The designed target gene oph , encodes for the enzyme organophosphate hydrolase (OPH), which degrades paraoxon into dimethyl phosphate (DMP) and p-nitrophenol (pNP) (Fig.1). Our oph gene is inserted in an enzyme plasmid in our pNP sensor cell, yet in order to monitor the level of paraoxon degradation by OPH in our sensor cell, we utilized the ability of pNP binding onto pNPmut1-1 to make a biosensor called pNP sensor. We referred our pNP sensor design directly to a research paper on organophosphate hydrolysis (Jha, Ramesh K., et al.). In our sensor plasmid, we included a dual-directional pobA/R promoter(BBa_K4271005), pNP RBS(BBa_K4271006), GFP sequence(BBa_I746916), pobR operator(BBa_K4271007), pNPmut 1-1 sequence for pNP binding(BBa_K4271004), and two double terminators of RrrnB1 terminator and T7 terminator(BBa_B0015) (Fig.2).

Once pNP binds to pNPmut1-1, the protein complex would act as an activator to the pobR operator, enhancing the ability of RNA polymerase to bind to the pobR promoter and initiate GFP transcription and translation (Fig.3). Therefore, as the level of pNP increases, more GFP will be generated to produce strong green fluorescence.

Our gene parts are synthesized by Twist Bioscience (ABreal Biotech Co., Taiwan), whose genetic synthesis is based directly on the sensor plasmid design we acquired from the paper(Jha, Ramesh K., et al.). Gene fragments that were synthesized were pNPmut1-1, dual-directional pobA//R promoter (including pobR operator and RBS), and sfGFP. Linear map of the genetic organization of the pNP sensor is shown in Fig. 2, which demonstrates all parts subcloned to the pFAST vector (Cat. TTC-CA15, Tools, Taiwan).

To confirm the efficiency of our pNP sensor in determining the amount of pNP produced, we measure the GFP fluorescence of E. coli BL21 (DE3) with and without pNP sensor in the presence and absence of pNP.

The result we acquired from the experiment is not consistent with the data previously published (Jha, Ramesh K., et al.). The difference in the level of GFP fluorescence with and without adding 125 µM of pNP is not significant enough to prove the effectiveness of our pNP sensor (Jha, Ramesh K., et al.). Given that the genetic organization and sequence of our pNP sensor is identical to the plasmid design in the research paper, we went back to further examine and check the pNP sensor design. As a result, we discovered the lack of commonly used RBS sequence in front of pNPmut1-1 in the sensor plasmid, from which we inferred that the poor transcription of pNPmut1-1 might be the reason behind the relatively weak and undetectable green fluorescence signals. In Redesign , we are planning to insert RBS by flanking 4 bases apart from the start codon of pNPmut1-1 in the pNP sensor backbone (Fig.5) to further observe if GFP expression will increase in the presence of the same amount of pNP.

The genetic organization and sequence of our pNP sensor plasmid is directly acquired from the published data in section 1G of supplemental data(Jha, Ramesh K., et al.). Observely, we did not acquire data that was consistent with results in the research paper. We have redesigned the plasmid sequences by inserting RBS in the sensor plasmid, which contributes to future research related to pNP sensor design.

Upon iptg induction, the lacI repressor protein will be detached from the lacI gene, leading to the transcription and translation of our target oph gene. In the process of paraoxon degradation, our target gene oph encodes for the enzyme organophosphate hydrolase (OPH), which hydrolyzes paraoxon into dimethyl phosphate (DMP) and p-nitrophenol (pNP) (Fig.1). Since the E.coli bacterial strain BL21 (DE3) has a high level protein expression with T7 RNA polymerase, we chose it as a host cell for our experiment. The vector we used is pET-22b, which includes a T7 promoter (BBa_I712074), lac operator (BBa_K2406019), RBS (BBa_K2924053), OPH gene (BBa_K4271000), and T7 terminator (BBa_K731721) (Fig.6). We also included a pelB signal peptide, which plays a significant role in our experiment by directing our target OPH enzyme to the bacterial periplasm, thereby enhancing the enzyme’s activity at the specific location (Jain, Monika et al.).

Synthetic oph gene we used in this study is derived from the opd (organophosphate degradation) gene in Agrobacterium tumefaciens and performed with codon usage optimization for E. coli heteroexpression. We digested the oph gene with BamHI and HindIII, subcloned it to pET22b vectors that underwent the same restriction enzyme digestion, then transformed the recombinant into E. coli DH5α. The transformation was conducted by plasmid extraction through mini-prep.

