Results - Blue Light

We decided to develop a photo-inducible genetic circuit that triggers the production of Ethylene in presence of blue light (470 nm), and stops at dark.

We thought to use blue light as our inducer because it would fit perfectly to our very own B. fruity vending machine, being the easiest way to control a genetic device in a totally automated scaffold. All parts have been transformed and characterized in E. coli (strain NEB10b).

The device

We aimed at first to get ethylene production at blue light (470 nm) and the off state at dark, so we went for the design of a blue light dependent device that includes an inverter cassette.

BBa_K1065310

We engineered in E. coli a blue-light sensor composed by:

• Anderson promoter BBa_J23100;
• the blue light receptor YF1, which consist of YtvA from B. subtilis fused to a kinase domain (fixL) from B. japonicum ("Möglich A., J Mol Biol. 2009, 385(5): 1433–1444)(Ohlendorf R., J Mol Biol. 2012, 414: 534-542);
• its response regulator, FixJ;
• a downstream promoter PfixK2, which is turned off by phosphorylated FixJ;
• an inverter cassette composed of cI and Plambda;
• a reporter (chromoprotein amilCP), which was later substituted by EFE (our ethylene forming enzyme).

To assemble this device we used the following parts from the registry:

• Bba_J23100 from Berkeley 2006 iGEM team;
• Bba_K592016 from Uppsala 2011 iGEM team;
• BBa_K592020 from Uppsala 2011 iGEM team.

We characterized this circuit along with the version without the inverter cassette (activated at dark and inhibited by blue light). Thus we also created the part:

BBa_K1065302

If you are interested in all the molecular details of these circuits, please check our datapage

Different sources of blue light induces AmilCP production in the "inverted circuit"

We first assembled the “inverted circuit” with a blue chromo-protein (AmilCP) downstream instead of EFE to obtain easy-to-watch and clear characterization results.
At first we compared the induction power of several light sources:

• 1 LED blue light;
• 1 blue light bulb;
• 1 white light bulb.

Fig. 1: Different light sources induction power. We had massive production of amilCP at LED light (4) and white light (3), instead exposure to the blue light bulb (2) induced a little less but still way more than the dark control (1): probably white light worked as well because it included the right wavelength (470), instead the blue bulb wavelength range is unknown, so we could assume that the right WL was not as much intense as in the other sources.

Then we decided to use only blue LED and normal light as inducers in further tests.

Fig. 2: Light successfully induced AmilCP production. We started the characterization test splitting the cultures (OD = 0.7) into 3 samples of 5 ml under 3 different conditions: Blue LED induced culture (1), Normal light induced culture (2), Dark control (3). We induced O/N at 37° and the next day we could admire a very conspicuous difference in color of liquid cultures and pellets .

Moreover, AmilCP has an absorbance peak at 588 nm so we measured the absorbance peak at the UV-VIS spectrometer (PerkinElmer lambda 25)in order to have concrete data. We sonicated samples for 10 seconds and resuspended the pellets in 2 ml of PBS.

Fig. 3: by measuring the absorbance of the three samples (588 nm), we observed a substantial difference between the dark control and the other two samples; this is a quantitative proof that the production of AMILCP at dark was strongly repressed. The graph also proves that white light induction is a little less efficient.

We carried out several tests in order to demonstrate the reproducibility of the behavior, although sometimes we observed amilCP production even in the dark control, so we could assume that the circuit doesn’t act like a perfectly controlled switch. We can make a speculation on the cause: probably the inverter cassette presence is responsible of this flawed behavior; Plambda is actually a strong promoter but CI transcription is at the end of a pretty long cascade that is likely to produce low CI; this means that there isn’t enough inhibitor to block Plambda activity. In order to confirm our theory we tested also the circuit without the inverter. This device, on the contrary, is designed to get switched on at dark and to be inhibited by blue light though. We extracted the part BBa_k952003, the circuit with the reporter AMILGFP (yellow fluorescent protein).

BBa_K952003

The part extracted from the registry missed a RBS sequence, resulting in a nonfunctional part. We decided to improve this part by inserting the missing RBS with a mutagenesis PCR. The mutagenesis was successfull.

BBa_K1065305

In order to have it tested and characterized, we also added the pLac promoter ahead as already shown (BBa_K1065302).

Better defined switch observed in the circuit wothout inverter

The test involved the induction of both, the improved circuit and the original part, in order to demonstrate the actual enhancement of the device. So we compared samples depending on two factors: induction/non induction & RBS/no RBS. We also took some quantitative measurements with a Cary Eclipse Varian fluorimeter considering that amilGFP is a fluorescent protein that emits at 512 nm and absorbs at 503 nm. For measurements we resuspended sonicated samples' pellets in 2 ml of PBS.

Fig. 4: Slight yellow shade appears only in the induced sample with RBS: After the culture with Bba_K1065302 reached OD= 0.7 we split it into 2 samples of 5ml at 37°: blue light exposed control (2) and induced sample at dark (1).We also made 2 sample (3 and 4) at the same conditions from a culture transformed with the original part missing the RBS, in order to compare the original part to the improved one. From both the image and the plot we can confirm that our part with RBS is undeniably improved and works as expected.

Summary

We achieved a successful characterization of both circuit with or without the inverter, noticing a substantial difference between controls and induced samples; Nevertheless we couldn’t appreciate this for every test performed on Bba_k1065310. Comparing the behavior of the 2 devices, we can notice that, under the conditions that we used, the one without the inverter definitely shows a sharper switch.
Supposed causes:

• as already explained probably the genetic cascade in BBa_K1065310 that brings the activation of cI isn’t strong enough, resulting in a limited production of inverter thus in the subsequent production of AmilCP;
• furthermore, we believe that the absence of a terminator after pFixK2 could cause the transcription of several different-sized segments, sometimes including the inverter cassette.
• a remaining element probably involved in the different behaviors is the different upstream promoters used.

We believe that more improvements need to be provided to the system in order to get it perfected and satisfying:

• add a terminator after pFixK2;
• substitute J23100 with pLac promoter;
• improve the transcription of CI;

To this day we are in the process of improving it and hopefully have some great enhancements for the championships!!!

However we were able to produce ethylene by placing EFE (BBa_K1065000) downstream the circuits. Please go ahead to our ethylene measurements section for these data.