PerleDeDiamant

PerleDeDiamant

How Gene Circuits Store Information

How Gene Circuits Store Information

Posted: 19 Sep 2010 09:10 PM PDT
http://feedproxy.google.com/~r/arxivblog/GmoU/~3/92dobRszV-0/click.phdo?utm_source=feedburner&utm_medium=email


Certain types of gene circuits can store information over many cell 
generations, according to a theoretical study. That will trigger a frantic 
search to find these circuits in real cells


One of the great puzzles of cell biology is how information is stored, 
processed and passed from generation to generation at the biochemical level. By far the most famous mechanism is the sequence of nucleotides in DNA. 
However, in recent years a number other data storage mechanisms have 
emerged, so-called epigenetic processes, and their role is under fierce 
debate. For example, the pattern of methyl groups attached to DNA seems to  be an important data storage system as do modifications to the proteins  that control how DNA is packaged.

Today, Georg Fritz at the University of Cologne and a few buddies put 
forward a new idea. They say that a simple network of genes can act as a 
conditional memory, that either stores or ignores information when it is 
told to do so. "The circuit behaves similarly to a "data latch" in an 
electronic circuit, i.e. it reads and stores an input signal only when 
conditioned to do so by a "read command," say the group.

This memory circuit is essentially a modified version of the well known 
genetic toggle switch which consists of two mutually repressing genes and 
can exist in one of two states. The difference is that this circuit is 
controlled by the concentration of another protein. At high concentrations, 
the circuit is primed to store data in the form of another protein. At low 
concentrations, the circuit ignores what's going on. That gives an extra 
layer of control over the simple toggle switch.

This a very simply design that has the potential to perform complex logic. 
Indeed, Fritz and co go on to show how such switches can combine to do just that. This is sequential logic which depends crucially on the history of 
input signals (as opposed to combinatorial logic which depends only on the  present input signals). That's potentially very important. "Conditional memory would then enable cells to manipulate information "collected" under different conditions at different times," say Fritz and co.  In fact, exactly this kind of behaviour is observed in certain types of cell. We've looked at one example of apparently "intelligent" cellular behaviour that has exactly this property.

Fritz and co even go on to work out the properties that a conditional 
memory circuit should have. It should be able to respond to changes in the environment over a timescale of 30 minutes or so and yet be able to store information over many cell generations.

But here's the rub: Fritz's work is entirely theoretical and that raises an 
interesting and important question. Do conditional memory circuits actually exist in real genetic networks? Nobody knows.

They easily could and some fairly straightforward work should reveal their presence. If they don't show up, that'll be interesting too. It'll mean 
that nature has found some crucial flaw in this circuit design or a better 
way of storing information.

Time get looking.

Ref: arxiv.org/abs/q-bio/0701011: Designing Sequential Transcription Logic: 
A Simple Genetic Circuit For Conditional Memory



20/09/2010

A découvrir aussi


Inscrivez-vous au blog

Soyez prévenu par email des prochaines mises à jour

Rejoignez les 3 autres membres