PRK

Achievements:

1. Construction and digestion of DNA gel shows that the size of it was right
2. The SDS-PAGE of PRK showed that the expression of PRK in the expected protein size
3. PRK toxicity test proves that the function of it varies when cloned into different plasmid

We constructed PRK fragments(BBa_K2762007) from IDT DNA synthesis. After PCR amplification, PRK is then cloned into pSB1C3 and transformed into DH5α. SDS-PAGE ensured that the protein expression was as expected. The results are shown below:

Fig. 1: Confirmation of prk digestion.
Fig. 2: Confirmation of PRK expression in DH5˚α. The expected protein size is 37.7kDa.

We initially decided to test its function by HPLC to measure the amount of RuBP inside the cell. Our instructors pointed out some difficulties in HPLC measurement such as excessive noise signal in our sample.We, therefore, determined to test its function with a toxicity test. The product of PRK-RuBP cannot be metabolite by wild-type E. coli. The accumulation of RuBP depletes the sugar from the native pentose phosphate pathway. Lack of carbon source, the growth of that strain may be repressed. We incubate the PRK expressing strain and control stain that contains no plasmid in M9 medium and altered M9 medium with 0.4% xylose as its sole carbon source. In normal M9 medium, glucose will not be converted into RuBP. In altered M9 medium, xylose will go through the native pathway and be converted into RuBP. Growth arrest of PRK strain should be observed.

We tested PRK in different strains. We first cloned prk into pSB1C3 and transformed into BL21(DE3). After 12 hours, the strain without plasmid could grow up to 1.4 O.D.600 in altered M9 xylose medium. The strain that contains PRK can grow up to 0.75 O.D.600 in normal M9 medium either. In contrast, the PRK strain that grew in altered M9 xylose medium showed no growth at all. The result shows that PRK can suppress/inhibit the growth, which matches to our expectation.

Fig. 3 The result of PRK test in BL21(DE3). The PRK expressing strain is incubated in both normal M9 medium and altered M9 xylose medium to compare with the strain without plasmid. The PRK expressing strain grown in altered M9 xylose showed merely no growth, which proves the function of PRK.

Although the function of PRK has been confirmed, we would like to lower the expression of it to minimize the growth arrest. We thus cloned the part into pSB3K3, a low copy number plasmid to lower its protein expression. We then compare the growth under high and low copy number plasmid. We found out that pSB3K3 shows a little growth arrest comparing to the strain without plasmid. The growth of it exceed that of PRK expressed in pSB1C3. We can regulate the expression of PRK via high or low copy number plasmid to optimize the growth and carbon fixation efficiency of the bacteria.

Fig. 4 Compares the growth in M9 xylose medium of PR K expressing strain in high and low copy number plasmid. The low copy number plasmid, pSB3K3, shows a little bit of growth retard compare to non-PRK expressing strain. However, the toxicity is much less than high copy number expressing strain.

We also transformed pSB3K3-prk into W3110 strain. W3110 is reported to have higher pressure tolerance. The trend of the results is similar to that of the BL21(DE3) but there is no statistically significant between the experiment and the control group. We deduce that PRK can still function in W3110 since the trend matches our expectation. As pSB3K3 is a low copy number plasmid, the expression of the protein may be lower than that of high copy number plasmid. The pressure tolerance of W3110 strain may also lessen the toxicity influence by PRK.

Fig. 5 The result of PRK test in W3110

CA

Achievements:

1. Construct the ca and transform it into BL21(DE3)
2. Run the SDS-PAGE to confirm its expression
3. Measure the activity of CA enzyme

We cloned the DNA fragments into pSB1C3 plasmid after the gene is amplified with PCR. We transform the plasmid into DH5-alpha and BL21(DE3). Next, we confirm its protein expression with SDS-PAGE.

Fig. 6 Confirmation of ca digestion
Fig. 7 Confirmation of CA expression in BL21(DE3). The expected protein size is 27.9kDa.

