Neurobiology 303 -- Chapter 24 Outline

Behavioral Plasticity: Learning

among all of its functions, the brain is most notable for its ability to learn
   and to store its learning as memory
    in a way, the function of the nervous system is to learn and remember
       things that could not be pre-programmed by the genome
    learning -- a change in the behavior of an animal based on experience
    memory -- the process of storage and retrieval of learned experience

habituation -- the simplest form of learning, is the cessation of a response to
   a stimulus after repeated presentation of that stimulus
    even reflexes, which are defined as stereotyped responses to specific
       stimuli, can show habituation
    you experience sensory habituation whenever you no longer notice a
       constant stimulus that initially commanded your attention, as
       when you get used to wearing itchy socks, for example

habituation, and other simple forms of learning, have been studied
   extensively in the invertebrate sea-slug Aplysia
    like other mollusks, Aplysia breaths through a gill and a siphon
    these parts are soft and exposed, so the Aplysia has a reflex, called the
       gill-withdrawal reflex, whereby it can retract its gill and siphon
       into a mantle at the first sign of mechanical insult
    a gentle touch of the siphon or mantle causes a slight withdrawal of
       the siphon and gill
    repetition of the gentle touch stimulus will result in a gradual
       cessation of the gill-withdrawal response -- it habituates
    the habituated state will reverse itself with time after the
       repetitive stimulation is discontinued -- dishabituation
    the longer the repetitive stimulus is applied the longer habituation will
       endure after stimulation is over
    work by Eric Kandel and coworkers has uncovered the
       neurophysiological basis of habituation of the gill-withdrawal
       reflex in Aplysia

neurophysiological basis of habituation of the gill-withdrawal reflex
   in Aplysia
    the Aplysia gill-withdrawal reflex pathway is really a disynaptic
       pathway in parallel with a monosynaptic pathway
    specifically, the Aplysia gill-withdrawal reflex consists of sensory
       neurons, interneurons, and motor neurons, and two pathways:
                     disynaptic -- sensory to interneurons to motor neurons
                     monosynaptic -- sensory to motor neurons directly
    of the two pathways the monosynaptic pathway contributes more
    the main underlying cause of habituation of the Aplysia gill-
       withdrawal reflex is depression of the synapses between the
       sensory and motor neurons
    synaptic depression -- reduction in effectiveness of pre-synaptic action
       potential from eliciting a post-synaptic action potential
    synaptic depression is caused by a reduction in the amount of
       neurotransmitter released from the sensory neuron
    reduction in transmitter release is attributed to two changes:
          inactivation of calcium channels, which reduces calcium influx
          reduction in number of neurotransmitter vesicles
    synaptic depression with repeated stimulation is observed at many, but
       by no means all, synapses in invertebrates and vertebrates

neurophysiological basis of dishabituation and sensitization of the gill-
   withdrawal reflex in Aplysia
    dishabituation -- full recovery of a habituated response by
       presentation of a strong, novel stimulus
    sensitization -- increase in the strength of a normal (i.e. not
       habituated) response by presentation of a stimulus
    dishabituation and sensitization can occur with stimuli of modalities
       other than that which usually evokes the response
    following habituation of the Aplysia gill-withdrawal reflex by
       repeated touch of the siphon, the reflex can be dishabituated by
       a strong tap on the tail
    a strong tap on the tail can sensitize the gill-withdrawal reflex in a
       normal (i.e. not habituated) Aplysia
    the underlying mechanism for both dishabituation and sensitization is
       called heterosynaptic facilitation
    heterosynaptic facilitation -- increase in effectiveness of a synapse
       between pre- and post-synaptic neurons that is brought about by
       input to that synapse from other neurons

