Remember yesterday when I said Solaris 11 goes way over my head? Well, today we’re talking about where operating systems and biology intersect. Scientists at the University of Nottingham are trying to develop the in vivo biological cell-equivalent of a computer operating system. The project’s name is AUdACiOuS.
What does this mean, exactly? “We are looking at creating a cell’s equivalent to a computer operating system in such a way that a given group of cells could be seamlessly re-programmed to perform any function without needing to modifying its hardware,” explains Professor Natalio Krasnogor, who leads the effort, “We are talking about a highly ambitious goal leading to a fundamental breakthrough that will, ultimately, allow us to rapidly prototype, implement and deploy living entities that are completely new and do not appear in nature, adapting them so they perform new useful functions.”
This technology could potentially have a huge impact on the world. It could lead to to the development of new food sources, special drugs tailored to patients, growing new organs for transplant patients. Furthermore, it could be used to solve environmental problems.
“Currently, each time we need a cell that will perform a certain new function we have to recreate it from scratch which is a long and laborious process. Most people think all we have to do to modify behaviour is to modify a cell’s DNA but it’s not as simple as that – we usually find we get the wrong behaviour and then we are back to square one,” Krasnogor further explains, “If we succeed with this AUdACiOuS project, in five years time, we will be programming bacterial cells in the computer and compiling and storing its program into these new cells so they can readily execute them. Like for a computer, we are trying to create a basic operating system for a biological cell.”
The research paper, titled “A computational study of liposome logic: towards cellular computing from the bottom up“, is available online – although I’m not sure if it’s free (my university account took care of it).
Even as a computational biologist, I really have no idea what they are planning on doing from reading this.
I can guess that they will be using models of regulation (and maybe metabolism if they are good, like this: https://simtk.org/home/ifba ) to aid in the design of some phenotype (function).
Biologically, they will probably be putting a lot of the transcriptional regulators under control of non-metabolic chemicals like doxycycline, though I wonder how many different chemical regulators they could use. If they really wanted to make this cell very programmable, they would need a lot. However, they could potentially engineer the TFs to use combinations of several such chemicals, and 10 chemicals could give you 2^10 – 1 = 1023 different signals, which is already a large portion of the E. coli genome. This approach would involve the overlap of signals (i.e. a superset would regulate a subset) so it probably wouldn’t be so useful.
A eukaryotic cell that could use shRNA to turn off any mRNA signal seems like it may allow more flexibility, though perhaps they could engineer in a DICER complex to E. coli (this was done in a baker’s yeast about a year ago, but that is also eukaryotic…). This is probably a pretty inefficient way of programming though, since it would assume the default signals were “ON”, which would result in a lot of RNA loss and I think shRNA is probably not so cheap iirc.
As for Solaris 11, I just updated my home server system to it last night while playing Skyrim. There seem to be quite a few differences in Solaris 11 and the last dev release of Nevada I was using. The news threads these days are hitting very close to home …
Edited 2011-11-11 20:29 UTC