We then ran the activity test of CA. In our bypass pathway, the function of CA is to convert proton and bicarbonate into water and carbon dioxide. CA activity was determined using the Wilbur-Anderson assay. Briefly, 9 mL ice-cold Tris−HCl (20 mM, pH8.3) buffer and 0.2 mL enzyme were mixed and transferred to a 20 mL sample bottle, with further incubation at 0 °C with stirring. Then, 6 mL of ice-cold CO2-saturated solution was added immediately into the sample bottle and the time course (in sec) of pH decrease from 8.3 to 6.3 was recorded. CA activity was calculated using a Wilbur–Anderson unit (WAU) per millilitre of sample. The definition for WAU is (T0-T)/(T0) in which T0 and T was the time required for the pH drop from 8.3 to 6.3, with and without CA, respectively. The enzyme activity of our CA is 21.8 unit/liter. To confirm the contribution of the CA to the whole pathway, we also ran the total solution which will be described in the following content.

Rubisco

Achievements:

1. Construction and digestion of DNA gel shows that the size of it was right
2. The SDS-PAGE of PRK showed that the expression of PRK in the expected size

We constructed PRK fragments from IDT DNA synthesis. After PCR amplification of the three subunits, rubisco is then cloned into pSB1C3 and transformed into DH5α. SDS-PAGE ensured that the protein expression was as expected. The results are shown below:

Fig. 8 Confirmation of rbcL digestion.
Fig. 9 Confirmation of rbcL expression in DH5˚α. The expected protein size is 52.37kDa.

After mining a lot of information from the publications, we found out a method to determine the activity of rubisco by thin-layer chromatographic has been reported. However, due to time concern, we are not capable of measuring the enzyme activity of Rubisco with this method. We finally confirm its function from the results of total solution test.

Rubisco

Achievements:

1. Develop an index to evaluate the carbon fixation ability of each constructed strain
2. Confirm the importance of Rubisco enzyme in bypass pathway
3. Check the growth and carbon fixation enhancement of CA enzyme
4. Compare the carbon fixation rate of W3110 and BL21(DE3) E. coli strains
5. Compare different CO₂ incubation environment

Overview

In the total solution experiment, we strive to measure the carbon fixation amount of each sample. After reading numerous publications, we found out that previous researches determine the efficiency of carbon fixation via measuring the decrease of carbon dioxide concentration in the closed system or measure the weight percentage of C14 radioisotope in the dry cell. However, due to biosafety constrain of our lab, we can barely use the radioisotope. Measuring the decrease of carbon dioxide concentration in the closed system is also impractical for us since we have too much test samples. A new method to measure multiple samples in the short period of time is developed by our team. We are able to evaluate the fixation efficiency of each sample with optical density O.D. 600 and xylose consumption. We have measure various construction to prove that the enzyme of our construction is necessary for carbon fixation.

The test samples below were incubated in an altered M9 medium which substitute glucose to xylose. 1/1000 of LB medium was added to support some rare elements. Since the concentration of LB medium is too low, it doesn’t contribute the carbon source of the bacteria.

We defined a new index, Xylose Utilization Index, to describe the potential of carbon fixation. We can compare this index of each strain to find out the strain that has highest capacity of carbon fixing.

To define the XUI index, we firstly made two assumptions:

1. O.D. 600 of the sample has linear relationship to dry cell weight (biomass). Optical density is frequently used as a means of describing the cell density in the broth. We measured the dry cell weight of samples in different O.D. value and discovered that it has linear relationship. We conclude that we can utilize O.D. value to estimate the dry cell weight. 1 0.D. of BL21(DE3) strain per litter yields the dry cell weight of 0.8 gram.

Fig. 10 shows the dry cell weight of BL21(DE3) incubated in altered M9 xylose medium. A linear relationship between O.D. and dry cell weight is observed.

2. The elemental formula of E. coli should be fixed or varies within a small range. Although there may exist slightly different in different growth condition, we assume that such error can be ignore during the following calculation.

After these two assumptions, the Xylose Consumption Index is designed to evaluate the carbon fixation ability of each strain. The definition of the index is xylose consumption over O.D. 600. O.D. 600 measurement can be viewed as the weight of carbon of the bacteria. The index shows the ratio of xylose consumption per biomass. For wild type E. coli, it only consumes xylose (the sole carbon source provided in our medium) as its carbon source. Although some native E. coli pathway may utilize CO₂ (such as lipid synthesis), the amount is too small to consider. As for engineered strain, carbon dioxide can be utilized as it’s carbon source. By producing same amount of carbon biomass, it requires less xylose. We can thus compare the XUI of each strain to determine the possible strain that fix carbon. The less the XUI in the sample, the more possibility that it fix carbon.

