Here's an example from my work at the moment. One of the pigments I'm looking at is violacein - a purple pigment found in marine bacteria. Four genes are used to make this pigment, the first two take a molecule of tryptophan and modify it, the third joins the two modified molecules together, and the fourth further modifies this structure. This can be shown diagrammatically:
VioA, vioB, vioC and vioD are the names of the genes. This diagram also shows that there is an intermediate colour within the pathway. If vioC is knocked out (so the gene can no longer function) the cells produce cyan pigment rather than violet.This gives two separate 'states' for the cell to be in. The engineers start getting excited about this, because vioC needn't be completely destroyed in order for cyan to be produced, it can be put under the control of a certain input system, e.g an arsenic sensor. This means that the cells will normally be cyan, but in the presence of arsenic vioC is expressed and they turn violet. Instant biosensor. Easy to use, and easy to interpret.
As well as being explained biologically this can also be represented diagrammatically:
I like this diagram, because it shows just how useful a set of four genes making two colours can be. The vioA and vioD genes act as an AND gate. Putting them both under different input systems means that you can engineer bacteria that only turn cyan when two conditions are met; e.g correct temperature and correct levels of (say) iron. The vioC is then a further AND gate. Or...if the vioC is controlled by a repressor instead of an activator, a NOT gate. The possibilities are endless...The really great thing about this is that it provides a level of abstraction for designers and engineers to work with biological systems. They don't need to know about the gene sequence, or the biological basis of gene activators and repressors. All they have to do is work with the blocks.
(Thanks to a fellow Lab Rat for the images!)
8 comments:
Ooooh. Very impressive. I'd heard of DNA computing before but I'd not heard of thinking of genes like this.
I'm guessing it might have trouble scaling up to 1000s of gates as you'd need a separate gene for each 'wire'/connection?
Yeah...genetic computers are a long way off...if they ever happen at all! Some of the iGEM teams in the past build things like memory stores and random number generators with bacteria but they were always less helpful/useful/workable than the electronic versions.
Less helful/useful/workable isn't necessarily a problem though? Isn't some science about just doing things because you can? Or is it all funding and commercial results driven now?
"Isn't some science about just doing things because you can? "
Add the words 'get funding' to the end of that sentance and you've pretty much got the truth :p The beauty of iGEM is that it's just kids playing around so you can get funding for all sorts of crazy ideas. in an actual lab setting it would be less easy ... even in science you still have to justify why you do what you do and "because it's fun" doesn't *quite* cut it.
Hi Lab Rat,
I'm curious about the last diagram depicting logical gates. Do you think of the genes themselves being the inputs to the gate? From the description in your paragraph, it sounds like arsenic and other signal molecules are the inputs, while the genes/complicated machinery of the cell are the implementation of the gates.
I believe construction of such projects requires knowledge of engineering and management principles and business procedures, economics, and human behavior.
@Markkimarkkonnen
We have two engineers and three biologists on our team at the moment, and this diagram was drawn by one of the biologists :) Yes...the input would be the arsinic, or whatever molecule we were sensing. However we found it easiest for gene design to think of *gene acivation* as the input, rather than random as-yet-to-be-decided molecules.
Gene translation and protein production would then be the process caused by this input, and the colour is the output.
~Lab Rat
got it. thanks.
Post a Comment