RE: Simple neural models

From: Gatherer, D. (Derek) (D.Gatherer@organon.nhe.akzonobel.nl)
Date: Wed Jul 26 2000 - 08:26:53 BST

  • Next message: Gatherer, D. (Derek): "new Gary Cziko book"

    Received: by alpheratz.cpm.aca.mmu.ac.uk id IAA06768 (8.6.9/5.3[ref pg@gmsl.co.uk] for cpm.aca.mmu.ac.uk from fmb-majordomo@mmu.ac.uk); Wed, 26 Jul 2000 08:30:52 +0100
    Message-ID: <A4400389479FD3118C9400508B0FF230040E6A@DELTA>
    From: "Gatherer, D. (Derek)" <D.Gatherer@organon.nhe.akzonobel.nl>
    To: "'memetics@mmu.ac.uk'" <memetics@mmu.ac.uk>
    Subject: RE: Simple neural models
    Date: Wed, 26 Jul 2000 09:26:53 +0200
    X-Mailer: Internet Mail Service (5.5.2650.21)
    Content-Type: text/plain; charset="iso-8859-1"
    Sender: fmb-majordomo@mmu.ac.uk
    Precedence: bulk
    Reply-To: memetics@mmu.ac.uk
    

    Mark:
    I'm not suggesting an entire nerve cell takes on a single binary state,
    only that binary elements at the synapse level play key roles in
    electro-chemical signal processing. Specifically, I'm alluding to
    Kock's description of autophosphorylating kinases (Biophysics of
    Computation). He suggested they are analogous to transistors. With a
    certain amount of voltage applied to them, they conduct. Without the
    voltage, they resist.

    Derek:
    I don't understand what you mean by 'With a certain amount of voltage
    applied to them, they conduct'. You must mean transistors? Surely not the
    kinases. Kinases add phosphate groups to proteins - if they are
    autophosphorylating, they add phosphate groups to themselves. For instance
    protein kinase A has 2 phosphorylation sites at positions Thr-197 and
    Ser-338. But it is activated by binding of cAMP to the C-terminus. I
    having trouble seeing the transistor analogy. Where and when is the voltage
    applied to the kinase?

    Mark:
    As to memory, if an autophosphorylating kinases is conducting, then its
    previous state was non-conducting.

    Derek:
    Again, I can't see what basis this transistor analogy has in actual biology.
    'Conducting'? Kinases don't conduct. Kinases are very subtle enzymes which
    can exhibit a whole spectrum of specificity and activity. They _aren't_
    some kind of molecular flip-switch. If you want an analogy/metaphor to
    electronics, then you'll need an analogue rather than a digital one. All
    enzymes have Michaelis-Menten kinetics, they don't just flick on and off, so
    when you say .....

    "They can only be one or the other."

    That really doesn't wash, at least not as far as a neurobiologist would be
    concerned.

    Mark:
    One
    might say that the cells doesn't know 'when' the previous state existed (1
    msec ago? 10 seconds ago?), but there is still a piece of inferential data
    available about the past which a signal processing system could use. Add a
    stable electrical wave pattern to the system and much more can be made of
    the inference.

    Derek:
    I have 2 problems with the above paragraph. Firstly, I still don't see what
    your mechanism is for the cell 'knowing' its past states, and secondly I'm
    not sure what 'Add a stable electrical wave pattern to the system' means.
    It seems you want the cell to be rather like an electronic circuit. There
    are too many soldering iron metaphors in here for a biologist to make sense
    of the proposed mechanism.

    Mark:

    Since listening to neural signals crossing synapse membranes is very
    difficult, most work on the role phosphorylating kinases comes from
    observations of physiological change during embryonic neural
    developments. They play key roles in neural differentiation, cell
    migration and connectivity. Knock out a kinases and the brain doesn't
    develop properly. For example, knocking out the gene for mDab1 (tyrosine
    phosphorylated during embryonic development) in mice does nothing to change
    the mice for the first week after birth. After that, they exhibit
    increasing motor deficits and grossly malformed brains. Apparently, the
    neural cells differentiate, but fail to migrate.

    Derek:
    Fair enough, but all this shows is that kinases have crucial roles to play
    in the cell, and specifically in neurogenesis, and that there is a genetic
    basis to both differentiation and migration.

    Mark:
    Edelman in 'Neural Darwinism' suggests neural systems develop according to
    'neuronal group' selection. This provides a level of complexity left
    'undetermined' by the genetic foundation. By binding Edelman's ideas to
    Kock's, we get neuronal groups with binary signal processing at their
    foundations (neural memetics).

    Derek:
    I still don't see how you make this leap. I grant you that neuronal
    plasticity demonstrates that there is a lot to the brain which is not
    genetic, but to go as far as to say .....

    "The neural system is a self-determining organ, with its own developmental
    memory founded upon its own mechanisms."

    seems to be a tep too far. Your own example of the mDab1 knockout above
    illustrates this. You say:

    "Going back to the mouse example above, the observed lack of neural cell
    migration shows cellular differentiation occurred (genetics) but not the
    normal migration (neural memetics)."

    The fact that the gene knockout inhibits the migration shows that the normal
    migration _is_ genetically controlled. Even for those aspects of
    neurogeneisis that are not genetically 'hard-wired', there is no evidence
    that they can be influenced by cultural processes, and therefore no
    justification for labelling them 'memetic'. Just because something isn't
    rigidly genetically determined doesn't mean that it is therefore memetic.
    Memetics, after all, whether one is an internalist or an externalist, is
    about culturally transmitted somethings (mnemons/behaviours/information
    delete as applicable). It isn't a bag into which everything non-genetic
    falls.

    Mark:
    While you might object that migration of neural cells to proper locations
    in the brain has little to do with cultural artifacts like Windsor knots,
    the ability to create Windsor knots reflect physiological changes in the
    brain of 'enabled' individuals, too. As an individual goes from 'blank' to
    'Windsor knot' enabled psychological states, their brain changes, neural
    cells extend themselves, connections are made. We are discussing a matter
    of degree, not of kind.

    Derek:
    Of course, I don't disagree that learning must have some sort of effect on
    the brain. What I'm saying is:

    a) your attempt to explain it in terms of an electronics-type analogy
    requires you to postulate things about kinases and/or other cellular
    components which won't stand up to scrutiny by molecular biologists.

    b) you still don't have a convincing mechanism for your hypothesised
    'memory' effect

    c) you are not justified in attaching the label 'memetic' to a process just
    because it is not genetic. Besides, your gene knockout example above
    actually argues the case for _greater_ genetic determination, not less.

    Derek

    ===============================================================
    This was distributed via the memetics list associated with the
    Journal of Memetics - Evolutionary Models of Information Transmission
    For information about the journal and the list (e.g. unsubscribing)
    see: http://www.cpm.mmu.ac.uk/jom-emit



    This archive was generated by hypermail 2b29 : Wed Jul 26 2000 - 08:31:53 BST