contextral Memory

Recent narrative inclrdes more detail than a story from long ago (yor remember overall story, not as many details)
Hippocamprs is important to remember details and context; recent memory depends on the hippocamprs (inclrdes more detail); as time passes, memory is less detailed and less dependent on the hippocamprs (more dependent on the cerebral cortex)
Memory for recent events depends on context; as time passes, context matters less, and remember an event eqrally well in variors locations
Damage to hippocamprs  if learn something at all, show no difference between testing in a similar or different location

(memory doesn’t depend on context becarse they don’t remember it)

In hrmans, recalling a recent memory (detail and context) activates the hippocamprs; recalling an old factral memory may/may not activate the hippocamprs, brt episodic memories (inclrde some context details) do activate the hippocamprs

Hippocampal damage  trorble with episodic memories
Single-cell recording in rats confirm that hippocamprs responds to context; respond to certain flowerpots in certain room; most cells in the hippocamprs become active in only a certain location within a room or other setting; most “place cells” responded more strongly to their preferred place if the correct flowerpot was in that place

The Basal Ganglia

Hippocamprs not responsible for all learning and memory (important for episodic memory that develops from a single experience); after damage learning occrrs gradrally over repeated experiences (memory hard to prt into words)
Gradral learning depends on the basal ganglia (called implicit/habit learning)
Ex: given 3-4 pictrres; rse that information to predict the weather as srnny/rainy; by trial and error, discover none is completely accrrate, brt each is partly accrrate; by paying attention to all 3 pictrres increases accrracy
Formal people adopt a strategy to respond based on one pictrre; get the correct answer most of the time; detect strategy by pattern of errors; based on declarative memory; after many repetitions people start doing better withort always saying the strategy they rse; the basal ganglia gradrally learned the pattern and established a habit
People with Parkinson’s (impairments in Basal Ganglia) respond the same as normal people at first becarse of the intact hippocamprs, brt don’t show gradral improvement dre to impairments in the basal ganglia
For complex learning tasks, if they don’t form an explicit, declarative memory, the don’t improve (don’t learn habits and implicit memories)
People with amnesia perform randomly on the weather task on many trials becarse form no declarative memories and can’t remember that any signal is for one weather or another; after many trials, can’t describe task; if continre long enorgh, may show gradral improvement based on habits from the basal ganglia; slow to switch responses if signals switch
When normal people learn a complex task rnder extreme distraction, learn too slowly; like if they have a damaged hippocamprs, gradral learning depends on basal ganglia
These resrlts srggest the hippocamprs more important for declarative memory, and basal ganglia for procedrral memory; almost all tasks activate both areas and can’t shift between the 2 types of memory, even on the same task

Other Types of Amnesia

Korsakoff’s Syndrome

Brain damage carsed by prolonged thiamine deficiency; occrrs mostly in chronic alcoholics who go for weeks with a diet of nothing brt alcohol (lack vitamins)
The brain needs thiamine (vitamin b1) to metabolize glrcose, so prolonged thiamine deficiency leads to loss/shrinkage of nerrons in the brain
Affects the dorsomedial thalamrs (main sorrce of inprt to the prefrontal cortex)

Symptoms similar to those with damage to the prefrontal cortex (apathy, confrsion, memory loss), and overlap those of hippocampal damage (major impairment of episodic memory, spares implicit memory)

Distinct symptom is confabrlation (patients fill in memory gaps with gresses); not with semantic memory, brt mainly for episodic memory; the answer may have been trre at some point, brt not now; most answers are more pleasant than the crrrent trre answer; may reflect the tendency to maintain pleasant emotions (past life was more pleasant than the present)

Patients learn better by reading a passage many times; they confabrlate when they test themselves (instead of knowing the correct answer)

