Proof of Concept

The first step of our project is to show the presence of starch in brewers’ spent grain. To do so, we first infused 3g of brewers’ spent grain in a solution containing 30mL of PBS buffer and 10 mL of distilled water. We set up two conditions. One tube was put at 60°C degrees overnight in a water bath, in order to mimic the temperature at which it would come out at a brewery. The other tube was put overnight at 37°C in an incubator since it's the optimal temperature for E. coli growth.

We did a negative control with only ultra pure water and a positive control with a starch solution with a concentration of 4g/mL. We used a lugol solution to reveal the presence of starch. Lugol is normally a yellow/orange color, and turns into a very deep blue in the presence of starch. For each of our conditions, we took 1mL of the solution, and added 30µL of lugol.

The lugol immediately changed colors in contact in both tubes with spent grain and the one with pure starch. As we can see in Figure 1, the lugol is darker in the tube with spent grain at 60ºC than the one at 37ºC, which means that treatment of brewers’ spent grain in PBS at 60°C degree is more efficient to extract a large amount of starch than the one at 37°C.

In order to obtain quantitative results of the quantity of starch we could obtain from the brewers spent grain, we decided to use a spectrophotometer. To do so, we created a calibration curve with various concentrations of starch colored with lugol. We used the spectrophotometer to calculate their absorbance at 530 nm.

Once we got our calibration curve, we calculated the concentrations of our two samples: spent grain at 60ºC and spent grain at 37ºC. Their concentrations were out of range, so we diluted the 37ºC sample by 4, and the 60ºC samples by 8. Here is the curve we obtained, with the 37ºC sample in red, and the 60ºC sample in green (see Figure 2).

Therefore, this confirmed our qualitative results that the spent grain at 60ºC releases more starch than the one at 37ºC. Indeed, when taking into account the dilution factor, the sample at 37ºC has a concentration of about 0.59g/L, and the one at 60ºC is at about 1.01 g/L.

Our next step was trying to determine the best conditions for E. coli to be able to grow with the brewer's spent grain.

In order to test the conditions under which E. coli would be able to grow, we decided to do multiple tests. First we wanted to know how contaminated the brewers' spent grain was. Secondly, we wanted to see if the brewer's spent grain was a good enough source of carbon for E. coli to grow with it alone.

a. Testing the contamination of brewer's spent grain

To do this, we decided to put spent grain in LB medium without E. coli to see what would grow.

We had three Petri dishes:

• autoclaved brewer's spent grain with LB
• non-sterile brewer's spent grain with LB
• PBS that had been infused with brewer's spent grain overnight

After 36 hours at 37ºC, we looked at the results:

From these results, we can see that the non-sterile brewer's spent grain and even the autoclaved brewer's spent grain caused contamination in our plate (Figure 3 and 5). However, in the plate with PBS infused with the brewer's spent grain, we see that it is not the case. Furthermore, we saw in the first step of our proof of concept that the PBS infused with brewer's spent grain contained good concentrations of starch that our E. coli could use to grow.

b. Testing E. coli's ability to grow with spent grain as a source of carbon

In order to see if brewer's spent grain would be a sufficient source of carbon for E. coli to grow. To do so, we created a M9 minimal media (see Experiments). Before pouring our Petri dishes, we added different sources of carbon. We had two Petri dishes with PBS with starch brewers' spent grain and two others with glucose.

As we can see in Figure 6 and 7, E. coli was able to grow on the minimal media with both sources of carbon, which means that E. coli would be able to develop using only our spent grain.

We assessed the ability of BL21 E.coli transformed with the AmyH (AmyH BL21) gene to degrade starch by performing a lugol staining assay on solid cultures. As this amylase’s activity depends on NaCl concentrations, we tested multiple salt concentrations (i.e. 0%, 5%). We cultured the bacteria as well-defined separated colonies on solid media containing 0.2 g/L of starch. We defined the degradation halos as the total unstained surface minus the surface of the colony: S(unstained) - S(colony) (see Figure 8). We measured these surfaces using ImageJ, an open source software developed by Wayne Rasband.

Our results show an increased starch degradation activity in E. coli BL21 transformed with amyH as the mean halo sizes of the colonies are higher in transformed colonies (Figure 9 and Figure 10). Indeed, on Figure 9, with 0% of salt, we have a mean halo surface area of 0.7cm² for NT E. coli BL21 and of 1.28 cm² for AmyH E. coli BL21. On figure 10, with 5% of salt, NT E. coli BL21 have a mean halo surface area of 0.05cm², and AmyH E. coli BL21 has a mean halo surface area of 1.25cm². This appears to prove that our transformed E. coli Bl21 has a better capacity of degrading the starch in the media.

We cannot conclude on the NaCl concentration effect on the amylase activity as the maximum mean halo sizes are similar in both NaCl concentration conditions. However we observe that with a 5% NaCl concentration, halo sizes of non-transformed E. coli BL21 are close to zero while halo sizes of transformed BL21 reach the same value as, in the 0%NaCl condition.

Now that we have shown that our transformed E. coli should be able to use the starch present in the brewer's spent grain, we want to prove that our E. coli has the ability to produce lactate.

