To make sure the plasmid is cloned into E. coli BL-21 strain, we extracted the plasmid from transformed E. coli.(figure 1) We validated the gene by PCR with specific primers and then we examined the result with Agarose gel electrophoresis. A successful cloning was verified from the results of a PCR performed with designed specific primers according to the theoretically expected length of horseradish peroxidase(927bp).

To obtain functional horseradish peroxidase we need to purify the protein from E. coli, but this kind of protein does not have any function and it is called apoprotein. After the first time purification, we refolded the protein to construct the correct structure and then we activated the apo-HRP with hemin to produce the functional HRP. We examined the existence of this protein by SDS-PAGE after purification. (figure 3)

To prove the degradation ability of HRP, we mixed 25 µg of HRP with 1 mM H2O2 and BPA or NP in the 1 mL water. The environment of degradation is suitable for HRP to degrade the phenolic compounds(40℃ and pH=6-7). After degradation for 24 hours, we denatured HRP by briefly heating up to 80℃ and use LC-PDA (Liquid Chromatography - Photodiode Array detector) to analysis the result of degradation.(figure 4 and 5)

From the result of LC-PDA, we found that there are two extra peaks showing after degradation and these extra peaks represent the by-products from degradation. Unfortunately, we can’t know how much EDCs is degraded by HRP because:

(1) The peak of the by-products overlap the peak of EDCs and we can’t get the information of remained EDCs. (There is a by-product have a broad peak from 2.2 min to 3.5min and its intensity also much higher than the peak of EDCs)

(2) The quality of the column from LC isn’t good enough to separate the by-products from EDCs due to their similar molecular properties. (This LC can’t distinguish between BPA and NP so we estimated that it can’t distinguish the degraded by-product from BPA or NP as well.)

Moreover, we use the mass spectrum to prove HRP can degrade BPA and NP by the signals of the large molecular weight of by-products.(figure 6-9) We found that there are lots of additional peaks show up after degraded by HRP, and this result can prove our HRP can degrade BPA and NP.

We cloned the sequence of ER-alpha and Monobody into the vector, and then we transformed the plasmid into BL-21 competent cells, respectively.(Figure 10 and 11) After transformation, we extracted both plasmids from the cells and the vector was validated by gel electrophoresis.

A successful cloning was verified from the results of a PCR performed with specifically designed primers according to the theoretically expected length of ER-alpha (2199bp) and Monobody (797bp).(Figure 12)

We compared the difference between E. coli expressed RFP and RFP-INP(figure 13). After the lysis of culture expressing RFP-INP, most of RFP was on the fragments of the membrane. Due to the larger structure of plasma membrane, it is easily spin down. If RFP-INP is on the E. coli plasma membrane when we centrifuged the cell lysate, RFP-INP is spun down with the plasma membrane, and we see a red pellet.

On contract, after cell lysis of RFP expression E. coli, most of RFP suspended in the solution. Since the centrifugation force, we set in this experiment is not enough to spin down particles as small as RFP. We see RFP suspended in the supernatant even after centrifugation.

This experiment showed that INP can bring RFP to the membrane of E. coli, and this result proved the function of INP.

To prove that our detection system can distinguish the sample has EDCs or not, we used IR spectrum to measure specific bonds on E. coli and verify the detection function.(Table 1)

In the beginning, we try to use live cells for the detection system. However, we found that all the samples have the information of E. coli. Showing that there is no difference between the sample with 5mM EDCs and the sample with no EDCs. (Figure 14)

To decrease the activity of E. coli, we tried to freeze E. coli in -80℃ for 24 hours and use it after thawing on ice immediately. We found that there is a significant difference between the sample with 5mM EDCs and EDCs-free and the intensity of IR signals also decrease due to the reduction of mobility of E. coli.(Figure 15) This result can prove that our detection system can detect EDCs in the water.

To prove our detection system can measure the concentration of EDCs, we use crystal violet to stain E. coli on the surface of gold and then observe the density of stained E. coli under microscope. Due to the surface tension force of water, there are lots of E. coli remained around the drop edge of sample, and we only observed the area in the middle of the sample to reduce the error. (Figure 16)

We observed the density of E. coli when the samples contained different concentration of BPA and NP from 5mM to 5nM (Figure 17 and 18). This result indicates that when the concentration of EDCs decreases, the density of E. coli will decrease as well.

For the practical propose, we added trehalose (0.1M) to prevent ER-alpha from denaturing and increase the time of conservation. We also observed the density of E. coli with samples containing the different concentration of BPA and NP from 5mM to 5nM.(Figure 19 and 20) From the result of observation, we can find the similar tendency as the sample without trehalose.

To clarify the relationship between the concentration of EDCs and the density of E. coli, we use image J to count the amount of E. coli on the surface. (Figure 21-24) From the results of imageJ, we can get the similar tendency as the result from the pictures of the microscope. However, the density of E. coli decreases in the presence of 0.1 M trehalose, and this result suggested that although trehalose can improve the time to store lyophilized E. coli, it can decrease the precision of our detection system.

To further understand the limitation of the concentration that our system can achieve, we observed the density of E. coli in four different conditions of background. We find that when there are no EDCs in the sample, the density of E. coli is close to the concentration of EDCs below 5µM. Showing that our detection system can’t measure the concentration of EDCs less than 5µM.(Figure 25 and Table 2)

1. Since we failed to express GFP and ER-alpha at the same time, we will try to construct the gene containing GFP and ER-alpha and express them together for the purpose of detection. (Figure 26)

2. Since we don’t have time to construct the detection system to measure the change of fluorescence or surface plasmon resonance, we will construct the detection system to estimate the concentration of EDCs in the water precisely. (Figure 27)