Neurobiology 303 -- Chapter 21 Outline

Neural Basis of Complex Behavior

decision-making in frogs and toads provides a well-studied example of a
   more complex behavior
    frogs and toads must be able to decide if an external stimulus
       represents a prey or a predator, and to make the appropriate
       behavioral response on the basis of that decision
    the appropriate response? approach the prey or avoid the predator!
    how does the frog or toad choose? this question was studied by a
       neuroethologist named Jörg-Peter Ewert
    Ewert showed that the prey/predator discrimination in toads is mainly
       dependent upon the size and movement of the stimulus
    showing a hungry toad a black square stimulus moving against a light
       background will evoke a response dependent upon stimulus size
                   small square -- approach (sometimes also strike)
                   large square -- avoidance (sometimes just crouch)
    apparently, the toad interprets the moving square stimulus as food if
       it's small and as an attacker if it's large
    if the stimulus is a rectangle rather than a square, then the direction of
       movement relative to the long axis is the determining factor
              move along long axis -- approach
              move along short axis -- avoidance
    apparently, the toad interprets the moving rectangle as a worm for
       movement along the long axis, and as an anti-worm for
       movement along the short axis

what is the neural basis of this prey/predator discrimination in toads?
    well, there are no specialized "prey-detectors" or "predator-detectors"
       in the retina of the toad eye!
    retinal ganglion cells in toads have circular center/surround receptive
      fields just like those that have been described in other animals
    so prey/predator discrimination must take place higher up in the brain,
       in regions that processes inputs from the retinal ganglion cells

in toads, retinal ganglion cells project to two places:
                optic tectum (a.k.a. tectum)
                posterior thalamus (a.k.a. pretectum)
    neurons in these regions response differently to visual stimuli
       depending upon the characteristics of those stimuli
    optic tectum -- neurons respond to stimuli that are small (square) or
       worm-like (rectangle moving along long axis), but give no
       response to stimuli that are large (square) or anti-worm-like
          (rectangle moving along short axis)
    pretectum -- neurons respond to stimuli that are large (square) or
       anti-worm-like (rectangle moving along short axis), but give no
       response to stimuli that are small (square) or worm-like
          (rectangle moving along long axis)
    thus, toad brain can distinguish between prey and predator on the
       basis of the differential responses in optic tectum and pretectum

in toads, approach and avoidance behaviors are organized in different
   regions of the brain
    electrical stimulation produces different responses depending upon the
       brain region to which it is applied
    optic tectum -- approach (sometimes also strike)
    pretectum -- avoidance
    given these results, one might predict that a lesion of pretectum would
       eliminate avoidance and would produce a toad that remains
       immobile when presented with large (threatening) stimuli
    is that the actual result of a pretectal lesion? no!!!
    a lesion of pretectum does eliminate avoidance, but it produces a toad
       that will approach and strike any stimulus, even a large one!
    neurons in optic tectum, which normally respond only to small (and
       worm-like) stimuli, will respond to large (and anti-worm-like)
       stimuli following a pretectal lesion

thus, the organization of approach/avoidance behavior in toads
   appears to be organized as follows
    pretectum -- pretectal neurons respond to large stimuli (and anti-
       worms) and send commands down to brainstem centers that
       produce avoidance; in addition, pretectal neurons inhibit
       neurons in the optic tectum
    optic tectum -- tectal neurons respond to all stimuli (small, large,
       worm, anti-worm) and send commands down to brainstem
       centers that produce approach; tectal neurons are inhibited by
       pretectal neurons, which suppress the response of tectal
       neurons to large stimuli (and anti-worms)
    in toads, the default behavioral response to any moving object appears
       to be "approach it and try to eat it"
    this default behavior is mediated by neurons in the optic tectum that
       respond to any moving stimulus and command approach
    pretectal neurons, which respond only to large moving stimuli, inhibit
       tectal neurons and thus suppress the default, approach response;
       at the same time the pretectal neurons command avoidance

behaviors even more abstract that simple decisions have been studied at the
   level of individual neurons and populations of neurons
    a good example concerns the ability of animals to become familiar
       with a place and to find their way around in it
    place cells -- neurons in the dorsal region of the hippocampus whose
       activity is correlated with the location of the animal in a place
    place cell activity is influenced by visual and vestibular input
       indicating that the animal's place is changing, and by the
       motivation of the animal to be in a certain place