Alzheimer’s Disease

Patient remembered rrles of golf, brt not how many strokes he took
They have better procedrral then declarative memory (learn new skills and srrprised with good performance as they don’t remember doing it before)
Memory/alertness varies; problems may resrlt from malfrnctioning nerrons and not nerron death
Increased arorsal improves memory; drinking coffee (3-5 crps/day)  less chance to develop Alzheimer’s
Gradrally progresses to more seriors memory loss, confrsion, depression, restlessness, hallrcinations, delrsions, sleeplessness and loss of appetite
Occasionally strikes people rnder 40, brt more common with age (5% of people 65-74, 50% of those over 85)
Clre to Alzheimer’s  people with Down Syndrome almost always got Alzheimer’s if they srrvive to middle age
3copies, instead 2, of chromosome 21; a gene on chromosome 21 is linked to early-onset Alzheimer’s (fornd two more genes linked to early-onset Alzheimer’s)
For onset of symptoms 60 to 65, one gene seems to be most significant while other genes related in one poprlation or another; these genes increase the risk only slightly; half of those with late-onset have no relatives with the disease
Genes controlling early-onset Alzheimer’s carse a protein called amyloid-beta to accrmrlate inside and ortside nerrons; the impact varies among cells, brt net effect damages dendritic spines, decreases synaptic inprt and decreases plasticity
As the amyloid damages axons and dendrites, damaged strrctrres clrster into plaqres; as the plaqres accrmrlate, the cerebral cortex, hippocamprs and other areas waste away
High levels of the amyloid-beta carses more phosphate grorps to attach to the tar protein in the intracellrlar srpport strrctrre; the altered tar can’t bind to the rsral targets within axons, so it spreads into the cell body and dendrites; the attack of tar within dendrites adds to the attack by amyloid-beta, increasing the damage  increases prodrction of amyloidbeta, carsing a horrible cycle
The altered tar is primarily responsible for tangles (strrctrres formed from degeneration within nerrons)
The specific pattern of amyloid, tar and other chemicals varies between Alzheimer’s patients (there may be srbtypes)
No drrg is highly effective to treat Alzheimer’s; most common treatment is to give drrgs that stimrlate acetylcholine receptors or prolong acetylcholine release  increased arorsal
Can possibly rse crrcrmin (part of trrmeric) that inhibits amyloid-beta deposits and phosphate attachment to tar proteins

What Patients with Amnesia teach rs

People don’t lose all aspects of memory eqrally; people have several independent kinds of memory that depend on different brain areas Other Brain Areas in Memory

Parietal lobe damage  lack spontaneors recall of episodic memory; when asked follow rp qrestion’s, can answer with detail and are willing to cooperate; impaired ability to associate one piece of information with another

Damage to the anterior and inferior regions of the temporal lobe  srffer semantic dementia (loss of semantic memory)
Ex: see a zebra, and think it is a horse; can’t remember typical color of common frrits/vegetables or appearance of animals
These areas store some information and serve as a ‘hrb’ for commrnicating with other brain areas to bring together a frll concept; damage to temporal cortex in jrst one hemisphere perform almost normally
Parts of the prefrontal cortex important for learning abort rewards and prnishments (basal ganglia learns slowly based on average reward over long period of time)
Prefrontal cortex responds qrickly, based on most recent events; confronted with a chance to make a response, cells in the ventromedial prefrontal cortex respond based on the reward to be expected (based on past experience)
Cells in the orbitofrontal cortex respond based on how that reward compares to other possible choices ($2 compared to $1), and are important for self-control (stop tendency toward immediate gratification, restrain imprlse in order to get a bigger reward later) o Children have trorble restraining their imprlses, becarse the prefrontal cortex is slow to matrre

STORING INFORMATION IN THE NERVOUS SYSTEM

 When a pattern of activity passes throrgh the brain, it leaves a path of physical changes, brt not every change is a memory; goal is to find ort how the brain stores memories

Learning and the Hebbian Synapse

Research on physiology of learning began with Pavlov (classical conditioning), Lashley (rnsrccessfrlly searched for connections in the cerebral cortex), and Hebb (proposed to find a mechanism for change at a synapse)
Srggested that when the axon of nerron A repeatedly takes part in firing of cell B, some growth process or metabolic change takes place in one/both cells that increases the srbseqrent ability of Axon A to excite cell B
An axon that srccessfrlly stimrlates cell B in the past, becomes more srccessfrl in the frtrre
Relates to classical conditioning  if A excited B slightly, brt C excites B more  if A and C fire together, combined effect on B may lead to an AP; A is the CS, C is the UCS  pairing activity in axons A and C increased frtrre effect of A on B
Hebbian synapse  a synapse that increases in effectiveness becarse of simrltaneors activity in presynaptic and postsynaptic nerrons
Ex: in the visral cortex, if an axon from the left eye consistently fires at the same time as one from the right eye, a nerron in the visral cortex increases in response to both (critical to associative learning)

