For the implementation, we aim to develop a device that assists with our synthetic biological solution to eutrophication. Eutrophication has been a significant environmental issue for worldwide aquatic environments. Excessively used fertilizers, detergents, and industrial discharge containing phosphorous flow into water bodies and cause algal bloom. The decomposition of the algae leads to the overproduction of CO2 and increased O2 consumption in the aquatic environment. Animals’ lives are threatened by the lack of O2. The fish kills and dead zones have ultimately resulted in significant ecological and economical losses.

Reservoirs in Taiwan have suffered from eutrophication mainly because of wastewater containing agricultural fertilizers and home-use shampoo with high phosphate concentration that is released into water bodies. In 2021, sixty-five percent of Taiwan’s reservoirs were eutrophic. Therefore, we target the local Taiwanese reservoirs for the implementation of our engineered bacteria.

We have also taken into consideration that temperatures below 21-degree celsius do not allow engineered bacteria to maintain the function or optimal enzymatic function. Since the temperature of Taiwan's local reservoirs has a temperature ranging from 20 to 30 degrees Celsius, we believe that they are the suitable choice of target for our product. In addition, the device is best suited to achieve purification of eutrophic water bodies in which the major factor exacerbating eutrophication is phosphorus, which is the case for our target reservoirs in Taiwan.

The goal of the implementation design is to decrease organic and inorganic phosphate levels in the water body. The idea of our device hardware consists of an artificial island with a filtering device containing our engineered E. coli, TripleP cells carrying compatible pACYCDuet::OPH/AsPhoU and polyP sensor, with the ability to hydrolyze organic phosphate and overtly absorb, as well as fixate, inorganic phosphate. The device is attached with filter papers on both ends that allow eutrophic water to flow through while preventing the engineered E. coli from leaking into the environment due to safety issues.

During the interviews by our human practice team, the advice and concerns from experts and stakeholders contributed to the development of our implementation in regard to our hardware and software designs, biosafety measures, and future goals. For more information on our human practice works, please visit our Integrated Human Practice Page .

In light of the issue, we wish to decrease the amount of phosphate in bodies of water to resolve eutrophication. Our bacteria “TripleP” are engineered to degrade organic phosphate and absorb inorganic phosphate from the water bodies. In order to complete our goal, we designed three main target genes subcloned into two plasmids coexisting in the bacteria.

The first plasmid is subcloned with the organic phosphate hydrolase (OPH) gene, which hydrolyzes organic phosphate pollutants such as paraoxon to lessen its harm to the environment, and the anti-sense PhoU (AsPhoU) gene, which reduces the expression of PhoU and increase the intake of inorganic phosphate. The second plasmid contains the polyphosphate sensor (PolyP sensor), which produces mCherry fluorescent protein for the detection of polyphosphate overaccumulation and, therefore, the need to notify the users to replace the filter.

Further information about the design of OPH and AsPhoU genes, as well as the PolyP sensor cells, can be found on the Engineering Success page .

Further information about the functions of OPH and AsPhoU genes can be found on the Proof of Concept page .

The main part of the device is a filtering tube containing the engineered E. coli, TripleP. We used filter papers with a hole size smaller than bacteria, and an acrylic tube to prevent bacteria leakage. While the eutrophic water will enter the filter tube through the filter papers, the engineered bacteria can absorb the external phosphate. As the phosphate concentration within the filter tube declines, the phosphate outside the filter will diffuse inward. However, the rate of phosphate absorbance will fall as the phosphate is fixated in the engineering E. coli. Since polyphosphate would inhibit the expression of mCherry fluorescence proteins, the light intensity of fluorescence proteins could reach a minimum when more phosphate is present in the bacteria. This in turn can be an observable indicator for the user to tell when to change the filter. To observe the minimum of fluorescent proteins we use the Arduino light sensor and wifi module. The chip is programmed to control the light beam and light sensor, transport the fluorescence data to device users, and notify them when the user should change the filter. Every six hours, the light source will beam lights on the filter and excite the fluorescence proteins; then, the light sensor will record the light intensity of the exciting proteins and send it to the chip. As the data collected from the sensor shows a minimum in light intensity, the chip will notify the user to change the filter so the device can continue to filter the eutrophic water at full rate.