We later confirmed the insertion of our oph gene into the enzyme plasmid by enzyme digestion, cutting the recombinant DNA with BamHI and HindIII respectively, and observing the same band sizes of 6.5 kilobases after gel electrophoresis (Fig.7). We later digested our pET22b::OPH again with both BamHI and HindIII, two of resulting DNA bands include the 1071 base-long oph and the 5479 base-long pET22b vector (Fig.8). Finally, the plasmid was transformed into the competent cells E.coli BL21(DE3) via heat shock, which we later used to examine the level of paraoxon degradation by our enzyme plasmid.

To test the degradation of paraoxon by OPH, we detected the level of pNP production with a spectrophotometer. Since pNP (yellow) reaches an absorbance peak at 410 nm, we assume that the absorbance at 410 nm of the colonies under different conditions will provide us with an overview of the efficiency of paraoxon degradation by OPH. We performed two experiments based on this assumption: the amount of pNP at various time points (pNP conc. v.s. Time) and in the presence of different IPTG concentrations at a fixed time (pNP conc. v.s. IPTG conc.).

The results met our expectations as the pNP concentration increased over time, showing that paraoxon is being degraded by the E.coli BL21(DE3) steadily. However, pNP concentration seems to increase rapidly only in the first 5 hours of observation, after which it proceeds to grow steadily, which demonstrates that the enzyme reaches optimal activity after 5 hours of culture.

We later measured the pNP concentration under exposure of different concentrations of IPTG. We discovered that the concentration of pNP reaches a maximum amount when around 250 μM of IPTG is introduced into E.coli BL21(DE3) engineered with OPH. We also inferred from the data that after pNP concentration reaches a maximum at 250μM of IPTG induction, the amount of pNP will not increase as the concentration of IPTG increases.

Since the overabundance of phosphate in water bodies is one of the major causes of eutrophication, we designed genetically engineered E. coli bacteria that are able to increase the uptake of phosphate into the cell. Normally, the Pho regulon in E. coli bacteria is responsible for regulating the amount of phosphate entering the bacteria to maintain the homeostasis of phosphate in the bacteria. To allow the bacteria to consume organic phosphate from the environment limitlessly, we engineered E. coli expressing the antisense AsPhoU (As PhoU), which binds to the phoU mRNA and blocks the translation of PhoU protein, thereby enhancing phosphate transportation into the cell.

In order to determine the amount of phosphate entering the bacteria, we utilized certain components of the PhoU regulon to measure the effectiveness of phosphate transportation. To evaluate the activity of the PstSCAB transporter, we conducted a preliminary experiment that measures the concentration of PhoA via its coloration in low and high phosphate environments. Since the activity of PstSCAB and PhoA are positively correlated, an increase in PhoA concentration will indicate the activity of the PstSCAB transporter. In this preliminary experiment, we added solutions of 5-Bromo-4-chloro-3-indolyl phosphate (XP) because PhoA will severe it into a phosphate ion and a 5,5′-dibromo-4,4′-dichloro-indigo, which makes the solution blue. Arabinose also plays an important role in our preliminary experiment, since it acts as an inducer that promotes AsPhoU to bind on the PhoU sequence.

Another experiment we conducted to measure the effectiveness of phosphate transportation into the cell is to measure the amount of extracellular phosphate in the bacteria via malachite green coloration. A complex of phosphomolybdic acid is formed when molybdate (MoO₄⁻²) interacts with phosphate (PO₄⁻³), which would later interact with malachite and form a green chromogenic complex.

We cultured three different groups of E.coli DH5α (E.coli DH5α, E.coli DH5α with AsPhoU, E.coli DH5α with AsPhoU and arabinose) in both low and high phosphate concentrations. The groups cultured in low phosphate concentration act as the positive control of our preliminary experiment, while the groups cultured in high phosphate concentration function as the negative control.

For the experiment of malachite green coloration, we incubated E.coli DH5α (0.1 O.D.) with AsPhoU and E.coli DH5α with AsPhoU and arabinose under fixed high-phosphate environment (2 mM of K₂HPO₄, 0.06% glucose, and MOPS buffer). We retrieved our E.coli colonies respectively after 1 hour, 2 hours, and 4 hours of incubation in a high-phosphate environment. We then added molybdate and malachite into the tested groups and used a spectrometer to detect the absorbance of phosphate at 600 and 620 nm, since molybdate has a max absorbance rate at 600 nm while that of malachite is at 620 nm.

From our preliminary result, we confirmed that the Pho regulon will only be active in a low phosphate environment as the positive control groups all turned blue, indicating PhoA enzyme activity and indicating PstSCAB activity. Our results also proved that arabinose has the ability to induce AsPhoU binding to the PhoU sequence, since the E. coli DH5α cultured with both AsPhoU and arabinose appears blue even in a high phosphate environment, showing PhoA and PstSCAB activity regardless of PhoU inhibition.

