Characterization of pET29B(+)-phaC1J4

The objective of our project was to genetically engineer E. coli to synthesize PHB. We designed a pET29b(+)-phaC1J4 construct to allow our engineered E. coli to employ the beta-oxidation pathway to break down volatile fatty acids (VFAs) and use it to synthesize PHB. We have successfully ligated and transformed this part into E.coli. The figure below shows the digestion confirmation of the vector pET29B(+)-phaC1J4. We have also sequence verified this part.

To confirm the production of protein, we induced gene expression in E. coli BL21(DE3) transformed with pET29B(+)-phaC1J4 with IPTG and performed SDS-PAGE. Our results are shown in the picture below. The SDS-PAGE results indicated that our protein PhaJ4 was being expressed. It was difficult to distinguish PhaC1 amongst the other protein bands. However, because PhaJ4 is downstream of PhaC1, it is likely that PhaC1 was being expressed as well, but further SDS-PAGE runs must be performed for confirmation. Furthermore, our PHB synthesis experiment with E. coli containing pET29B(+)-phaC1J4 produced PHB as shown in Figure 3. The successful synthesis of PHB indicated the functionality of our exogenous proteins.

To confirm the ability of E. coli transformed with our pET29B(+)-phaC1J4 construct to produce PHB, we grew O/N cultures for 24 hours and induced them with 0.1 mM of IPTG while supplementing them with chemical media mimicking fermented synthetic poop supernatant (FSPS). Cells were then left to culture for 16 hours. To collect the resulting PHB, we performed sodium hypochlorite extraction on the cultures, which can be found on the Experiments page.

To confirm that the identity of the white powder made by the transformed E. coli was actually our desired product of PHB, we ran our results alongside a sample of pure industrial-grade PHB provided to use by PolyFerm by dissolving PHB in sulphuric acid, which can be detected via HPLC as crotonic acid. The analysis confirmed the identity of our PHB and its similarity to industrial-grade material because the presence of crotonic acid was indicated by the peaks in the spectrum below.

Detailed results of these experiments can be found on this part's Registry page.

Characterization of pET29B(+)-phaCBA

As part of our project, we also used a pathway that relied on acetyl-CoA, a metabolite from the breakdown of glucose and some VFAs. We designed a pET29B(+)-phaCBA construct so that the E. coli could convert acetyl-CoA to PHB. We have successfully ligated and transformed this part into E. coli. The figure below shows the digestion confirmation of the vector pET29B(+)-phaCBA. We have also sequence verified this part.

We then cultured these transformed cells following the same method described above to product and extract PHB.

Again, To confirm that the identity of the white powder made by the transformed E. coli was PHB, we ran our results alongside a sample of pure industrial-grade PHB provided to use by PolyFerm by dissolving it in sulfuric acid. This leads to the production of crotonic acid peaks after HPLC analysis. The analysis confirmed the identity of our PHB, as the presence of crotonic acid was indicated by the peaks.

Characterization of pSB1C3-Phasin-HlyA Tag

In order to secrete PHB, we engineered the fusion protein phasin-HlyA tag first into E. coli DH5α for plasmid propagation then into E. coli BL21(DE3) for IPTG-inducible protein expression. Figure 8 shows the successful digestion of the phasin-HlyA tag construct with EcoRI and SpeI ligated in the pSB1C3 backbone.

To confirm the functionality of the phasin-HlyA protein, a "super" plasmid that contains both PHB-producing and PHB-secreting genes was made. This plasmid was created by ligating the phasin-HlyA construct into pSB1C3-phaCAB. PhaCAB is a PHB-producing BioBrick from Imperial College, 2013). This plasmid was transformed into E. coli BL21(DE3). The successful ligation of the "super" plasmid was confirmed with NotI restriction digest, as shown in Figure 9.

For a secretion assay, cultures of E. coli BL21(DE3) containing the "super" plasmid were inoculated in LB media + chloramphenicol + 3 % glucose. pSB1C3-pPhaCAB without the phasin-HlyA construct in E. coli BL21(DE3) was used as a control, and the experiment was run in triplicates. The bacteria were induced with IPTG, then incubated at 37°C for either 24 hours or 48 hours. After that time, the samples were separated into intracellular and secreted PHB fractions. CaCl₂ was added to promote PHB agglomeration in the secreted fraction so that it could be separated from cells via differential centrifugation. The amount of PHB collected in each fraction was measured and those results are presented below (Table 1).