We use the Dinitrosalicylic Acid (DNS) reducing sugar assay to measure the xylose concentration in the medium. Under base solution, DNS will turn to brown color while reacting with reductive sugar in high temperature. In the specific temperature range, the color will have linear relationship with the reductive sugar concentration. We can thus measure the xylose concentration at O.D.540.

Fig. 11 Shows the calibration line of DNS assay kit.

Before measuring the XUI, we observe the growth curve of each strain. We found out that W3110(L5T7) constructed strain cannot grow in altered M9 solution. W3110(L5T7) is a newly constructed strain, we are not quite certain its characteristic. We eliminate this strain from the following experiment. BL21(DE3) and W3110 constructed strains show little growth after 24 hours.

Fig. 12 shows the growth of W3110(L5T7), BL21(DE3), W3110 incubated in normal incubator for 24 hours. The growth of W3110(L5T7) is not obvious while other strains shows growth after 24hours.

Total solution check: Function of Rubisco

We then utilized XUI to evaluate the function of each enzyme in the pathway. We first check the function of Rubisco in BL21(DE3) strain. Rubisco enzyme with promoter P_T7 (BBa_K2762011) was cloned into pSB1C3 and PRK with promoter P_LacI (BBa_K2762007) was cloned into pSB3K3. Both plasmids were then co-transformed into BL21(DE3). We measure the XUI of the strain and compare to the control that IPTG was not added and BL21(DE3) without plasmid. IPTG can induce the promoter P_T7 to produce the downstream enzyme. The growth of each strain is first examined. The IPTG induced strain showed growth retard. We assume the cause of growth retard is due to the pressure from overexpressing the protein rubisco. The control strain without IPTG induction produce less rubisco enzyme than the experiment and had less pressure. We then compare the XUI of each strain and discovered that control strain without IPTG induction produce less rubisco enzyme than the experiment. Without rubisco, the bypass pathway is not capable of using CO₂. We found out that the strain without Rubisco has higher XUI, symbolizing that rubisco is essential in carbon fixation pathway.

Fig. 13 Shows the growth and XUI measured in 5% CO₂ incubation of 12 hours respectively. Lower growth of the strain that contains. The XUI of the strain that contains both Rubisco and PRK shows statistically significant decrease compare to strain without both enzymes.

Total solution check: Function of CA

From the above results, we discovered that although Rubisco and Prk alone can enhance the utilization rate of carbon dioxide, the growth and utilization ability didn’t meet our expectations. The third important enzyme came into play: CA enzyme. We cloned Rubisco (BBa_K2762011) into pSB1C3 and cloned PRK with P_LacI promoter and CA with P_T7 promoter(BBa_K2762013) into pSB3K3. Two plasmids are then co-transformed into BL21(DE3). We measured the XUI of this strain and compare with the previous strain that only contains PRK and Rubisco. We found out that CA can raise the growth and lower the XUI. We infer that CA can enhance the intracellular CO₂ concentration and thus increase the carbon flux of the bypass pathway. The efficiency of the bypass pathway is thus been increased.

Fig. 14 Shows the growth and XUI comparison of each strains. All the tested strains are incubated in 5% CO₂ incubator for 12 hr. 0.1mM of IPTG was added to induce the protein expression. We can observe that growth speed of the construction has been increased with the CA, and the XUI of the strain that contains complete three enzymes was the lowest compared to the strain without plasmid or the strain that only contains PRK and Rubisco, stating that three enzymes are required to optimized the carbon fixing bypass pathway.

XUI Comparison between BL21(DE3) and W3110

We then compare the XUI value between BL21(DE3) and W3110 constructed strain. When we design our IDT sequence, we link the CA directly to the promoter P_LacI, so we could not transform CA construct into W3110 strain. We thus compare the XUI of strains that only contains Rubisco and PRK (BBa_K2762011) , (BBa_K2762007). We found out that both strain shows similar trend: the XUI will be lower with the expression of the constructed protein. The growth condition of both constructed strains is similar for the first 12 hours. We then compare the difference of XUI between two E. coli strain. We found out that both strain shows similar trend: the XUI will be lower with the expression of the constructed protein. HoweverW3110 has a higher XUI compared with BL21(DE3) in constructed strain as well as the strain without plasmid. We infer two reasons that cause the difference of XUI:

1. W3110 “wildtype” strain has more flexible metabolic network but consumes more xylose compare to lab strains such as BL21(DE3).
2. The constructed protein expression in W3110 may be less than BL21(DE3) lab strain. BL21(DE3) commonly used to express protein. We inferred that with more protein been expressed, the bypass pathway in BL21(DE3) will be more favored than the W3110 strain.