mechanism of heterosynaptic facilitation of the Aplysia gill-withdrawal
   reflex
    synapses onto motor neurons (from sensory neurons and interneurons)
       receive pre-synaptic inputs from facilitating interneurons,
       which in turn receive sensory input from the tail
    when the facilitating interneurons are activated by sufficiently strong
       sensory input, they secrete serotonin pre-synaptically onto
       motor neuron synapses
    serotonin has two main effects on the synapse:
       it increases the amount of calcium that enters the synapse upon
          depolarization and it does this via the following second
          messenger cascade: serotonin binds membrane receptor,
          which activates a G protein, which activates adenylyl
          cyclase, which converts ATP to cAMP, which activates
          cAMP-dependent PKA, which phosphorylates a
          potassium channel thus blocking the channel; with the
          potassium channel blocked, the synapse stays depolarized
          longer, so more calcium gets in and more transmitter is
          released
       serotonin also makes more transmitter available for release at
          the synapse and it does this via the following second
          messenger system; serotonin binds membrane receptor,
          which activates another G protein, which activates
          membrane-bound diacylglycerol, which activates PKA
          and PKC, which mobilize transmitter vesicles from a
          storage pool to a release pool, so more transmitter is
          available for release
    with more calcium entering the synapse to release transmitter, and
       with more transmitter available for release, the synapse
       becomes more effective
    these mechanisms of heterosynaptic facilitation are short term
    persistent activation of the facilitating interneurons brings about more
       long term facilitation, and the mechanism is as follows:
          the activated PKA is translocated to the nucleus where it
          initiates synthesis of two proteins; one of these increases the
          availability of PKA, which blocks more of the potassium
          channels; the other protein promotes the formation of new
          synaptic connections

more complex learning -- associative conditioning
    first demonstrated by Ivan Pavlov and his famous dog (Pavlov's dog)
    in associative conditioning, an association is formed between two
       types of stimuli:
    unconditioned stimulus (US) -- the stimulus that
       normally evokes a given behavioral response
    conditioned stimulus (CS) -- another stimulus that
      normally does not evoke that behavioral response
    Pavlov paired the presentation of meat (unconditioned stimulus) with
       a ringing bell (conditioned stimulus) to his dog; the dog
       normally salivated when he saw the meat; after repeated
       pairings of meat and bell, the dog salivated when Pavlov rang
       the bell, even if he proffered no meat -- Pavlov had conditioned
       the dog to associate the bell with the meat

associative conditioning of the Aplysia gill-withdrawal reflex
    weak touch of the siphon or mantel will not evoke much gill-
       withdrawal, but strong electrical stimulation of the tail always
       evokes a full gill-withdrawal reflex response
    Kandel paired strong tail shock (unconditioned stimulus) with weak
       siphon touch (conditioned stimulus) and conditioned the
       Aplysia to fully withdraw its gill in response to weak siphon
       touch, which normally produced very little gill withdrawal
    after conditioning weak siphon touch to produce a full response,
       Kandel showed that weak mantel touch still produced very little
       gill withdrawal -- this showed that the conditioning was specific
       to siphon touch

mechanism of associative conditioning of the Aplysia gill-withdrawal reflex
    the conditioning produces heterosynaptic facilitation of the synapses
       from the sensory neurons onto the motor neurons, but only the
       synapses that are co-activated with the tail shock are facilitated
    tail shock activates the facilitating interneurons, which produce short
       term heterosynaptic facilitation as previously described, via
       serotonin
    weak activation of the siphon weakly activates the siphon sensory
       neurons, not enough to produce much gill withdrawal, but
       enough to depolarize the synapse a little and allow calcium to
       enter, which then activates calmodulin; calcium/calmodulin
       activates adenylyl cyclase, which converts ATP to cAMP
    now there are two separate second messenger cascades that have
       activated adenylyl cyclase -- calcium/calmodulin (from direct
       depolarization of the synapse) and serotonin (from the
       facilitating interneurons)
    the two cascades work synergistically to produce greater activation of
       adenylyl cyclase than the sum of the two cascades working
       alone, and the result is much more cAMP production
    the greatly elevated levels of cAMP then work to enhance calcium
       entry and transmitter availability, as cAMP did for
      heterosynaptic facilitation but to a greater extent
    for associative conditioning to occur, the CS must precede the US but
       only by a short interval -- the precise timing is what allows the
       facilitation of conditioning to be specific to particular synapses

genetic mutations in fruit flies can prevent associative conditioning, either
   by causing too much or too little cAMP
    duc (dunce) -- mutation of cAMP phosphodiesterase, which normally
       breaks down cAMP; duc causes over accumulation of cAMP
       that prevents associative conditioning because cAMP cannot be
       further elevated by pairing CS and US
    rut (rutabaga) -- mutation of adenylyl cyclase itself; rut prevents all
       cAMP production, so conditioning cannot occur