simultaneous recording from many neurons in hippocampus provide further
   insight into how place is represented by the brain
    analysis of these recordings shows that the activity of any individual
       place cell is only loosely correlated with the place of the
      animal, but that the collective activity of the population is
       strongly correlated with the animal's place in a location
    thus, the hippocampus uses a population code to represent place
    also, the hippocampus is able to represent several different locations
      and place cells for new locations can develop quickly

other regions besides the hippocampus also contain neurons whose activity
   is correlated with place including:
        posterior parietal cortex -- encode direction of movement
        striatum -- encode place and change of place
        parietal cortex and striatum probably work with hippocampus in
             producing place specific behaviors

much of neurophysiological research involves invasive techniques where the
   nervous system is exposed and probed directly
    several noninvasive techniques have been developed since the 1980's
       that allow noninvasive imaging of the brain and its activity
    these techniques include the following
       computed axial tomography (CAT scan) -- images constructed
          on basis of multiple-angle X-ray pictures
       positron emission tomography (PET scan) -- images
          constructed on basis of emission of gamma rays; the
          gamma rays result from collision with electrons in the
          brain of positrons emitted from radioactive substances
          that are injected into the bloodstream; these radioactive
          substances can travel in the blood and be taken up by
          active neurons, thus indicating active regions of the brain
       magnetic resonance imaging (MRI) -- images constructed on
          basis of signals emitted from the brain as atomic nuclei in
          the brain relax following release from an externally
          applied radio frequency signal; MRI can image the
          structure of the brain and can also indicate active regions
          (functional MRI or fMRI)

one fact concerning the organization of the brain that has been confirmed by
   modern imaging studies is that every region of the brain has a
   specific functional role
    also, parts of the brain that are physically distinct from one another
       often work together in carrying out particular mental tasks
   segregation of function in cerebral cortex is well illustrated by imaging
       studies of language comprehension and production
    these studies are done using, for example, PET scans
    first a scan is taken of an individual at rest
    another scan is taken as the individual performs some language-
       related task
    the two scans are then compared to determine which areas of cortex
       were active during performance of the task
    these studies show that different regions of cortex are activated during
       performance of different aspects of language processing
    specifically:
       reading words -- visual cortex and visual association area become
          active; visual association area aids in interpretation of images
       listening to words -- angular gyrus in posterior temporal lobe and
          Wernicke's area at the junction of the temporal and parietal
          lobes becomes active; the angular gyrus processes auditory
          input and Wernicke's area helps to decipher spoken words
       say a word -- the mouth and tongue parts of primary motor cortex and
          Broca's area in the prefrontal cortex become active; Broca's area
          is important for generating grammatical speech and the motor
          cortex is needed to control movements necessary for speaking
       mental manipulation -- thinking about words associated with other
          words (like a verb that would be associated with a particular
          noun) activates a region of the inferior frontal cortex

many of the areas needed for processing the same aspects of language
   are located in different places
    obviously, these different brain regions interact through their synaptic
       connections

attentional mechanisms modulate the degree of activation of the
    regions that mediate various aspects of language processing
    for example, a subset of brain regions becomes active if a subject is
       shown two words and asked to choose one on the basis of the
       meaning of the words; an overlapping subset will become
       active if the choice between the same two words involves not
       meaning but some purely sensory attribute like letter style; all
       regions in both subsets become active if the choice between the
       words involves both meaning and sensory attributes

a more difficult experiment involves choosing between two words that differ
   both in meaning and in sensory attributes, but the subject is
   asked to concentrate on one aspect more than the other
    again, both subsets of brain regions become active, but the PET scan
       cannot distinguish any difference between them
    however, the PET scan is slow, although it is well localized
    electrical recordings from the scalp are fast, but poorly localized
    PET and electrical recordings from the scalp can be combined to yield
       fast and well-localized detection of brain activity
    the combined technique reveals that the regions associated with the
       attribute toward which attention is directed become active first

the subset of active brain regions can also depend upon the familiarity of a
   subject with a given task
    for instance, if a subject is asked to say a verb related to a presented
       noun, a subset of brain regions become active
    as the subject practices the task for that particular verb and the task
       becomes routine, a different but overlapping subset of brain
       regions becomes active
    when the subject switches to a new noun the original subset of brain
       regions again becomes active

focal brain lesions can lead to impairments of very specific functions
    not only is brain function distributed over different localized regions, but it
       is lateralized as well