Fig. 15 Shows the growth and the XUI of BL21(DE3) and W3110 strains.

We finally concluded that the efficiency of the bypass pathway in BL21(DE3) is better than that in the W3110 strain.

Incubation under different CO₂ concentration

Finally, we compare the XUI under different CO₂ concentration. We incubated the bacteria in normal incubator without CO₂ input and the cell culture incubator that maintains 5% CO₂ concentration. We observed that the strain in 5% CO₂ incubator has lower the XUI. The supply of sufficient CO₂ can increase the efficiency of the bypass pathway and enhance the growth. We can concluded that our constructed pathway can be utilize carbon dioxide as one of its carbon source from this result.

Fig. 16 The comparison of the growth and the XUI of the BL21(DE3) that contains all three enzymes in normal incubator and 5% CO₂ incubator. The strain grown in CO₂ incubator shows better growth and lower XUI, which indicates that our strain can use CO₂ as a carbon source in the presence of high CO₂ level.

To find out how much and how efficient genetically engineered E. coli can fix carbon dioxide, we use the material balance concept to evaluate the heterotrophic CO₂ fixation process. Consider a system composed of a single component, the general material balance can be written as: $${\{Input\ to\ the\ system\}\ –\ \{Output\ to\ the\ system\}\ =\ \{Accumulation\ in\ the\ syste\}}$$ A system can be defined as an arbitrary portion of a process considered for analysis, in which in this case, is an engineered carbon capturing E. coli.

The engineered E. coli BL21(DE3) are cultured in M9 medium with formula adjusted so that xylose is the sole carbon source. The aforementioned M9 Medium contains 0.4% xylose and 1/1000 LB medium (the carbon consumed from LB medium can be ignored). By applying the law of conservation of mass, which states that mass may neither be created nor destroyed, the material balance for carbon in an engineered E. coli may simply be written as $${\{C_{CO_2}\ in\}\ +\ \{C_{xylose}\}\ -\ \{C_{CO_2}\ out\}\ -\ \{C_{waste}\}\ =\ \{C_{biomass}\}...(1)}$$ Considering the difficulties in measuring carbon in E. coli metabolic waste and that C_waste would be positive, the equation reduces to $${\{C_{CO_2}\ in\}\ -\ \{C_{CO2}\ out\}\ ≥\ \{C_{biomass}\}\ -\ \{C_{xylose}\}...(2)}$$ Let {C_CO2 net}= {C_CO2 in} - {C_CO2 out}, equation (2) further simplifies to $${\{C_{CO_2}\ net\}\ ≥\ \{C_{biomass}\}\ -\ \{C_{xylose}\}...(3)}$$ If C_waste is very small and negligible, we can obtain the net amount of carbon dioxide fixed over time. If, on the contrary, C_waste cannot be neglected, equation (3) allows us to estimate the minimum net amount of carbon dioxide fixed.

C_biomass can be calculate by multiplying O.D. 600 to DCW and mass percent of carbon in E. coli biomass. The O.D. 600 of engineered E. coli is measured after a 12-hour cultivation and the result obtained is 0.45O.D. . Yin Li et al. reported that dry cell weight (DCW) of E. coli is $${0.35g\over L ∙ 𝑂.𝐷. 600}$$ , determined by experiment. E. coli biomass contains 48% of carbon by mass. $${C_{biomass}\ =\ 0.45\ ×\ 0.35\ ×\ 48\%}$$ $${=\ 0.0756\ g/L}$$

