As an engineer, I think the counter is completely useful and totally cool as well. I'm really
excited at the prospect that it might actually work (which I frankly still doubt a little bit at
the moment). The SO-Constructor definitely is cool as well, but I don't see it's usefulness yet,
but as I'm no bilogist, it is not mine to judge.
Talking about modularity, I see that the SO-Constructor has a lot of modules, but they are not
'parallel' modules, but 'serial' modules, i.e. if one module fails, the overlaying modules won't
work as well. They strongly depend on each other. I also see the problem of tuning, because there
is a real lot of tuning to do and nearly everything needs to be tuned. For me, some questions
arise like: Can you simply change the threshold for a certain protein in quorum sensing? Anyway,
the probability that we have some results to present in the end is still relatively high.
I don't think the counter is really modular, altough it is a combination of one basic part (the
'thing'). The point is, if that basic part works, the whole counter probably works as well. If it
doesn't, then we don't have anything at all. But shouldn't we just take the risk and try it? No
risk, no fun :-)
Conclusion: The SO-Constructor is in my opinion the 'safer' option, meaning that we are much more
probable to get something that actually works. Whereas the Counter is much more useful and
revolutionary, at least when we get to work it somehow.
As I already have said during the meeting, I am a little bit concerned about the effects of the
different presentation styles and levels of detail: it is hard to judge the feasability of one or
the other, since for the SO-Constructor I can see a clear stepwise approach with controllable
biological units to implement, but I am doubtful about the tuning and proper interplay of all these
components - although I am confident that to some degree this can be achieved.
As for the Counter I am also not sure about the basic biological assumptions, e.g. whether the
toggle switch units are available and will work as well as the symmetric repression/activation
that is necessary, although the modeling part suggests everything is straightforward - but this
might be deceiving.
It is hard for me to judge the biological unpredictabilites and risks, but from an engineering
point of view the Counter building block is of the more useful thing to do, since it has a broad
application range thanks to the clear interfaces.
Last but not least such a technical unit seems more suitable in the context of Synthetic Biology.
I think if the whole thing works this project is astonishing. But I'm in doubt about it. The fact
that you could model each step seperatly is of course a big advantage of this project. The
probability that at least a few of the steps work is quite high and so you will see some results at
the end. I'm not so sure about the usefulness. To encapsulate some bacterias which produce a
certain substance could be usefull but I'm not an expert in this stuff.
Addition: Hervé, I think those who said, that they doubt about the feasibility of the
Constructor Bacteria didn't mean, that it doesn't work at all. I'm rather sure that at least
some parts will show the wanted behaviour but I also think that it is unlikely that really the
whole thing works. Further ideas and knowledge about how we could implement both projects in detail
will increase the probability that we will succeed (for both projects).
I like the general idea and concept of the SO-Constructor project very much, even if many team members
expressed their doubts about the practical benefits (as Urs mentioned above it could be used for
substance (e.g. drug) encapsulation. I'm not sure, but I think that one may also use the whole thing
for detection and localization purposes (e.g. of toxic substances)). As others pointed out already,
overall feasibility, especially the expected difficulties in connecting the different modules, could
be a drawback here. Dominic, Giorgia and Herve have really elaborated an excellent model, but to me it
seems to be very complex and I don't know if we could manage to set the whole thing up in time (on the
other hand this does not seem to be a general concern because of the "intermediate results"). But
since my knowledge of (applied) molecular biology is very limited and I don't know how (non-)volatile
the counter circuit is, my personal feasibility estimation stands on thin soil.
The presentations made on Thursday had a different profile in terms of details and analysis.
It is obvious that the constructor group had focused its energy on the feasibility and the
options available to circumvent the problems that will arise during the development of this project.
It is also clear that we refused to present the details about all these options because we wanted
to discuss about the concept only. I would recommend people having some doubts about the feasibility
to check the detailed information available on the wiki. Concerning the usefulness, I do think that
both projects are extremely useful and I have some problems to understand that people are questioning
the usefulness of the constructor bacteria project (sorting and purification is a key issue for biotech
companies for example).In order to make the appropriate choice, I would still need more information
about the availability of the modules involved in the "thing" project. Do we really have the repressors
and activators required for this project? Which ones are we gonna use? Considering that it is dificult to
have activators and repressors with the same "strenght", how could this affect the model and the
feasibility of the "thing" project? It is really interesting to notice that people have better feeling
about the feasibility of this project (compared to the constructor bacteria) even though the biology
part (which is, in the end, the core and working part of the project)has been more or less skipped
during the "thing" presentation.Regarding the concept of the synthetic biology, I have to agree
that the "thing" project sounds much more appropriate for this kind of competition. This would be
in the end the criteria that would push my vote into this project (after a deeper analysis of the
feasibility and availability of the needed activators and repressors).If we decide to go for this
project we can of course reduce the risk factor by taking different activator - repressor candidates
and maybe trying 2 or 3 combinations that have passed through the modelization test.
Herve, we have not looked for actual repressors/activators because we believe that we should first concentrate
on the design, at the "functionality" level so to say. Now that that part is more or less clear, well yes, you are
right, we should start looking at the implementation. And when I say "we", that's everybody, and in particulary you
and those who have more experience in the lab. I just had a quick look at pubmed, and i found a paper that
mentions a protein ("TyrR") that acts as repressor and activator (see ref below). Another potential protein is
"Cra", that can activate or repress depending where it binds to the DNA. Maybe we can work with a fusion
protein from an activator and a repressor. Maybe we can solve the duality by using an activator that is required for
the expression of genes 1/3 but represses genes 2/4, that are otherwise expressed constitutively, by placing its
binding site at a position that overlaps with the polymerase binding sites, etc...
