Production of materials is relatively easy - genetic engineering has been doing and refining this for a relatively long time. Producing materials in useful 'formats', however, is slightly harder. This page lists some possible methods for directing bacteria, allowing them to construct tertiary structures or lay down materials in a pattern.
This page includes an overview of ideas that were touched upon in our brainstorming sessions, and expands on them.
Motile Control Mechanisms
It could be possible to steer bacteria using a magnetic field, if they were first polarised somehow. A shifting magnetic field should allow fine control of the movement of bacteria - the field would need to shift in line with the bacteria's movement speed however, and to regulate it may require a reporter gene being present (so they can be directed as they go).
Here is a paper which confirms the possibility of introduction of genes coding for magnetosome proteins, from naturally occuring magnetotactic bacteria into E.Coli. http://www.calpoly.edu/~rfrankel/NatRevMicro.pdf
Magnetosome, like the other organelles found in a cell, i.e. mitochondria, smooth ER etc. is responsible for the production of fero-magnetic substances within magnetotactic bacteria. These substances include magnetite and greigite, of which magnetite has a stronger response to magnetic fields. The production of such materials allow bacteria to be polarised and allow them to move along magnetic field lines of the earth. This is achieved by placing the organelle, magnetosome close to the flaggelum which contributes to bacterial motion.
MAGNETOSOME FORMATION IN PROKARYOTES, Dennis A. Bazylinski* and Richard B. Frankel‡ 
Monodisperse FePt nanoparticles and ferromagnetic FePt nanocrystal superlattices
Similarly to magnetism, using an electric field could allow control over the movement of the bacteria - this may be easier to regulate than magnetism too, as you can change the strength of the field with ease.
Chemotaxis is an interesting possibility for controlling direction. Erika's mini-iGEM project (presentation on Thursday!) involved a chemotactic "dot-to-dot" set up, whereby bacteria were directed to a point (where a nutrient source was placed that they were attracted to) up a gradient. Quorum sensing was then used to trigger both an excretion of a marker protein (to draw the dot) and a shift in the attractant - so they would start moving toward a second point, the position of a second attractant.
In B.Subtilis, is was shown that CheY null mutatns were in a constant state of tumbling, whereas CheY-P was responsible for the continuous swimming of B.Subtilis. If we can find a way to control the expression of CheY in B.Subtilis, then we will be able to control the tumbling motion of bacteria and thus guide bacterial clusters to areas of interest.
Quorum sensing is a very valuable tool for this situation. With it we can trigger clustering or dispersal via a link to the flagellar motion of the bacteria (see the flagellar clutch below), and theoretically we should be able to arrange our bacteria in a monospaced array... This could be very useful for producing e.g. collagen or a similar product.
The Flagellar Clutch
1. 2005 Penn State Uni team did a similar project, but they didn't use EpsE/SinR gene expression. Instead, they played around with the actual motor of the bacteria, motB gene which generates the torque required to give flagellum its rotating power. http://ieeexplore.ieee.org/xpls/abs_all.jsp?arnumber=4290622
2. Here is a simple explanation of how the clutch system works: http://www.sciencemag.org/cgi/reprint/320/5883/1599.pdf What happens is that the gene SinR upregulates the expression of flagellar genes i.e. MotA/MotB etc. and downregulates expression of biofilm forming genes. If SinR is absent, biofilm takes over and the bacteria loses its motility. It was found that SinR represses the epsE gene, thus the absence of SinR causes the derepression of EpsE.
It was later determined that EpsE is responsible for disengaging the clutch in the flagellar motor. See http://www.sciencemag.org/cgi/reprint/320/5883/1636.pdf for more. This means that instead of turning the motor on or off, bacteria's motor is continuously spinning, but disengaging or engaging the flagella to the motor is regulated by EpsE and thus the epsE gene. If we can characterise the epsE gene, we should be able to control whether the bacteria moves or not, irregardless of its orientation, which brings us to the next problem.