According to the data above, we concluded that E.coli engineered with AsPhoU and induced by arabinose has a significantly higher efficiency in absorbing phosphate. At all three time periods, the absorbance rate of our engineered E.coli cells was higher in the absence of arabinose. From this result, we concluded that arabinose, along with the AsPhoU that induces, does increase phosphate absorption through the PstSCAB transporter even under exposure at 2 mM of phosphate. In addition, the presence of AsPhoU is also proven effective at increasing phosphate absorption, as the two groups of DH5α engineered with AsPhoU show levels of phosphate significantly higher than the other two groups of normal DH5α. Both of the conclusions we obtained from the data further prove that our engineered bacteria has the ability to absorb phosphate from the eutrophicated water bodies, thus reducing the concentration of phosphate in the polluted water.

The design of the polyP sensor plasmid includes genes encoding for mCherry fluorescent protein and RpoD sigma factor with a P region that easily binds to polyphosphate, respectively. Without polyphosphate accumulation, the sigma factor could successfully direct the RNA polymerase to the promoter, resulting in the expression of mCherry fluorescent protein. However, if polyphosphate is fixated, accumulated, and attached to the P region of the sigma factor, the sigma factor loses its function and would be unable to direct the RNA polymerase for transcription of mCherry. Therefore, the increase in phosphate absorption and, therefore, polyphosphate fixation results in reduced mCherry fluorescence.

By measuring the mCherry fluorescence, we could monitor the accumulation of polyP in the cell. The method is employed in the design of our implementation. To ensure that our AsPhoU cells perform at their highest efficiency, we designed the hardware to detect reduced fluorescence levels to the minimum, which signals that the cell has reached maximum phosphate fixation. A notification would then be sent through the designed software to remind the users to replace the filter in time, thus maintaining the effectiveness of our device and further ensuring biosafety. For further information on the implementation design, please visit our Implementation page .

During our partnership with NYCU_Taipei, we adopted their advice of flanking the mCherry gene with a degradation tag in order to increase the specificity of the sensor. The degradation tag enhances the rate of fluorescent protein degradation so that the protein would not accumulate and generate false signals for detection. For more information on our team’s partnership with NYCU_Taipei, please visit our Partnership page .

We committed to Twist Bioscience (ABreal Biotech Co., Taiwan) for synthetic ropD gene (BBa_K4271010), comprising N-terminal 69 amino acid for polyP binding (Yang et al., 2010), E. coli RpoD (1-555 a.a), C-terminal 4.2 region for UreA promoter recognition (BBa_K4271011) (Beier et al., 1998). We conducted reverse PCR to amplify the plasmids, consisting of mCherry (BBa_J18932), and transcriptional terminators (BBa_B0015). Two fragments, Tac promoter (BBa_K4271009) and transcriptional terminators (BBa_B0015) were prepared by PCR. We also amplified the genetic synthetic UreA promoter (containing RBS, BBa_K4271012) by PCR, respectively.

Due to time constraints, we weren’t able to complete the engineering cycle of this particular construct. Yet we have drawn up a future plan for the building of the polyP sensor. Tac promoter (189 bps), synthetic ropD gene (1986 bps), transcriptional terminator (173 bps), and plasmid (2885 bps) will be connected by Gibson assembly to generate the polyP sensor (Fig. 14). The assembly product will then be transformed to E. coli DH5 alpha via the heat shock method. Colony PCR with two primers, PTac _forward and Tt_reverse, will be used to confirm and generate a 3.7 bps fragment if the assembly succeeded.

Beier, D., Spohn, G., Rappuoli, R., & Scarlato, V. (1998, October). Functional analysis of theHelicobacter pyloriprincipal sigma subunit of RNA polymerase reveals that the spacer region is important for efficient transcription. Molecular Microbiology, 30(1), 121–134. https://doi.org/10.1046/j.1365-2958.1998.01043.x

Jha, Ramesh K., et al. “A Microbial Sensor for Organophosphate Hydrolysis Exploiting an Engineered Specificity Switch in a Transcription Factor.” Nucleic Acids Research, vol. 44, no. 17, 2016, pp. 8490–500. Crossref, https://doi.org/10.1093/nar/gkw687.

Jain, Monika et al. “Recombinant organophosphorus hydrolase (OPH) expression in E. coli for the effective detection of organophosphate pesticides.” Protein Expression and Purification, Volume 186, 2021, 105929, ISSN 1046-5928, https://doi.org/10.1016/j.pep.2021.105929

Yang, Z. X., Zhou, Y. N., Yang, Y., & Jin, D. J. (2010, June 11). Polyphosphate binds to the principal sigma factor of RNA polymerase during starvation response in Helicobacter pylori. Molecular Microbiology, 77(3), 618–627. https://doi.org/10.1111/j.1365-2958.2010.07233.x