associative conditioning in vertebrates may depend upon a phenomenon
   known as long-term potentiation (LTP)
    LTP is a long-term (many days) increase in the amplitude of a post-
       synaptic response brought about by co-activation of the pre-
       and post-synaptic neurons
    LTP can occur when the co-activation of the post-synaptic neuron is
       brought about either by repetitive activation of the pre-synaptic
       neuron, or by a third neuron that separately activates the post-
       synaptic neuron
    LTP is associated with activation of NMDA glutamate receptors
    depolarization of the post-synaptic neuron causes magnesium to be
       displaced from the mouth of the NMDA channel -- this allows
       more calcium to enter the post-synaptic cell, which then
       activates various calcium-dependent kinases, which then make
       non-NMDA receptors more responsive to glutamate

the study of the synaptic basis of learning in vertebrates has been greatly
   facilitated by the brain-slice technique
    to make brain slices, a brain is removed from a live animal (usually
       following a quick decapitation by guillotine) and sliced into
       slabs about 0.5 mm thick
    the slab is put in a chamber containing glucose, minerals, and oxygen
       necessary to keep the neurons in the slab alive for many hours
    the brain slice can be viewed under a microscope and intracellular
       recordings can be made from parts of selected cell types -- for
       example, recordings can be made simultaneously from a post-
       synaptic cell and a pre-synaptic terminal
    LTP in hippocampus has been studied in this way

learning, memory, and the hippocampus
    the hippocampus is necessary for several types of learning and
       memory in rats and other mammals
    in humans, damage to the hippocampus causes deficits in learning
       about people, places, and things
    in rats, lesions to hippocampus result in deficits in spatial learning
       (i.e. locating a particular place using spatial cues)
    LTP may underlie learning in the hippocampus

knockout mice
    mice in which the function of a single gene is eliminated,
       or knocked out
    the procedure for making a knockout mouse essentially involves the
       following steps
          create a mutant copy of the gene under study
          flank the mutated gene with marker genes, and flank the marker
             genes with the genes that flank the normal gene
           incubate this strand with mouse embryo stem (ES) cells
          ES cells will take up the mutant gene and some will swap it for
                the normal gene by homologous replacement
          grow ES cells in culture and screen them cells for the marker to
             find ES cells were the swap has been made successfully
          inject ES cells carrying the mutant gene into normal mouse
             embryos at the blastocyst stage
          this produces a chimeric blastocyst containing some normal and
             some mutant cells
          place the chimeric blastocyst into pseudopregnant mice and
             allow them to develop into chimeric offspring
          as adults, the chimeric offspring will have some cells, including
             germ cells (eggs and sperm) that carry the mutant gene
          mutant skin cells will identify chimeric mice by their marker
             genes that might code, for example, black fur
          mate chimeric mice with normal mice to produce some
             offspring that are heterozygous for the mutant gene
          mate the heterozygous offspring to produce some mice that are
             homozygous for the mutant
          a mutant strain of mice has now been developed that is normal
             except for the complete elimination of a single gene
    this basic knockout technique has been useful in studying how various
       genes contribute to learning and memory

learning in the hippocampus may involve NMDA receptors and LTP
   as mentioned above, LTP involves passage of calcium through
    activated NMDA channels, which then activates various protein
       kinases, which then make the synapse more effective
    one of these kinases is called alpha-calcium-calmodulin-dependent
       kinase II (alpha CaMK II)
    Susumu Tonegawa made knockout mice with the alpha CaMK II gene
       knocked out -- these mice could not synthesize alpha CaMK II
    recordings in hippocampal brain slices from these knockout mice
       showed that they had no LTP, yet the hippocampal neurons
       were normal in all other respects
    the alpha CaMK II knockout mice behaved normally, except that they
       had great difficulty in performing spatial learning tasks
    in another strain of mice, Tonegawa knocked out a gene that codes a
       specific type of NMDA receptor only in the CA1 region of the
       hippocampus -- NMDA receptors in other parts of the
       hippocampus, and brain generally, were not affected
    these CA1 NMDA receptor knockouts lacked LTP in the CA1 region
       and also had great difficulty in performing spatial learning tasks
    these experiments suggest that alpha CaMK II is necessary for LTP
       and that LTP is necessary for spatial learning in hippocampus