If one begins with a regular array of bacteria on a plate, the triggering of production and secretion of the biomaterial required could be linked to light - this could, conceivably, allow us to produce materials in a very complex pattern, and also allow gradients of production (should that be required). Repeating this multiple times for different layers could, therefore, allow us to build up a construction (not unlike a 3D printer ) of our material.
In 2000 a light-activated proton pump called Proteorhodopsin was discovered. In the absence of cellular respiration, this pump is activated by illumination, generating a proton motive force which drives the flagellar motor. However, there is no effect of light when respiration is not impaired, and there are no direct links between this light sensitive protein and EpsE which we are trying to control.
With respect to B.Subtilis, the following were found: YtvA from B.subtilis is a flavo-protein related to the plant blue-light receptors phototropins. YtvA is a positive activator of the general stress transcription factor σB. In B.subtilis, the transcription factor σB controls more than 200 genes. Blue and not red light induces σB activity. We could use σB which is expressed upon blue light illumination to exert clutch control. YtvA-SigmaB activation pathway:
Adapted from Photosensing in chemotrophic, non-phototrophic bacteria: let there be light sensing too Michael A. van der Horst, Jason Key, and Klaas J. Hellingwerf
New Insights into Metabolic Properties of Marine Bacteria Encoding Proteorhodopsins (Would be good if someone could decode what this paper is saying)
Light powered E.Coli with Proteorhodopsin 
Blue Light Activates the σB -Dependent Stress Response of Bacillus subtilis via YtvA. Marcela A´ vila-Pe´rez, Klaas J. Hellingwerf, and Remco Kort*
First Evidence for Phototropin-Related Blue-Light Receptors in Prokaryotes. Aba Losi, Eugenia Polverini, Benjamin Quest and Wolfgang Gartner 
Listening to the blue: the time-resolved thermodynamics of the bacterial blue-light receptor YtvA and its isolated LOV domain. Aba Losi, Benjamin Quest and Wolfgang Gärtner 
Photosensing in chemotrophic, non-phototrophic bacteria: let there be light sensing too Michael A. van der Horst, Jason Key, and Klaas J. Hellingwerf 
Combinations of the above techniques allow for powerful possibilities for control.
Production of fibres would be useful, as we use fibrous materials a lot (textiles/nylon for example). Spider silk is the most notable example of a fibrous material that would be incredibly useful to make. Production of fibres, however, would be very hard from a Synthetic Biology point of view... If one were to attempt to make a fibre by combination of fibrils, each being produced by a different bacteria, exceptionally fine directional control would be needed. In addition it would be very hard to produce a long fibre from a bacteria, considering their mode of expulsion of product (sort of like a reverse-phage, a needle of protein from within pierces the outer membrane and product exits through the needle but the needle can pierce anywhere on the bacteria. Additionally you'd need all the bacteria to swim the same way while spinning out the fibrils at an approximately constant rate and it's probably not feasible. It may be better for us to focus on other materials, therefore.
Construction of laminar patterns, designs or sheets of biomaterial would be very valuable indeed - particularly, perhaps, in the medical field (production of collagen for instance, skin grafts, and so on). This might be the most exciting application or area of production of biomaterials that is feasible, with the complexity of 3D construction being inhibitory! Having said that...
Perhaps the most exciting idea is that of a biological 3D printer, as referred to above; building up a structure, layer by layer. With the addition of a number of different materials being produced, it may be possible to form complex structures with different components. Layer by layer construction would take a long time, however, and it's uncertain whether each excreted protein could be connected to the ones around it.
Approaching from a different angle, one might envision a block of agar laced with nutrients that trigger production of a biomaterial; and bacteria forming up via chemotaxis/quorum sensing and excreting their product in a shape as defined by the presence or absence of certain nutrients, perhaps.
A key possibility for construction of a complex shape is the utilisation of a scaffold. It has been suggested that a scaffold of chitin would be the most likely possibility, as it is produced by fungi (as well as insects, crustaceans) and therefore probably does not require any complex post-translational modifications. Building a scaffold and then laying a more detailed product on top is a valuable idea.