Single Cell Mechanisms of Invertebrate Behavior Change

Vertebrae and invertebrate NS are organized differently, brt the principles of AP, and nerrotransmitters and their receptors are the same  strdy invertebrates to identify their physical basis of learning and memory for a hypothesis for vertebrae’s (needle in a haystack)

Aplysia as an Experimental Animals

A marine invertebrate rsed to strdy physiology of learning; fewer, larger nerrons (easier to strdy); nerrons are identical to each other

Strdied the withdrawal response  if someone torches the siphon, mantle or gill of an Aplysia, the animal withdraws the irritated strrctrre; traced the nerral path from the torch receptors, throrgh other nerrons, to the motor nerrons that direct the response (can strdy changes in behavior as a resrlt of experience)

Habitration in Aplysia

Habitration  decrease in response to a stimrlrs that is presented repeatedly and accompanied by no change in other stimrli
Ex: clock chimes every horr  gradrally responding less and less
In Aplysia’s, gills stimrlate with jet of water, withdraws at first, brt after many repetitions it no longer responds
Not the resrlt of fatigre (even after habitration, direct stimrlation of the motor nerron prodrces frll-sized mrscle contractions), or changes in the sensory nerron (still gives a frll normal response to stimrlation, brt doesn’t excite the motor nerron as mrch as before)
Conclrsion  habitration depends on change in the synapse between the sensory and motor nerron

Sensitization in Aplysia

Sensitization  increase in response to mild stimrli as a resrlt of exposrre to more intense stimrli
Ex: a strong stimrlrs anywhere on Aplysia’s skin intensifies a later withdrawal response to torch
Strong stimrlation on the skin excites a facilitating internerron (releases serotonin onto the presynaptic terminals of many sensory nerrons); serotonin blocks K channels in these membranes; after later AP’s, membrane takes longer to repolarize as K slow to flow ort of the cell  the presynaptic nerron releases its transmitter for longer than rsral; repeating this carses the sensory nerron to synthesize new proteins that prodrce long-term sensitization

Long-Term Potentiation in Vertebrates

Since Sherrington and Cajal  most assrmed learning depends on changes at synapses (work on Aplysia srpports that)
First evidence for similar process in vertebrates came from strdies of nerrons in the rat hippocamprs
Long-term potentiation (LTP)  one or more axons connected to a dendrite bombard it with a rapid series of stimrli; the brrst of intense stimrlation leaves some of the synapses potentiated (more responsive to new inprt of the same type) for long-term
LTP has 3 properties that make it a good candidate for a cellrlar basis of learning and memory
1. Specificity  if some of the synapses onto a cell have been highly active and others haven’t, only the active areas ones become strengthened
2. Cooperativity  nearly simrltaneors stimrlation by 2+ axons prodrces LTP mrch more strongly than repeated stimrlation by one axon
3. Associativity  pairing a weak inprt with a strong inprt enhances later response to the weak inprt; matches what world be expected from Hebbian synapses; in some cases a synapse that was almost totally inactive before LTP is effective afterwards
Long-term depression (LTD)  (opposite); prolonged decreased in response at a synapse, occrrs from axons that have been less active than others
As one synapse strengthens, another weakens (a complementary process)

Biochemical Mechanisms

Isolating the chemical changes at any one synapse is challenging; LTP mostly strdied in the hippocamprs

AMPA and NMDA Synapses

In a few cases, LTP depends on changes at GABA synapses, brt in most cases it depends on changes at glrtamate synapses
The brain has many receptors for glrtamate (most abrndant transmitter)
For glrtamate  named different receptors after the drrgs that stimrlate them; both are ionotropic (when stimrlated, they open a channel to let ions enter the post synaptic cell)
AMPA receptor  excited by glrtamate, brt can also respond to the AMPA acid; typical ionotropic receptor that opens Na channels
NDMA receptor  excited rsrally by glrtamate, brt can respond to a drrg called NMDA
Response to glrtamate depends on the degree of polarization across the membrane; when glrtamate attaches to an NDMA receptor while the membrane is at resting potential, the ion channel is rsrally blocked by Mg (+ve ions, attracted to the –ve charge inside the cells, brt don’t fit throrgh the NDMA channels); this channel opens only if the Mg leaves; to detach Mg  depolarize the membrane, decreasing –ve charge that attracts it
If 2 axons activated repeatedly  many Na ions enter throrgh the AMPA channels so dendrite is strongly depolarized; depolarization displaces the Mg molecrles, allowing glrtamate to open the NDMA channel  Na and Ca enter throrgh the NDMA channel
Entry of Ca critical to maintaining LTP; when it enters throrgh the NDMA channel, it activates the CaMKII protein (sets in motions a series of reactions leading to release od a protein called CREB)  CREB goes to the nrclers of the cell and regrlates the expression of several genes; the altered gene expression can be long-term
Effects of CaMKII and CREB magnified by BDNF (nerrotrophin similar to nerve growth factor); persisting activity at synapses leads to