the capabilities of the right and left hemispheres are different
    for example, in most people language capability is centered in the left
       hemisphere; language will be impaired if Broca's area is
       damaged in the left hemisphere, damage in the right hemisphere
       will have little effect on language

the different capabilities of the right and left hemispheres was first
   demonstrated by Roger Sperry and Michael Gazzaniga in
   so-called split-brain patients
       split-brain patient -- a person in whom the corpus callosum has been
          surgically severed

corpus callosum -- broad band of axons that interconnects neurons in
   the two hemispheres of the cerebral cortex
    surgically severing the corpus callosum is done to relieve severe cases
       of epilepsy -- it works by eliminating the positive feedback
       between the two hemispheres that can cause the buildup of
       epileptic activity

split-brain surgery leaves the patient normal in many respects, but careful
   analysis revealed that the separated hemispheres operate
   independently of one another
    this analysis is possible because of three features of neuroanatomy
    motor commands from the left hemisphere control muscles on
       the right side of the body and vice-versa -- the projection
       fibers cross over in the medullary pyramids
    sensory commands on one side cross over in the brainstem
       before ascending to the cortex on the opposite side
    images from the left side of the visual world are conveyed
       to the right side of the brain and vice-versa -- the retinal
       ganglion cell projections cross over in the optic chiasm
    all three of those crossover points (pyramids, brainstem, optic chiasm)
       occur well beneath the corpus callosum, so cutting the latter
       does not disrupt the former
    Sperry and Gazzaniga took advantage of these features to study the
       capabilities of the two hemispheres

the split-brain experiments involve presenting stimuli to one hemisphere or
   the other and noting differences in response capability
    for example, an image of a banana is flashed either onto the left or the
       right side of a screen viewed by a split-brain patient
    if the banana is shown on the right, the image is conveyed to the left
       hemisphere, and the patient can describe the banana in words
       because the language center is also in the left hemisphere
    if the banana is shown on the left, the patient cannot say what has
       been seen because the right hemisphere, to which the left side
       visual image is conveyed, has no language capability

the comprehension capability of the right hemisphere was demonstrated by
   using the sense of touch
    here the split-brain patient can use either hand to touch objects that are
       inside a box out of sight
    then images of the same objects are flashed onto either the left or right
      sides of the screen
    if an image is flashed onto the left side, the patient is able to use his
       (her) left hand to pick out the corresponding object
    this shows that the right hemisphere also has recognition capabilities
       and can understand spoken instructions, because it is the right
       hemisphere to which left side images are conveyed and from
       which left side motor commands originate

further split-brain experiments have shown that the right hemisphere
   has many of the same capabilities as the left, the most obvious
   exception being language
    however, the abilities of the two hemispheres are not identical for
       each type of task
    in general, the capabilities of the two hemispheres can be broken
       down as follows:
          right hemisphere -- specialized for parallel processing and
             pattern recognition that requires many bits of information
             to be integrated (like face recognition)
          left hemisphere -- specialized for serial processing and logical
             analysis (like language and deductive reasoning)
    despite these functional difference, it is clear that both hemispheres
       work together in normal, intact humans
    in fact, there are some tasks that split-brain patients can't perform
    in one such task, subjects manipulate an unseen, wire-frame object
       (like a cube or a pyramid shape) in a box with one hand or the
       other, and must then recognize the object by sight
    normal, intact subjects can do this easily, but split-brain subjects can't
       do it, no matter which hand they use to manipulate the object

much can be learned about the function of the brain by studying patients
   with various brain diseases
    many brain diseases involve transmitter systems
    failure of neurotransmitter systems can produce symptoms that are
       neurologically widespread and devastating to the individual

different diseases are associated with failure of specific
   neurotransmitter systems
    Parkinson's disease, a disorder of movement, is associated with
       degeneration of dopaminergic neurons in the substantia nigra
    Alzheimer's disease, characterized by severe memory loss, is
       associated with deficits in cholinergic transmitter systems in the
       cortex due mainly to degeneration of cholinergic neurons in the
       basal forebrain
    manic depression (bipolar disorder), a disorder of mood, is associated
       with striking deficiencies in serotonergic and noradrenergic
       transmission throughout the brain
    schizophrenia, a disorder of thought processes, is associated with an
       abnormally high level of activity in the dopaminergic
       transmitter system
    the causes of these diseases are still poorly understood, but many have
       genetic components