The addition of CaCl₂ to the secreted fractions results in a pellet being obtained that contains both PHB and CaCl₂. To account for this, the mass of secreted PHB was corrected for the mass of CaCl₂ added (0.05549 g). The corrected results are presented in Table 2.

The assay showed that cells with PHB-secreting genes secreted more PHB than control cells which did not contain these genes, but this difference only became apparent after 48 hours of incubation at 37°C. At 24 hours, there was no meaningful difference between the amounts of PHB in the media of both E. coli variants (0.12851 g for the control and 0.12417 for the positive PHB-secreting samples). At 48 h, control cells had only 0.08051 g of PHB in the media, but the PHB-secreting cells secreted 0.17251 g. This is a 114% increase in the amount of PHB secreted by cells with Phasin-HlyA Tag compared to the control. This indicates that our part, Phasin-HlyA Tag, improves secretion of PHB after an incubation time of at least 24 hours.

These results can also be found on this part's Registry page.

Methods for VFA quantification and characterization

As mentioned in our Notebook, determination of the total VFA concentration in the solution was an important step in the process – knowing how to quantify total VFAs in the solution helped to prove that the fermentation of human feces with naturally-occurring bacteria increases the VFA concentration, as well as it helped to prove VFA presence in both fermented and unfermented synthetic feces.

Titration is commonly employed by wastewater treatment plants to give a rapid estimate of the VFA concentration in sample solution. Toe measure VFA concentration, we performed Simple Titration procedures, which can be found on our Experiments page. The results in Table 3 indicate that this method tends to give a slight overestimate of the total concentration; however, it can be used for quick estimations and show changes in overall VFA concentration.

HPLC is another method commonly employed in laboratory setting for the VFA concentration determination. The advantage of the method is the fact that it provides the concentration of different volatile fatty acids in the solution.

Table 4 shows VFA concentrations in fermented synthetic feces supernatant as measured by the HPLC. Since the amount of VFAs in the supernatant was unknown, different dilution factors were tested to ensure the VFA concentrations are within the optimal HPLC detection range.

VFA fermentation results

We fermented synthetic feces to mimic the fermentation by natural gut flora that produce VFAs in the supernatant. In our process, these VFAs can later be used as a feedstock for our PHB-producing E. coli. First, we determined VFA production by leaving synthetic feces to ferment at 22 and 37°C. This experiment was performed before the Synthesis team had successfully created our PHB-producing parts, so we borrowed those of other teams available in the iGEM Registry (from Imperial College 2013 and Tokyo 2012). The first VFA fermentation experiment showed higher VFA production at 37°C than at 22°C (Figure 11). The control condition was E. coli BL21(DE3) transformed with pET29B(+) vector without any inserts (non-PHB producing), "Imperial condition" was PHB-producing E. coli transformed with a PHB-producing part from Imperial College 2013, and "Tokyo condition" was PHB-producing E. coli transformed with a PHB-producing part from Tokyo 2012. The reported results are an average of 3 titrations performed for each sample.

We performed a second VFA fermentation experiment confirmed higher VFA production at 37°C than at 22°C (Figure 12). In addition to a 3-day fermentation (denoted as D3), a 5-day fermentation was also introduced during this experiment (denoted as D5). Higher VFA concentrations were observed after 5 days of fermentation. Similarly to the first experiment, the reported results are an average of 3 titrations performed for each sample.

In subsequent experiments, the supernatant from fermented synthetic feces was collected after the VFA fermentation experiments, sterilized, and cultured with PHB-producing bacteria for 2 – 3 days at 37°C. PHB-producing bacteria cultured in the supernatant from synthetic feces fermented at 37°C had little to no growth, resulting in little to no PHB produced. Although the supernatant from synthetic feces fermented at 37°C had higher VFA concentration (Figure 11, Figure 12), we believe the low pH of these samples prohibited bacterial growth, resulting in no PHB production. Consequently, the synthetic feces supernatant fermented at 22°C was selected as preferred VFA fermentation temperature due to optimal pH for bacterial growth, although it had a lower VFA concentration.

Liquid-solid separation results

The first experiments we performed for solid-liquid separation of synthetic feces from supernatant were Gravity-Driven Sedimentation and Gravity-Driven Filtration; both protocols can be found on the Experiments page. Both methods rely on gravity to settle solids in a phase separate from that of the liquid. The results of these experiments are summarized in Table 5 below:

It is clear the gravity alone did not sufficiently separate solids from liquids; therefore, the Staged Filtration Experiment was conducted using 25 g of synthetic feces (Recipe 2). This method relied on using a series of filters with sequentially smaller pores to separate solids from liquid in multiple steps. The original sample contained 15 g of water, yet only 10% of it was recovered, meaning that a more advanced and power intensive technology has to be considered for this stage of the process.

