IGEM:Melbourne/2008/ProjectIdeas/BiologicalClock

Return to Melbourne Homepage

I have been doing some thinking on this idea and there are a few things to consider.


 * First, the 2006 team from ETH Zurich were working on a half adder that may be worth reviewing as a clock could be thought of as a series of adders that you keep adding one to.
 * Thanks for this info, I'm trying to have a look what they have done, at the moment experiencing problems opening their page. I'm keeping trying.


 * Secondly, the 2007 Peking team were working on a push-on/push-off switch (single signal switches state in both directions) that could be very useful, as we want each component to change its current state if the previous component switches off.


 * Alternatively, for a single signal switch we could consider the mating type switching system in yeast in which the expression of a single gene induces a cell to switch to the opposite mating type. This would be quite complex to do however and may in fact be impossible with the time we have.


 * With regard to sensing when the previous component switches off, I have one idea for an implementation but don't know if it is feasible. If the protein for the ON signal for each component was not degraded too quickly but the OFF signal was, the component essentially has a short-term memory of when it was switched on. Thus, going from ON to OFF, both proteins would be present for a short time. This would not be seen going from OFF to ON as the OFF protein would be degraded quickly. Therefore the presence of both the ON and OFF signals could be used to indicate that a component has just switched itself off, acting as a trigger for the switching mechanism of the next component in the system. Any thoughts on how practical an idea like this is?
 * YES! I think it makes good sense. So bacteria of the next bit only change state when they see both ON and OFF proteins present from previous toggle switch, yeah? (and that's when previous bacteria change from ON to OFF, yes?) But I just what to clarify one thing, when you said "the component essentially has a short-term memory of when it was switched on", you meant the component as the toggle control not the bacteria that's holding the timing information right? (Because each bacteria (apart from B0) must have a long-term memory of its current state (i.e. timing bit) until its previous toggle switch tells it to change state.) I'm looking forward to hearing more of your explainations at the meeting.
 * however with the setup as peking 2007's pushon/off switch, is there need of different signalling molecules for ON/OFF? because as peking did, all they did was using UV to switch on/off, so if say we use UV to control on/off for B0, having GFP as off, mRFP as on, replace BFP with the signalling molecule (x) for B1, which acts the same way as UV for B0, then x is only expressed when B0 is switched off, and can turn B1 either on or off. same thing can be done for B2. also if we put everything onto one plasmid, so one single bacterium can express all three types of fluorescent proteins, does it solve the synchronisation problem?
 * For B0, I am thinking to let B0 to generate some signal to turn itself on and off periodically, instead of using an outside control, so that the clock can sustain itself. We could use the repressilator as discussed in our last meeting or something else that could ossilate much faster. I'm still trying to follow up your idea of solving the synchronisation problem. Also I'm trying to read the Zurich's 2006 Half Adder to see how different their actual way of implementing binary operations is to our proposing way, praying it would be different...


 * Just to extend the counting idea a bit more. Currently the digital clocks or pocket calculators have a theoretical hard-wired limit to which they can count or add up to (e.g. if we type $$10^{100}$$ it will give an error), because they have fixed number of bits in their registers/adders. We could make our bacteria such that they know to differentiate into a new type when a new bit is needed, for example if we start with 3 bacteria B2B1B0, they can count up to 111 (i.e. 7), in order to count 8 (1000), there need to be 4 bacteria B3B2B1B0. If we could make B2 to automatically generate B3 (B3 is initially in OFF state) when B2's active state is turned on, and B3 to generate B4 (B4 is initially in OFF state) when B3's active state is turned on..., then there is no theoretical counting or the computation size limit of the system. After B4 is made it also needs to repress B3 to make more B4 since it's already there. I just have some very primative idea on using the lengths of something e.g. telomerase (there will be better choices I believe) as the tagging label and communication information between different types of bacteria to tell each other which type of bacteria already existed. I know this will be hard. But if we could make it work, our clock can supercede a normal clock on a theoretical level:)


 * Bacteria dont have telomere and so you may want to think of synthesis/cleavage of other biomolecules.


 * Regarding the problem with synchronisation, what if we were able to design bacteria so that every time they underwent binary fission, the genetic material went into only one daughter cell? Fission actually occurs once the DNA has been replicated, so perhaps by inhibiting this replication, we could ensure that only one of the two daughter cells contains the DNA and so the other one will automatically die. Thus at any time only one bacterium will exist of each type (i.e. 1 of B0, 1 of B1 etc.) But I’m not sure if it is possible to isolate a single bacterium to start off with. And also, I’m wondering now if the daughter cell that survives will keep to the same timing as the parent. Maybe we could somehow synchronise the fission of the 3 (or how many ever types of) bacteria so that at least relative to one another they will keep the same time. Hmm not sure how practicable this is.
 * But inconsistency could still result between generations once they are all there, wouldn't it? I mean if we had three generations of bacteria 2: B2 B2` B2``, both B2 and B2`` received update information from B1 but B2` didn't, then there's still going to be inconsistency.
 * If we were able to ensure that every time fission occurred, only one of the daughter cells survived, we should never see the existence of different generations of the same type of bacteria at the same time. 1 generation must replace the next and so on. But I’m not sure how and if we could create this (I guess it will have to be along the lines of stopping DNA replication as I mentioned earlier and at the meeting). My concern is that even if we synchronised the fission of the 3 bacteria that the daughter cells would not necessarily keep the same time, but I guess that depends on the design of the ‘clock’ itself. What do you think?(Note, creating this synchronisation is independent of controlling fission so that only 1 daughter cell survives.)
 * Yes it makes good sense.