Theory

In this section we will review the theory behind our biological work.

Phototaxis

Bacterial flagellar motility 101

Bacteria have evolved many modes of propulsion for the microscale environments they inhabit. At these scales materials behave very differently than at the macro-scale, and of particular interest to us is the way liquids seem more viscous. One of the many ways bacteria move around in liquids, is by means of flagella. A single flagellum is a thin filament around 0.1um thick, that extends many cell lengths out from the cell. It consist mainly of flaggelin subunits that assemble into a helical structure forming a long hollow cylindrical filament. In E. coli the mean number of flagella per cell is 4, but there is a wide variance between strains, and even between individual cells of each strain. The environment around the cell also has a large influence on how many flagellae are pressent, or if they are pressent at all, as we have learned! Flagella rotate to generate force that allows bacterial cells to swim through fluids in characteristic patterns, more on which later, in search of better conditions for proliferation or survival. Normally flagellated strains of E. coli can achieve speeds up to 20um/sec, and considdering a cell length of only 1-2um this is an impressive feat indeed.

A flagellum is anchored to the cell body by a large, wheel-like protein complex spanning both inner and outer membranes, through which the subunits are secreted to the tip of the flagellar tube, thus elongating the filament. (This mechanism is slightly different in archaea, where the filament is assembled from the base of the flagellum) The membrane anchor also functions as a rotary engine, driven by the proton motive force in much the same way the ATP-syntase is turned to create ATP from ADP. In fact the flagellar motor shares a lot of structural homology with the ATP-syntase, suggesting a common evolutionary ancestor. The flagellar motor rotates at up to 1000hz and can turn either clockwise or counter-clockwise, both resulting in a distinct movement pattern for the cell.

Spinning in the counter-clockwise direction , the flagella will twist into a bundle in the shape of a corkscrew, and create a linear driving force, propelling the cell in a straight line through the liquid. This form of movement is termed run. Spun in the clockwise direction one might then expect the cell to reverse, but this is not the case. Instead the flagellar bundle will unwind and each flagellum will flail wildly, creating chaotic movement. This movement reorients the cell randomly and is termed tumbling. A cell will typically run for ?? seconds at a time, then change it’s orientation by tumbling for ?? seconds, and then run again. The direction of flagella rotation is controlled by binding of a cytosolic protein CheY, more on which later.

Different taxis pathways that steer cells towards favorable conditions and away from danger work by regulating the frequency of tumbling events. We can take an example where a cell is getting close to a toxin it can sense and react to. As it gets closer to the source of the toxin, intracellular pathways will increase the frequency of tumbling events, in effect preventing the cell from rushing into certain doom, and since the frequency of tumling events will decrease if the cell is going in a direction away from the toxin, it will ”encourage” the cell to continue in that direction. In the case of an attractant such as an increase in nutrient concentration, the pattern will be oppisite, so that the cell is encouraged to continue towards the source of the attractant. This form of movement, combining tumbling and running, with regulation of the tumbling frequency is termed a biased random walk.

To understand how this can work we need a simplified understanding of the chemotaxis pathway at a molecular level. Chemotactic receptors can both increase and decrease tumbling frequencies to generate biased random walk behaviour. Increased tumbling is achieved through a phosphorylation cascade beginning with the binding of a repelant to a transmembrane receptor . The receptor is linked to two proteins CheW and CheA. CheA is a histidine-kinase that will autophosphorylate when the repelant binds. The phosphoryl group is then transfered to CheY activating the protein. The flagellar has high affinity for CheY-p, and binding reverses the mode of movement from run to tumbling. CheY-P is continuosly dephosphorylated back to CheY by CheZ which is present in the cytosol. A receptor sensing an attractant might instead switch from a default active CheA state to an inactive state when it’s ligand is bound, thus decreasing CheY phosphorylation.

Photosensor

In our system we want to be able to control the amount of flow in the channel through a remote signal. The signal we have chosen is light since we want to avoid altering the chemical composition of the fluid running through the channel. Having looked at previous iGEM work on light sensitive systems which have all been focused on transcriptional regulation, we realised that we would need a different approach for the fast response times our system requires. We have therefore focused our work on proteorhodopsins that integrate into the chemotaxis pathway, giving us very fast response to light stimulation.