Moreover, all of the devices will be attached to a giant artificial island set in the water body by threads, so the device can float in the water. We have also designed solar panels generating energy to propel collecting the energy which the electronics and the chip need. For more information about the hardware design, please visit our Hardware Page .

During our interviews, several concerns and advice were raised by the experts and stakeholders that led to the improvements of our device. First of all, we decided to address the secretion blockage problem by adding bacterial houses made of sodium alginate, which allows E. coli to adhere and grow on the surface of the houses, thus preventing secreted substances from covering the filter pores and blocking water infiltration. Secondly, aquatic plants will be planted on the artificial islands to improve the phosphorus extraction efficiency of our device since the roots of plants are able to simultaneously absorb phosphorus in bodies of water.

The designs of our implementation hardware and software have taken biosafety into consideration. Instead of adopting the more common biosafety design, such as the kill switch, we provide three ways to prevent the problems, including a bacteria filter, taking advantage of bacteria’s natural growth inhibition, and engineering an alerting biosensor. To ensure that the engineered bacteria in the filtering device will not be accidentally released into the environment, we attached filtering paper with a pore size smaller than the bacteria on both sides of the tube. In the event of an unexpected bacteria leakage, the over absorbance and fixation of polyphosphate would also inhibit the growth of bacteria, which lessens the threat it might cause to the environment. In addition, the detection of fluorescence from the expression of the mCherry biosensor and the alert that would then be sent to the user, which decreases the risk of the release of phosphate back into the environment after bacteria death and lysis.

Our approaches ensure the comprehensive biosafety of our implementation design without causing any decrease in bacteria function’s efficiency as a leaky expression of a kill switch gene might. Please visit our Safety Page for more information on the safety and security of our implementation design.

Concerning the software implementation of our project, our intention is to elaborate on the topic of eutrophication and raise public awareness of this worldwide aquatic issue. Our software features the calculation of nitrogen and phosphorus in a body of water to determine the status of eutrophication. Moreover, we also construct a Taiwan map for the user to tap on the annotated reservoir, and the system will return sets of information manifesting the extent that the reservoirs are influenced by eutrophication. Please visit our Software Page for more information on software design.

Our interviews have also contributed to the improvements in our software. Specifically, we incorporated the Carlson Trophic State Index (CTSI), as advised by experts, into the software application to help users gain a better understanding of the precise water quality of the reservoir being tested. To be specific, by adding the three parameters of Transparency (SD), Chlorophyll-a (Chl-a), and Total Phosphate (TP) along with the three classifications of eutrophication levels: oligotrophic, mesotrophic, and eutrophic, our software ensures data quality and the variety of information presented to users.

To allow our hardware device to reach the highest effectiveness, we plan to design physical and chemical methods for our device to be able to measure the level of eutrophication, especially the concentration of phosphorus. With our improved design, the filtering device could be deployed at specific locations to improve the efficiency of phosphate elimination.

To bring our implementation to a bigger scale with the advice we received from the human practice works, we are planning to further prevent the membrane blockage problem by applying additional appliances that would spout water and wash or a brush that would scrub unwanted materials off the surface of the membrane. Another idea would be to use filter sand or silver-coded filter papers to prevent bacteria from growing on the surfaces. Going for an even greater leap, we learned that molecular methods could be incorporated into our implementation. Through ionic attractions between anion and cations, negatively charged phosphorus would adhere to the positively charged exterior at the bottom of our device. This idea has great potential in improving the efficiency of our product; therefore, we are planning on incorporating it into our future plans and implementing it after further research and redesign.

For the software, we hope to achieve the connection between the hardware device and the user’s mobile device through the wifi module of the Arduino chip in the future. The users could then receive notifications from the hardware's monitoring of the bacteria's condition.

Our implementation provides an alternative to the current solutions to eutrophication, such as MSL. While MSL is only suited for smaller water bodies, such as ponds, or sewers before the effluent enters water bodies, our device could be implemented in larger water bodies including reservoirs.

In the future, we hoped to expand the implementation of the device from a local to a global level and alleviate eutrophication worldwide. We will continue to consult experts’ advice, as well as other stakeholders’ opinions, and improve the design of our device further.

牟麗娥、游琇如、張博雅。〈台灣主要水庫歷年水質變化特徵〉，第96頁。行政院環境保護署、清華科技檢驗股份有限公司。