On the other hand, C_xylose can be calculated by multiplying the amount of xylose consumed per unit volume of broth to the mass percent of carbon in xylose. Xylose consumption is calculated by using a DNS kit that measures the concentration of reducing sugar and the result obtained is 0.172324g of xylose consumed per litre of M9 medium. Carbon mass percentage of xylose is 40%. $${C_{xylose}\ =\ 0.172324\ ×\ 40\%\ =\ 0.0689296\ g/L}$$ By equation (3) $${C_{CO_2\ net}\ =\ 0.0756\ -\ 0.0689296}$$ $${=\ 0.0066704\ g/L}$$ Since the E. coli has been cultured for 12 hours, we can calculate the rate of carbon fixation by $${Rate\ of\ carbon\ fixation\ =\ {𝐶_{𝐶𝑂_2\ 𝑛𝑒𝑡}\over 12}}$$ $${=\ {0.0066704\over 12}}$$ $${=\ 0.5558\ {mg\over L ∙hr}}$$ To find out how much carbon in biomass comes from the carbon in CO₂ captured by the heterotrophic microbes, divide the net amount of carbon fixed by the mass percent of carbon in biomass. $${Ratio\ of\ carbon\ in\ CO_2\ fixed\ to\ carbon\ in\ biomass\ =\ {0.0066704\over 0.0756}}$$ $${=\ 8.82\%}$$

We construct both promoters with PCR, using primer as templates since the size of it is small. For the construction of P_asr, we cloned into the plasmid that contains both rbc and GFP. For the construction of P_gadA , we took the constructed part from 2016 Dundee iGEM team as a reference. Both parts were then cloned into pSB1C3 plasmid and transformed into BL21(DE3).

Pasr is reported to be induce in acidic condition. We think that in can be used to report the abnormal acidity of the medium. We thus determine to measure the fluorescent intensity in a short period of time. We first incubated the bacteria to log phase (within 2 hour) with LB medium. We then centrifuged the broth and suspended the pellet with pH modified M9 medium (the pH value is modified with 1M HCl). We then took the sample and incubate in the 96 well and measure its fluorescent intensity for every 3 minutes. We found out that the promoter Pasr will be induced at the pH value below four within 30 minutes. The different fluorescent intensity can be observed within 30 minutes. The fluorescent had the peak at pH value of 4.25.

Fig. 17 The data shows the fluorescent intensity (absorbance: 485 nm, excitation: 535 nm) expressed by Pasr in different pH.

Based on the data has shown above, we could conclude that Pasr is an acidic promoter as it has a high expression of fluorescent at pH 4.25 and pH 5. The results show that Pasr constructed pH sensing system can be used as an alert. When the medium turns acidic, fluorescent can be easily observed. We believe that this system can also be applied to various bio-detection system.

Fluorescent intensity measurement of P_gadA

P_gadA was previously reported to be induced under neutral and mild acidic environment. We measure the fluorescent intensity for 14 hours. We pre-cultured the strain and incubate the strain with pH modified M9 medium (the pH value is modified with 1M HCl). The induction of P_gadA is observed under neutral and mild expression.

Fig. 18 The data shows the fluorescent intensity (absorbance: 485 nm, excitation: 535 nm) expressed by PGadA in different pH.

We found out that the fluorescent intensity of P_gadA is much lower than the P_asr and would like to improve the sensitivity of this biobrick. We thus add a riboJ sequence at the downstream of PGadA. For more information, please check the Improvement page.

References

1. Gong, F., Liu, G., Zhai, X., Zhou, J., Cai, Z., & Li, Y. (2015). Quantitative analysis of an engineered CO2-fixing Escherichia Coli reveals great potential of heterotrophic CO2 fixation. Biotechnology for Biofuels,8(1). doi:10.1186/s13068-015-0268-1
2. Stockar, U. V., & Liu, J. (1999). Does microbial life always feed on negative entropy? Thermodynamic analysis of microbial growth. Biochimica Et Biophysica Acta (BBA) - Bioenergetics,1412(3), 191-211. doi:10.1016/s0005-2728(99)00065-1
3. 2016 Dundee iGEM team
4. Chakrabarti, S., Bhattacharya, S., & Bhattacharya, S. K. (2002). A nonradioactive assay method for determination of enzymatic activity of d-ribulose-1,5-bisphosphate carboxylase/oxygenase (Rubisco). Journal of Biochemical and Biophysical Methods,52(3), 179-187. doi:10.1016/s0165-022x(02)00072-6