Federico C 27 may 2008 (EDT)
As far as I have understood the project, we are trying to artificially reproduce Turing patterns; a kind of patterns that could be generated from a homogeneous media, that have been studied extensively, reproduced chemically and that are thought to have an important role in some process of morphogenesis.
Apparently some naturally occurring patterns seem to match and resemble Turing patterns, such as those seen in the organization of tricomas in Arabidopsis thaliana’s leafs or the jaguar spots, yet it’s still controversial whether or not they are actually that kind of patterns for the genetic network that underlies its formation is unknown.
While we could analyze the variation in naturally occurring patterns and dwell into the genetics of an organism (basically, hammering the organism by knocking out genes), a different approach to the problem; the construction of synthetic networks that produce patterns in organisms that previously didn’t have them, might to be an insightful and refreshing alternative.
Our synthetic devise may not resemble the natural one and might pale in comparison, yet its successful construction would allow us to establish whether or not the basic elements necessary for the process are complete and well understood and even reveal what would be needed for cells to generate complex patterns.
Federico C 27 may 2008 (EDT)
Which construction are we going to use? I’ve been unable to design a devise to replicate the widest known Turing pattern, the one that is presented in most papers, it involves a node that is repressed by the action of one diffusible signal and induced by the action of another node and there doesn’t seem to be promoters that behave in such a way, actually that leads me to think that most Turing patterns are not formed in the way that most papers suggest… simply it’s not biologically feasible.
Anyway, unless someone else comes with a silver bullet I’ll suggest affronting the problem in a different way: To get theoretical guys [sic] to investigate what is needed for a devise to reproduce Turing patterns. I have designed a lot of devices, yet I only intuitively know that they can generate Turing Patterns. If it can be showed that those devises are able to produce Turing patterns we would have made a big step. I want to remark that, in my opinion we should not be attached to the model that is shown in most papers, after all no one knows the mechanism that underlies the formation of “Turing patterns” in organisms.
Federico C 27 may 2008 (EDT)
Does it work? While talking with Juan Arias he expressed his concerns about the functionality of the devise and suggested analyzing and using parts and devises that other teams have used. Here I invite everyone (not only team members but also the OWW community) to post links to papers and works that could be useful.
- Prakash group IAP 2003 Designed some oscillators dependant on diffusible signals that are now available in the kit plates, a parallel work to the one carried by the Mexican team on iGEM 2007
- McGill iGEM 2007Designed a simple oscillator dependent on diffusible signals.
- Tokio Tech iGEM 2007 Designed devise that would allow a population of bacteria to differentiate based on diffusible signals. Theoretically the population would be able to maintain an equilibrium between two cell lines.
- Paris iGEM 2007Designed and assembled a device that allows a population to differentiate onto a germinal line and a somatic line. Very interesting, besides the DAP could be used for cell communication.
- Cambridge iGEM 2007Developed some parts for cellular communication based on peptide signals, they might solve some problems associated with Lux/Las systems. Unfortunately Cambridge parts did not made it to this year’s kit plates.
- Tsingshua iGEM 2007 Designed a devise that would allow communication between bacteria by carrying elements during conjugation. Also designed a simple oscillator.
A short answer to the Problem #1
Luis de Jesus Martinez 28 may 2008 (EDT)
Well, first to say that the literature about Turing patterns explicity specify how they should be generated and emerged (it has been proved with different chemicals); obviously the model is expressed in Partial Differential Equations... I really think that we do not need a promoter with that behaivor, the model of the autocatalizer-inhibitor is esentially an abstraction, we can go down in it and create a system that behaves like the autocatalizer with more than one promoter; Juan and me are working in it. And about the silver bullet... our teacher Pablo has told us about it (with the help of the physics), the key is to see that we do not necesary need difussible signals, only a device that with some probability could get some stage of equilibrium... I will explain it later. I must go to sleep now.
Federico C 29 may 2008 (EDT)
Surely we can get patterns without intercellular communication...the rounded colonies that grow in petri dishes show a structure (rounded!) and a pattern (dispersed colonies), albeit a simple one.
If you add a devise for differentiation you will be able to differentiate single colonies or parts of one colony but if you want that you might just mix green and red bacterias and you will get the same result.
You could generate complex patterns without cellular communication with an external signal that would trigger a differentiation (such as the Turkey's las year iGEM project) but then you will not be using the putative homogeneous media that Turing patterns require.
You could also obtain patterns by regulating the way and rate by which bacterias divide (which is kind of difficult) or by killing some (easier) but I don't see how that is related with Turing patterns.