To prove that our E. coli transformed with LDH overproduced D-lactate, we decided to perform the D-Lactate Assay Kit from Sigma Aldrich (see Experiments). We did a calibration curve using the standard solutions from the kit. In order to have greater accuracy, we created two sets of standard solutions (Figure 11 ) and did a mean standard curve to find the concentration of samples.

From these two curves (Figure 11), we created a mean calibration curve. Fortunately, our two calibration curves were quite similar. We could then test our samples. For the test, we had two separate LDH sequences to test: LDH1 and our part LDH2 (BBa_K4187017). Moreover, for the assay kit, there were two enzymes: enzyme A and enzyme B. We could either test with both enzymes, or with just enzyme B. Enzyme A is only used for samples with endogenous enzyme activity, which is the case of our transformed E. coli. In both cases, it is recommended to do a control using only enzyme B, which we did.

As seen on Figure 12, LDH1 and LDH 2 both have significantly higher concentrations when tested with both enzymes, meaning that LDH truly has an endogenous activity. Moreover, E. coli transformed with LDH2, meaning our sequence, achieved higher concentrations of LDH compared to E. coli transformed with LDH1, which is quite positive for us.

In order to complete these results, we plan on doing more lactate assays once our bacteria will be transformed with the polymerization plasmid. Indeed, we will grow our culture of transformed E. coli in a minimal media with starch to see if we obtain similar results regarding the production of lactate.

Finally, we wanted to test the production of electricity by Shewanella oneidensis in presence of lactate, and with varying conditions.

We wanted to evaluate the effect of different conditions on Shewanella oneidensis's ability to produce electric currents to determine the best ones for our prototype. In total, we tested 5 conditions all at once in our preliminary prototypes (see Figure 13). Firstly, it's important to note that this experiment was done after a successful transformation of Shewanella oneidensis to overexpress mtrC (composite part BBa_K4187026). For clarity purposes, wild type Shewanella oneidensis will be called NT, and the Shewanella oneidensis transformed with mtrC will be named mtrC.

The first one was simply NT. The second was NT supplemented with 10mM of lactate. The third one mtrC with 10mM of lactate. The fourth one was mtrC with 10mM of lactate and riboflavin (also known as vitamin B2), which is an electron acceptor known to increase the electricity production of Shewanella oneidensis. Finally, the fifth one was mtrC with 10 mM of lactate, with a culture of E. coli producing Phenazine-1-carboxylic acid, which is also supposed to increase electricity production.

All conditions plateaued at their maximum tension variation 20 to 30 minutes after the beginning of the experience, as can be seen on Figure 14. The maximum tension variation of 28mV was observed with Shewanella oneidensis transformed with the mtrC gene and supplemented with 37nM riboflavin. This shows us that riboflavin might be the best way for improving electron transfer as it reaches a tension of over 10 mV above all the other conditions.

No significant difference was observed between Shewanella oneidensis transformed with mtrC and non-transformed Shewanella oneidensis which both plateau at around 16mV. Shewanella oneidensis transformed with p150 plateaued at around 5mV. Non lactate-supplemented wild type shewanella oneidensis did not show any consistent tension plateau and showed a max tension of 3mV. After 9 hours, all MFC returned to their baseline tension.

The results from the co-culture of E.coli secreting Phenazine-1-carboxylic acid and Shewanella oneidensis were not what we were expecting, especially compared to Shewanella oneidensis with Riboflavin. Indeed, we expected to see better results coming from the Phenazine-1-carboxylic acid. However, the test with Phenazine-1-carboxylic acid involved a co-culture of Shewanella oneidensis and E. coli, and we suspect that this is why we obtained weak results. We believe that E. coli may have overgrew Shewanella oneidensis. Therefore, in the future, we plan to re-do this experiment by purifying the Phenazine-1-carboxylic acid from the E. coli culture to avoid this problem.

Later on, we plan on continuing to do our proof of concept by showing the functionality of each of our parts. For the moment, we are missing EL222, PLAase, PHAC1 and PCT.

To show PLAase, we will first need to activate its transcription by activating the pBling receptor with EL222. Then, we will use the Nile Red protocol (see Experiments) to show the degradation of PLA over time. To do so, we will begin with a certain quantity of PLA that we will show thanks to the Nile Red coloration. After a period of incubation, we will see that the quantity of PLA has depleted.

On the other hand, to show that PCT and PHAC1 work, we will also use the Nile Red protocol, but this time to show its production. This will work by growing a culture of transformed E. coli bl21 with PHAC1 - PCT ligation, letting it incubate for 24h to 48h. Afterwards, we will color the Nile Red to show the production of PLA.

For EL222, we would create a construct in which a reporter gene, such as RFP, is expressed under the pBlind receptor. That way, if EL222 is expressed, we would know since it would activate the pBlind receptor, which would lead to the production of the reporter gene.

Moreover, we want to re-do most of our experiments in order to be able to do statistical tests on our results, to make sure that they are relevant.

Finally, we plan on assembling our two plasmids, Depolymerization and Polymerization, transforming E. coli BL21 with each of them, and performing the same tests to make sure everything works together.