Our construct centers around a synthetic protein created by Spudich et. al. It consists of an archaeal proteorhodopsin SRII (Sensory Rhodopsin II) and it’s transducer protein htrII from Natronomonas pharaonis coupled to a tar domain from a transmembrane receptor from Salmonella enterica. The tar domain is the part of the receptor that couples with CheA and CheW, and although it is taken from a different species, it has been shown to work in E. coli as well. In the construct we are working with light acts as an attractant, reducing the tumbling rate upon illumination. This might help us to control our pumping power, by decreasing the fraction of bacteria tumbling in the channel by increasing light stimulus, thus promoting linear drive. The photosensor should be most active in light with a wavelength of about 500nm, according to the original article.

Note that although the bacteria will be stationary in our system, since they are glued to the inner surface of the flowchannel, our construct in reality confers phototactic ability to E. coli.

Retinal Generator

Retinal requirements of light-sensing proteins

Proteorhodopsins and other related light-sensing proteins such as the rhodopsins found in human retinal pigment often require a supply of retinal to function. In fact retinal is the molecule responsible for the initial activation, as it undergoes photoizomerization after being struck by a photon. It is this change in conformation of the retinal molecule that is relayed through the entire rhodopsin-transducer complex to activate/deactivate the CheW/A complex in the cytosol. Thus either an external supply of retinal or an internal supply of genes coding for enzymes in the retinal biosynthesis pathway are required, if we wish to se phototactic behaviour in our cells.

Many plants and microbes have complete retinal biosynthesis pathways integrated into their genomes, to help drive their rhodopsins. In these organisms rhodopsins play an essential role, not only in photosensation but also directly in energy production, in fact in some organisms rhodopsins are used to create proton motive force directly by pumping protons out into the extracellular space using light energy to drive the process. Humans and other animals on the other hand often only have enzymes coding for the final steps of the pathway, more on which later. They rely on a supply of retinal precursors or Vitamin A (a group consisting of retinal and it's metabolites) in their diet. This is why Vitamin A defficiency causes night-blindness as an early symptom in humans.

Retinal biosynthesis

<p style="text-align: justify;">

In our construct, part of the synthesis is carried out by the BioBrick K274210, created by the Cambridge team in 2009. It consists of four genes crtE, crtB, crtI and crtY from P. ananatis that together make up the pathway that converts farnesyl pyrophosphate to beta-carotene, which is a precursor for retinal.

• crtE encodes the protein geranyl-geranyl pyrophosphate synthase that converts farnesyl pyrophosphate to geranyl-geranyl pyrophosphate by elongating it by one unit of isopentenyl.
• crtB encodes the protein phytoene synthase and synthesizes phytoene by putting together two molecules of geranyl-geranyl pyrophosphate whilst cutting off 2 molecules of pyrophosphate.
• crtI encodes the protein phytoene dehydrogenase and converts phytoene to lycopene by converting the trans bond to a cis bond and adding more cis double bonds.
• crtY encodes the protein lycopene B-cyclase and converts lycopene to beta-carotene.

The pathway (including the step that generates retinal) is summed up below:

To introduce the final step from beta-carotene to retinal, we use the gene ninaB from D. melanogaster. This gene encodes the protein beta-carotene 15,15’-monooxygenase, which cleaves beta-carotene to produce two molecules of retinal under the consumption of oxygen.
We have inserted the part 343006 into a different plasmid from the K274210 part since both parts are very long, so a plasmid containing both wouldn't have been viable.