Many of the problems are similar across all methods. Most obvious, perhaps, is that bacteria divide and that this could mess up our steering. One solution might be to produce a two-state bacteria - in one state it can divide but not move, and in the other it can not divide but can move about. Thus you could start with a fixed colony of immobile bacteria, breed them up to the desired amount, then flip the bistable switch with a trigger and put them into the motile state. They would stop dividing, and would move as directed (well, hopefully). Peking's 2007 iGEM project's push-on push-off switch could be ideal for this.
YtvA activates sigB which then activates the stress response in B.Subtilis. Few of the concerns are with respect to the kind of stress response elicited upon blue light illumination.
1. What are the other regulators of sigB (besides blue light)? In general, sigB is activated by diverse environmental and energy stresses to direct the expression of a large set of general stress genes. We must be careful not to activate the sigB transcription factor under these stressful conditions while culturing B.Subtilis. Perfect conditions w/o activation of sigB must be provided if we are to use the pathway in controlling motility.
2. There are around 200 genes controlled by sigB. Will these genes affect cell motility and interfere with biomaterial synthesis? Main category of genes activated by sigB include genes whose products aid against oxidative damage, protection of DNA and controlling solute transport. The general stress response has an impact on the expression of genes coding for known or portential transcriptional and post-transcriptional regulators, for transporters involved in ctronolling solute influx and efflux, proteins involved in carbon metabolism, envelope function and macromolecular turn-over.
3. Are there other ways to make B.Subtilis responsive to light w/o going through the sigB pathway? ΔsigB mutant strains PB153 was constructed in '91, other strains which have been used include those where sigB is turned on by IPTG.
The Pitch (11/07/08)
I think we are much more likely to be successful in our pitch if we break the whole idea down into separate "checkpoints" toward the final goal of directed material synthesis. For instance if we try and achieve the following aims, in order...
- Culture B.subtilis in the lab and get it working for our experiments, make it a viable vector/chassis;
- Characterise and control the epsE pathway (perhaps also SinR) and biobrick that part;
- Control expression of epsE, and hence movement of the bacteria, via an outside stimulus (probably light, and try and biobrick that too);
- Use that control to arrange the bacteria and have them express a reporter when triggered to do so, if epsE is present (the bacteria are stopped) thus producing patterns (a la Cambridge 2006) on a plate;
- Swap out the reporter for a biomaterial synthesis pathway, completing our entire project and producing some material in a pre-defined pattern; phage display..?
...then even if we didn't get all the way through each step is an important advancement for Synthetic Biology - from the first to the last.
The pitch will begin, therefore, with an overview of our grand plan, the possibilities of completion of the project (the wow factor!). After that, however, we'd demonstrate the breakdown of our project and show that each step is achievable and that really, any progress along the chain of aims listed above would have a shot of winning - the more the better! The project is therefore very tractable, because it doesn't require a set time-frame. We work with what we've got and present whatever we come up with - the rest being theoretical (but the groundwork research having been done, most probably).
An additional problem with B. subtilis, specifically making a chassis from it, is that Newcastle are doing that for their project this year... They're engineering a B. subtilis chassis to detect and destroy Gram-positive bacteria (C. difficile, S. aureus for instance). This could be a benefit, however, as we may be able to collaborate to some extent - Gold medals all around..?
Major hurdles will probably include the first step (see Cambridge 2007). If we find B. subtilis to be unworkable (and all evidence points to Cambridge... shooting themselves in the foot, rather than the bacteria being belligerent!) the we will still have time to cut and run - switch to filtration, or something. Additional hurdles will be encountered with trying to control the bacteria with light, in all probability, as there aren't any biobricked parts out there for light sensors in subtilis.
Lastly the phage display possibility will be very complex and possibly cutting-edge newly experimental. We'd probably have to leave this theoretical but it'd be a great thing for a "future work" slide. Matthieu seems to think we'd be better off pursuing this in parallel with the main project, so I guess we're going to need more on this it turns out!
Self-assembling Peptide Nanofiber Scaffolds for 3D Tissue Cell Cultures (12/07/08)
Ile-Lys-Val-Ala-Val (IKVAV)-containing laminin A1 chain peptides self assembly to form amyloid-like fibrils when cations/cells are introduced.