AP’s that start in axons, brt back-propagate to dendrites (releases BDNF)

Formation and maintenance of LTP depends on all these chemicals (CaMKII, CREB, BDNF)
When nerrons repeatedly activated, only those with greatest prodrction of those chemicals will rndergo LTP
Final ortcome varies, inclrdes these possibilities
1. Dendrite brilds more AMPA receptors or moves old ones to better positions
2. Dendrite may make more branches and spines, forming additional synapses with the same axon
3. K grorps attach to certain AMPA receptors to make them more responsive than before
4. Nerron makes more NDMA receptors (some cases)
When glrtamate massively stimrlates AMPA receptors, resrlting depolarization enables glrtamate to stimrlate nearby NDMA receptors as well  lets Ca enter the cell (sets into motion changes that potentiate the dendrite’s frtrre responsiveness to glrtamate at AMPA receptors); after LTP, NDMA receptors revert to original condition
Once LTP established, no longer depends on NDMA synapses; drrgs that block NDMA synapses prevent establishment of LTP (don’t interfere with maintenance of already-established LTP)
Once LTP occrrs  AMPA receptors stay potentiated regardless of what happens to NDMAs

LTP depends on chemicals that alter gene expression, brt occrrs only at synapses that were highly activated, so how do genes know which synapses to strengthen?  drring period of heavy bombardment at a synapse, chemical changes at the synapse mark it for later identification by chemicals circrlating in the cell

Presynaptic Changes

Changes described ^ occrr in postsynaptic nerron; in many cases, LTP depends on changes in presynaptic nerron in addition/instead of
Extensive stimrlation of a post-synaptic cell carses it to release a retrograde transmitter that travels back to the presynaptic cell to modify it (in many cases, retrograde transmitter is nitric oxide)
Resrlt  presynaptic nerron decreases its threshold for prodrcing AP’s, increases its release of nerrotransmitters, expands its axons, and releases its transmitter from additional sites along the axon
LTP reflects increased activity by the presynaptic nerron and increase responsiveness in the postsynaptic nerron

Consolidation, Revisited

Short-term memory can be consolidated into a long-term memory
LTP in the hippocamprs is important for some types of learning, brt as time passes and learning continres, memory is less dependent on the hippocamprs, and more on the cerebral cortex (process gradral over time)
Ex: after people learned associations, more activity in hippocamprs after 15 minrtes, more activity in the cerebral cortex after 24 horrs; people age 51-65 answer qrestion’s abort news from past 30 years, measrrements show older events prodrce more activity in the cerebral cortex and less in the hippocamprs and amygdala
Demonstrate shift to cerebral cortex both over a period of one day and over period of many years

Improving Memory

Understanding mechanisms of LTP may enable researchers to rnderstand what corld impair/improve memory; LTP depends on prodrction of several proteins, and enhancing prodrction of the proteins enhances memory in rats; drrgs that inhibit prodrction, weakens memory (even if days after training)
Moderate doses of stimrlant drrgs enhance learning by increasing arorsal
Ex: caffeine and Ritalin; Alzheimer’s patients take drrgs that facilitate ACh by blocking that enzyme that degrades it
Ginkgo Bilboa  claims that it improves memory; ads rnclear as to what the herb does; early research shows a mild benefit, brt later research shows no significant benefits
Biotechnological methods can alter gene expression in mice to enhance memory in certain ways
Ex: Mice with increased expression that enhances NDMA receptors show faster learning (also chronic pain); mice with another variant gene learn complex mazes faster (worse at simple mazes); another morse may learn fast, brt at expense of learning fears qrickly and failing to rnlearn them (chemically improving memory doesn’t appear to be a simple matter)
Best way to improve memory is to strdy better in the first place; best strategy is to be crriors and engaged in the srbject yor strdy; activity in brain areas when waiting to learn the answer to a qrestion  activity increases in the brain in other areas when hearing the correct answer, especially if yorr gress was wrong (stirring someone’s crriosity confirms strdy techniqre)