Characterizing the retinal BioBrick

We characterized our BioBrick using two methods: Photospectrometry and high performance liquid chromatography (HPLC). While characterizing our own construct, we simultaneously characterized the Cambridge part, K274210 seeing as we wish to find out if the two parts work in concert.
Seeing as both beta-carotene and retinal have unique and characteristic spectres when studied by UV-vis spectrometry, this is a good place to start. The spectra can be obtained by harvesting beta-carotene- and retinal-producing cells, resuspending them in acetone and lysing them, which we chose to do by sonication. Afterwards, the cell debris can be pelleted and the supernatant can be examinated.
The acetone suspension of beta-carotene and retinal can then be subjected to photospectrometry, and the obtained values and spectra can be compared to those of pure beta-carotene or retinal in known concentrations. This method provides both a qualitative answer to whether or not the desired compound is present and a qualitative indication of the concentration in the cells.
This is the method by which the Cambridge team characterized their beta-carotene-generating BioBrick in 2009.
Apart from photospectrometry, the resuspension of beta-carotene or retinal in acetone can be subjected to analysis by HPLC. HPLC can be used to separate retinal from beta-carotene, to get a better indication of whether or not our retinal-generating part actually produces retinal from beta-carotene. In the experiments we used these protocols
See our results here.

BioBrick design

Retinal coding biobrick: K343001
Retinal composite part: K343002

Further use of the retinal BioBrick

Role in light-based signal transduction

Since retinal plays such an essential role in photosensing in both eukaryotes as well as bacteria and archaea, all work with rhodopsins and proteorhodopsins will need a retinal supply to function. This supply might come from the external environment, but it is an appealing thought that we might be able to supply the retinal from an internal source. Our project centers around phototaxis, but other constructs combining photorhodopsins with other membrane associated tyrosine kinases may also be imagined, opening vast posibilities for regulation of phopsphorylation cascades using light as input. In such systems, retinal biosynthesis could play a very valuable role.

Hyperflagellation

Background

The flagella regulon in E. coli is composed of at least 50 genes organized in no less than 14 operons that all contribute to the synthesis and operation of flagella. The operons are synthesized in a three-level transcriptional cascade where the FlhDC operon is the master regulator at the top of the cascade. The flagella regulon is tightly controlled by nutritional and environmental conditions, E. coli starved of amino acids showed temporarily decrease of the flagella regulon transcripts which are needed for the synthesis and operation of the flagellum.(1) The synthesis and assembly of flagella are regulated by the transcriptional cascade composed of three levels of gene products (class I, -II and –III). Class I genes consist of a single operon encoding the proteins FlhD and FlhC that form a multimeric (FlhD4C2) transcriptional activation complex. This ‘master regulator’ stimulates transcription by binding upstream of Class II promoters. Class II genes encode proteins that assemble to form the basal body and hook of the flagellum, as well as the fliA gene that encodes the alternative σ factor σ28, also called σF. σ28 binds to RNA polymerase (RNAP) core enzyme and directs it to Class III promoters. Class III genes encode the rest of the structural genes of the flagellum, including fliC encoding flagellin, as well as the chemotaxis apparatus. (1)
It has been shown that overexpression of the FlhDC operon restores motility in mutants that have been made immotile (2). Also, overexpression of FlhDC in the E. coli K12 strain MG1655 made the cells hypermotile.(3)

Our system
To maximize the microflow system's effectivity, we want to increase the force each single bacterium can generate. Since the main motor for the flow is the flagellum we will have to modify this factor for increasing the force. Flagella in E. coli rotate at a maximum speed around 6000 rpm, which can not easily be exceeded. So if we wanted to increse the generated force we would have to opt for more flagella on the surface of our bacteria, instead of faster flagella. This is called hyperflagellation, which is a process that is not all too well studied in normally-flagellated E. coli, so we will have to test it.
The way we are hoping to achieve the hyperflagellation is by upregulating the FlhDC operon. Since FlhDC is sitting on top of the regulating cascade we want to overexpress it, so that we will get an increase in flagellar count.
The way we are going to achieve that is by isolating the operon from an E. coli and inserting the coding sequence into a BioBrick where we can control both the ribosome binding site and the type of promoter. We are going to use a constitutive promoter, so that flagella will be constantly expressed. The effects of this on other areas of the organism like the cell cycle are not entirely clear, but we will first try to overxpress the operon and then analyse the effect on the cells behavior.

Biobrick design

FlhD,C mutated coding sequence: K343000
FlhDC composite part: K343004

And now for all of the truly brainy stuff!