Finally, we decided to investigate the efficiency of centrifugal-based extraction methods using the Centrifugation for Solid-Liquid Separation Experiment. When a 50 g undiluted sample of synthetic feces (Recipe 2) was tested, the mass of water recovered was 19.6 g, while the mass of initial water present in the sample was 30 g, meaning that there was 65% recovery of water. This result indicated that centrifugal based solid-liquid separation technology would be the best fit for our application.

PHB Extraction Results

Chemical Coagulation Results

We demonstrated in the lab that adding calcium chloride and then centrifuging increased the amount of PHB removed from a suspension in water, as CaCl₂ assists in PHB agglomeration. We carried out centrifugation at various speeds and then collected absorbance data for the samples to measure relative quantity of PHB in solution. The lower the absorbance reading, the higher the amount of PHB that was removed.

Details concerning the methods of this experiment can be found here. The absorbance readings and the standard deviations of the various conditions tested are shown in the figure below:

As expected, centrifugation at higher RPM resulted greater removal of PHB from solution. The addition of CaCl₂ increased PHB agglomeration, resulting in additional PHB extraction.

Electrocoagulation

We then looked at the possibility of using electrocoagulation to remove PHB from solution. We tested electrocoagulation on microscale PHB particles in suspension in water, nanoscale PHB particles in suspension in water, synthetic feces supernatant, and a 1:1 mixture of microscale PHB suspension and synthetic feces supernatant.

From our preliminary experiments with just PHB particles in a water suspension, we were able to demonstrate that PHB particles do settle out via electrocoagulation. After we stopped running electricity through our apparatuses, we were able to observe a layer of PHB at the bottom of the container. However, after a few hours a layer of brown powder settled on top of the white layer of PHB. We hypothesized that this was probably iron(III) hydroxide, formed from the excess iron ions released by the anode that did not bind to the PHB.

From our experiments with synthetic feces supernatant, we found that a layer of brown sludge settled at the bottom each time. Even with the 1:1 mixture there was no discernable layer of PHB within the sludge. We tried washing the sludge with dilute acid to remove the metal salts that might have been present in the sludge. However, we were still unable to separate PHB from the sludge.

The electrocoagulation experiments led us to conclude that while it is possible to settle out PHB using electrocoagulation, this method was not suitable for our media which contained a number of salts that interfered with the coagulation process and caused the formation of sludge.

High Performance Liquid Chromatography (HPLC) Analysis of PHB

Crotonic acid is the product of PHB digestion with sulphuric acid, and itcan be detected using HPLC. During the first HPLC run with our PHB samples extracted from cell culture, we didn’t observe crotonic acid peaks as expected. This was likely due to digesting our PHB in sulphuric acid for 3 hours, which was likely too long, and resulted in degradation of crotonic acid in solution.

For our second run of HPLC trials, we tested two different sulphuric acid digestion times (20 minutes and 30 minutes) and observed peaks for crotonic acid for samples obtained from our phaCBA part and confirmed that PHB was present in the our samples. We also ran HPLC trials on PHB produced using Imperial College's 2013 BioBrick and observed crotonic acid peaks as well.

Although we were successful at detecting PHB using HPLC, quantification of PHB using HPLC would require further development. In particular, an optimal digestion time needs to be selected where all of the PHB in the sample is converted to crotonic acid and crotonic acid does not start to degrade. Nonetheless, we attempted to quantify our PHB samples by assuming an 80% conversion of PHB to crotonic acid, a value reported by Karr et al. (1983). However, conversation factors in our HPLC runs ranged from 40% to 60% for the standard PHB samples, resulting in inaccurate dilution factor predictions, and we were not able to quantify our samples. Accurate dilutions are important to ensure the sample concentration is within the HPLC detection range.

For our third HPLC run, we digested PHB samples from our beta-oxidation BioBrick for 15 and 30 minutes and observed clear peaks for crotonic acid. Again, we based our dilution factors on 60% conversion of PHB to crotonic acid. More details about the HPLC results can be found on the Beta-Oxidation and Glycolysis part pages.

Tools from PHB

We used commerical PHB, the same type of bioplastic that was produced in the lab by genetically engineered bacteria, to make a 2.5 cm (length) by 1.1 cm (width) wrench (Figure 17). The wrench was made by melting PHB into a wrench-shaped mold and letting the plastic to cool.