What are *your* suggestions? Can you look into that issue? Don't you think we can solve that problem?
update: maybe there is a trivial solution to this problem: we could simply have *both* a repressor and an activator,
controlled by the same promoter (i.e. heat shock promoter by hand, cell cycle promoter, repressilator) in an operon
type of scheme. The repressor and activator would target respectively g2/g4 and g1/g3. g2/g4 would have a constitutive
promoter that is suppressed by the repressor, while g1/g3 would require the activator for expression?!?
Refs: TyrR protein of Escherichia coli and its role as repressor and activator.
The catabolite repressor/activator (Cra) protein of enteric bacteria
Although I like our baby very much :-), I must agree with Dominic and Hervé that "the thing" would possibly be a
conceptually more appropriate project than ours.
Speaking of usefulness, I think both “the thing” and the “constructor x” could turn up to be very useful, so that's not
really my point. What I am afraid about is that even the model looks perfect the implementation of the thing in the real
world would be very tricky. As Dominic stated during our presentation, we introduced the most complicated variant of our
concept. Therefore, many could have had the impression that feasibility is low. However, we looked for alternatives and made
up some easier ways to implement each module. Moreover, debugging in this project is possible for each module, what would
ensure having some result to present (that would also be nice!).
It is also my opinion that the presentation of “the thing” lacked of important details on biological background and
implementation. I don't really think Cra would be an option. Since it controls the expression of many genes concerned with
carbon and energy metabolism, every time that this protein is present in the cell it will affect the status of the cell and
its behavior. Moreover, cra promoters are also catabolite repressed or activated (e.g. when fructose-1-phosphate is present
in the cell, then cra activated operons are repressed and vice versa). I think it is very important that activator/repressor
does not interact with possibly any other gene promoter in the cell, and that promoters regulated by our activator /repressor
should not be activated by the interaction with other molecules that could be present at any moment in the cell. Also Tyr
would not be an ideal option because repression is mediated also by tyrosine and other amino acids (e.g. Phe), which
concentration in the cell is difficult to assess and regulate. Moreover, TyrR is also responsible for repression/activation
of transcription of genes responsible for biosynthesis and transport of aromatic amino acids.
Well, I don’t want to give you the wrong impression: I don’t think implementation of this concept is impossible! I just want
to point out that this kind of information is very important (at least for me) in order to be able to make a decision...
The constructor is a neat idea and could potentially be useful (e.g. drug delivery, component of paint, etc.).
The quite elaborate design of the constructor could be hard to bring to function and in such case we should
focus on one part that we believe is feasible. Which part should this be? perhaps aggregation and cellulose
expression => balls? I like the innovativeness of the project, but i believe that the coolness would sink if
we were not to realize the complete project. But then again, we should vote for the project that everybody
like the most and make it happen, i.e. not argue over details. After we have a decision we just make it happen.
From the engineer's point of few a counter like this is of course a crucial brick. As you can take
the concentration of any substance you like (by transforming it to our input S) and also add for
every single counter step a specific gene needed for an application this counter can be used
whereever you want. The big disadvantage is of course that if the counter doesn't work there will be
no result you can see.
Addition: I think the question we have to ask our selves (if the meeting with Randy Rettberg
doesn't show any results) the following: Do we want to create something very useful that will be
reused very often and (in combination with the cell division) meet in a nature article
according to Sven but risk that we fail and come up with no positiv results? Or do we want to create
something maybe very useful for the industry, which has a couple of parts more or less likely to
work but will show at least some reactions and results? Both have their pros and cons and in the end
it is really (as Dominic said on Thursday) about which direction the Synthetic Biology wants or
should go: Build a biological super computer or create special behaviour out of bacterias.
About feasibility: i don't agree with much of what has been said here. The counter is made of two
toggle switches. These switches exist in the nature (lambda repressor) and have also been synthetized
(Gardner et al., partly also on MIT registry). Now, the only things we need to come up with are A) a way
of having 4 different types of road blocks, which is, as Sven said, possible using zinc-fingers that
target different dna seqs, and B) a way to have S working as an inducer and as a repressor, in a
roughly symmetrical manner. And that's all. I am not saying that it is trivial, but seriously, it is
really doable. And even if we don't manage it to do it for the jamboree, finishing it later could
still be academically very rewarding.
Bottom line: The difference between the two projects, i guess, is that most of the work in the QS project
would be at the assembling/tuning of existing parts for a very cool final effect, while in the the counter,
we would focus the energy of the whole group into realizing a little module that is reliable and reusable. Both
projects are cool, both teams have done an increadible amount of preparatory work until now and so at this
point, i would be very excited working on either projects.
A stable counter would be a very nice thing to have. In my eyes this project seems to be a bit more feasible,
but maybe that's a misjudgement. The model as elaborated by the group is detailed and the simulation seems to
be quite promising. On the other hand the "all-or-nothing" scenario makes it a bit risky.
PS: even though I didn't give a 10 for coolness, I do believe that it's a Nature candidate.
An counter is such an obvious part in the realms of synthetic biology it is
almost surprising that it has not been described earlier. Does anyone know if
any other team is working on something similar? A counter would be tremendously
useful. The design and idea do not look too difficult to implement, it is
has modularity and could be used in conjunction with many other devices.
I like it!
We asked Randy if such thing as a counter exists already, and he said that he is not aware of a counter.
For sure, there is no guarantee that no other team is trying to build a counter in this moment, but I
think chances are good that we're the only team that would do this in the moment.