memory is an essential concomitant of learning
    there are several, perhaps many, different types of memory
    declarative (explicit) memory -- can be broken down into:
       episodic memory -- memory of experiences
       semantic memory -- memory of names and definitions
    non-declarative memory -- can be broken down into:
       procedural memory -- skill acquisition, and simpler forms like
          conditioning and habituation
       priming -- recognition of having seen an item before, like a
          word, without remembering its meaning, context, etc.
       short-term (working) memory -- memory retained for a relatively
          short time (up to several hours)
       long-term memory -- memory that can be retained indefinitely
    all animals have short-term and long-term forms of non-declarative
       memory, which occur throughout the nervous system
    only vertebrates have declarative memory, which is stored either in or
       by the hippocampus, and can also be short- or long-term

the formation of long-term memory requires protein synthesis
    if protein synthesis is blocked right after some experience, the
       experience will not be learned on a long-term basis
    the mechanism of new protein synthesis that underlies long-term
       memory involves the following:
          an appropriate stimulus induces activation of cAMP,
             which then activates PKA
          a subunit of PKA breaks off and is translocated to the nucleus
          in the nucleus, the PKA subunit phosphorylates a regulatory
             protein known as cAMP response element binding
             protein (CREB)
          CREB is a transcription factor, a molecule that promotes
             synthesis of RNA from a strand of DNA
          CREB binds with a segment of DNA called the cAMP response
             element (CRE)
          binding of CREB to the CRE promotes transcription of the gene
             and hence synthesis of new protein
    the new proteins that are synthesized enhance synaptic transmission
       by directly or indirectly influencing the amount of transmitter
       or the structure of the synapse
    up or down regulating CREB in transgenic animals (e.g. fruit files)
       induces or suppresses long-term memory
    in Aplysia, injection of CREB enhancers will induce long-term
       memory even if only short-term training is applied, and
       injection of CREB inhibitors will suppress long-term memory
       even if long-term training is applied
    an analog of CREB has been identified in mammals

memories are stored in various regions throughout the nervous system
    in early studies, Karl Lashley made small and large lesions throughout
       the brain and found that memory loss was proportional to the
       amount of brain tissue lesioned, regardless of where the lesions
       were made
    Lashley concluded that memories are distributed throughout the brain
       rather than localized in one place
    now we know that the truth is somewhere in-between -- lots of brain
       regions can store memories, but specific memories are stored in
       specific brain regions
    for example in flies, memory of olfactory avoidance learning is stored
      in the mushroom bodies, a particular region of the brain

eye-blink conditioning in rabbits provides an excellent example of the site
   specificity of learning in vertebrates
    this work was pioneered by Richard Thompson
    for eye-blink conditioning, the US is an air puff applied to the cornea
       of the eye, which naturally produces a blink
    the CS is a brief tone, which is conditioned to produce an eye-blink by
       repeatedly pairing it with the air puff (US)
    the reflex eye-blink (in response to the US) is mediated by a three-
       neuron reflex pathway, from corneal sensory neurons to
       interneurons in the trigeminal nucleus to motoneurons in the
       facial nucleus that control the eyelid muscles
    information about the US and the CS is also transmitted via various
       brainstem nuclei to the cerebellum, both to Purkinje cells in the
       cerebellar cortex and to neurons in the interpositus nucleus (one
       of the deep nuclei) -- the US arrives via climbing fibers and the
       CS via mossy fibers
    an association between the CS and US can then be made synaptically,
       both at the Purkinje cells and at neurons in the interpositus
    thus, the pathway for the conditioned response goes from brainstem
       nuclei to the interpositus, and from there through the red
       nucleus to eyelid motoneurons in the facial nucleus
    various pieces of evidence have been offered in support of the theory
       that the cerebellar cortex is necessary for eye-blink conditioning
       to occur, and that the site of eye-blink conditioning is the
       interpositus nucleus
    after training, electrical stimulation of different parts of the pathway
       can elicit the eye-blink response
    also, conditioning of CS to US can occur with properly timed
       electrical stimulation of mossy and climbing fibers, in the
       absence of actual tones and air-puffs
    the eye-blink cannot be conditioned to a tone if the interpositus
       nucleus is lesioned
    mice with mutated Purkinje cells cannot have their eye-blink response
       conditioned to a tone