Reader Comments
Post a new comment on this article
Post Your Discussion Comment
Please follow our guidelines for comments and review our competing interests policy. Comments that do not conform to our guidelines will be promptly removed and the user account disabled. The following must be avoided:
- Remarks that could be interpreted as allegations of misconduct
- Unsupported assertions or statements
- Inflammatory or insulting language
Thank You!
Thank you for taking the time to flag this posting; we review flagged postings on a regular basis.
closeuse specific guidance mechanisms or just random connectivity? hmm
Posted by sydneyunivisphys on 08 Jun 2007 at 23:22 GMT
Many LGN cells in cat get input from more than one retinal ganglion cell, and these inputs are almost always of the same type (X-, Y-, ON-, OFF-). Thus there are guidance mechanisms in the developing visual brain that allow the formation of precise functional connections. The statistical model here shows that these guidance mechanisms may not be required to form the precise maps of orientation in visual cortex from LGN inputs. But on the other hand, specific guidance is apparently quite possible so it seems plausible that there is more than just the relatively passive probabilistic process suggested in this paper.
RE: use specific guidance mechanisms or just random connectivity? hmm
dario replied to sydneyunivisphys on 11 Jun 2007 at 21:32 GMT
I may need some specific references to the studies you are referring to.
The article cites the simultaneous recordings and mappings of cells in the retina and LGN by Usrey et al, (1999), as showing that, from a functional point of view, geniculate cells tend to be strongly dominated by one input.
I believe that a combination of BOTH 'specific guidance' and 'statistical connectivity' mechanisms will need to be invoked to explain the full wiring of maps and receptive fields.
As a matter of fact, the model already relies on an early establishment of a well-defined retinotopy (presumably based on molecular gradients). I don't think one could argue that statistical connectivity ALONE can do the job.
On the other hand, one must be very careful to specify exactly what type of data would support the "formation of precise functional connections".
Consider a hypothetical example of one area A that containing cells of two types: Type 1 and 2. Cells in these area project to area B. Assume further that cells in Area B are somehow programmed to accept one and only one input from A. In other words, after the first synapse from A is formed, cells can refuse further attempts at synaptogenesis from other inputs.
The result would be that each cell in B is driven by a single cell in A, either type I or type II. Which one it is is, of course, is random. It is possible, in principle, that some molecular marker in area B attracts type I inputs to some cells and type II to others. However, this is not a conclusion one can arrive at based on one-to-one connectivity described above.
Similar concerns about "specificity of connections" arise as to the interpretation of thalamo-cortical connectivity data (you can read more about these arguments in my paper "Haphazard wiring [...]").