< Previous | Home | Up | Table of Contents | Next >

Soft Uploading:

Neuroscience Basis of Uploading

Below are some neuroscience facts and theories, some well-established and some speculative, that are relevant to uploading and brain emulation.


Housekeeping vs information processing

What is the criteria that a chemical signal is relevant to consciousness? A person may not be consciously aware of a chemical process and yet it could still be relevant (eg subconsciousness). A "housekeeping" process is one that does not contribute to mechanisms of information processing in the brain.


Neural information consists of 3 types:

  1. Electrical -- action potentials; dendritic integration
  2. Chemical -- neuromodulation and gene expression etc
  3. Morphological -- neuronal connectivity

Biological systems are often hierarchical

The brain has many levels of organization [Synaptic Organization of the Brain, ed Gordon M Shepherd]:

  1. systems
  2. topographical maps
  3. laminas -- 6-layer isocortex
  4. vertical columns
  5. neurons
  6. axonal and dendritic trees
  7. synaptic microcircuits -- small clusters of synapses
  8. synapses and dendritic spines
  9. molecules and ions

The action potential filters out noise

The all-or-nothing principle of action potentials was first proposed by Edgar Douglas Adrian in 1914 at the age of 25. The threshold for generating an action potential is a mechanism (an example of positive-feedback) to filter out small random variations in the membrane potential. Using phase-space analysis of PDEs (the Hodgkin-Huxley equations), it can be shown that some traveling-wave solutions (corresponding to action potentials) are stable in the sense that "deviations from it relax to zero with increasing time"; as opposed to sub-threshold potentials that are unstable in the sense that "deviations from it diverges with increasing time: slightly smaller solutions decay to zero, and slightly larger solutions grow to become the fully developed nerve impulse" [Neuroscience: A Mathematical Primer, Alwyn Scott 2002].

This fact is of pivotal importance because we can deduce from it that neural computation is noisy, which lets us determine what kind of information is relevant to consciousness.

Since the shape of action potentials is stereotypical, it cannot convey information. In the brain, however, the shape of action potentials can be affected by neuromodulation.


Neuropeptides

Currently we do not fully understand the purpose and mechanisms of peptide neuromodulation. Some clues are as follows:

Evolution of signaling pathways

A probable evolutionary sequence of events [Leo W Buss 1987 The Evolution of Individuality]:

  1. Chemical signalling among unicellular organisms (protists)
  2. Emergence of multicellualrity (cell colonies eg volvox (phytomastigophora))
  3. Emergence of cellular differentiation (plants, fungi, animals)

Primitive Animals

According to current theory, choanoflagellates (a protist) gave rise to sponges (Porifera) and these in turn to the diploblastic and triploblastic animals (metazoans). Sponges do not have nervous systems. The most primitive nerve nets are found in Cnidaria that are believed to be close to the base of the metazoan tree, as shown [Tree of Life Web Project 2001].

"The first multicellular organisms apparently had no nervous system (like volvox). Thus, it was a problem how to coordinate the movements of the entire organism. Hydra and many other Cnidaria have no central nervous system, but an almost uniform web of nervous cells. These cells have specialized in signal transmission. Yet, all nervous cells are uniform.The maximum they can do is to cause the contraction of the entire body. Here there is no specialization among nervous cells, no higher-level control." [© 1998 Principia Cybernetica "Biological examples of MSTs"]

Hydra contains several hundreds (~500 known) of peptide signalling molecules, some of which are neuropeptides and others emanate from epithelial cells [Bosch & Fujisawa 2001]. But the view that Hydrozoa was the ancestral group to Cnidarians has recently been refuted.

Ctenophores are likely to be the simplest metazoans possessing definitive muscle cells.

Why are there peptides in primitive nerve nets?

To answer this, one should look at what kind of behavioral responses are possible in organisms basal to the metazoan tree, eg sponges, Ctenophora, or Cnidaria.

1) What can choanoflagellates (protists) do? Do they communicate with each other?

2) What can sponges do? What kind of intercellular communication?
Most are sessile, some can move slowly (up to 4 mm per day). One family of sponges can feed by capturing and engulfing small crustaceans.
1) reproduce a/sexually; 2) growth; 3) feeding; 4) perception of environment (eg calcareous material)

3) What can ctenophores / cnidarians do?
1) - 4); 5) swim; 6) escape / defense; 7) chemotaxis?

Neuromodulation and fast synaptic transmission: which evolved first?

All the classical transmitter ligand molecules evolved at least 1000 million years ago. With the possible exception of the Porifera and coelenterates (Cnidaria), they occur in all the remaining phyla [Walker, Brooks, Holden-Dye 1996].

Reference

[Bosch & Fujisawa] Polyps, peptides and patterning, Bioessays. 2001 May;23(5):420-7
[Walker, Brooks, Holden-Dye] Evolution and overview of classical transmitter molecules and their receptors, Parasitology. 1996;113 Suppl:S3-33


< Previous | Home | Up | Table of Contents | Next >

Feb/2004 Yan King Yin