Team:Bielefeld-CeBiTec/Results/HeavyMetals

Heavy Metals

Results

Figure 1: Influence of heavy metals on the growth of E.coli KRX. The tested concentrations were 20 µg/L lead, 60 µg/L mercury, 60 µg/L chromium, 80 µg/L nickel, 40 mg/L copper, which represent ten times the WHO guideline. The influence of arsenic was not tested as E. coli is known to be resistant to arsenic.

We tested our heavy metal biosensors in Escherichia coli as well as in our cell-free protein synthesis.

Prior to the in vivo characterization, we tested whether the heavy metals have a negative effect on the growth of E. coli.

As can be seen from the figure, we observed no significant difference between the growth in the presence of heavy metals and the controls. This first experiment showed us that in vivo characterization of these sensors is possible. Most cultivations for in vivo characterization were performed in the BioLector. Due to the accuracy of this device, we could measure our samples in duplicates. Subsequently, all functional biosensors were tested in vitro.

Click on the test strip for the results of our biosensor tests in E. coli and in our CFPS:

Arsenic

Figure 2: Genetic build-up of the arsenic biosensor we used for in vivo characterization. Both the repressor (arsR) and the reporter (mRFP1) are under the control of the same promoter, which is controlled by ArsR.

We choose to work with the chromosomal arsenic operon of E. coli, which was used by the team from Edinburgh in 2006. This operon encodes an efflux pump which confers resistance against arsenic. The expression is controlled by the repressor ArsR, which negatively autoregulates its own expression. As^III can bind to three cysteine residues in ArsR. The resulting conformational change deactivates the repressor (Chen, Rosen 2014).

By placing a reporter gene downstream of arsR, an arsenic biosensor can be constructed. In this case, both the repressor and the reporter are under the control of the same promoter (Figure 2). In this respect, the arsenic sensor is different from the other heavy metal biosensors we worked with, as their repressors or activators are expressed constitutively. However, the genetic build-up of the arsenic sensor is well-established. Consequently, we decided to keep this deviating design.

in vivo

We tested an arsenic sensor with mRFP1 as reporter gene in vivo to confirm that the sensor is functional. Furthermore, we examined whether it is possible to detect the safety limit as defined by the WHO. The results are shown in figure 3. We observed a reaction approximately five hours after addition of arsenic. The safety limit of 10 µg/L could clearly be distinguished from the negative control and the fluorescence signal increased up to a concentration of 500 µg/L. The signal in the presence of 1000 µg/L was slightly lower than in the presence of 500 µg/L.

Figure 3: Time course of the induction of an arsenic biosensor with RFP for different arsenic concentrations in vivo. Error bars represent the standard deviation of three biological replicates.

in vitro

Figure 4: Genetic structure of BBa_K1758300, which was used for characterization of the arsenic sensor in vitro. The arsenic operator was placed downstream of the T7 promoter in order to control the expression of sfGFP in the presence of the arsenic repressor, ArsR.

E. coli is resistant to arsenic because it possesses an efflux pump. The cell extract is not protected by such mechanisms, therefore we tested the effect of arsenic on the synthesis of sfGFP. We observed no significant effect for the relevant safety limits of 10 µg/L and 50 µg/L.

Figure 5: Influence of arsenic on cell-free protein synthesis. We measured the expression of sfGFP (BBa_K1758102) in the presence of the two internationally relevant safety limits of arsenic. The relative fluorescence was normalized to the fluorescence of a reference sample (culture supernatant of cells expressing sfGFP). Error bars represent the standard deviation of three biological replicates.

In order to test the arsenic sensor in our cell-free protein synthesis, we cloned a device containing sfGFP under control of the T7 promoter and the arsenic operator combined with our optimized untranslated region (UTR). We tested this device in a cell extract that had been generated from cells expressing the arsenic repressor. As our modeling had shown us, this approach is superior to co-expressing the repressor in the reaction. We observed an induction when adding arsenic up to a concentration of 1.87 mg/L. As high arsenic concentrations inhibit the performance of the CFPS, we normalized the results to this effect. In the final application, this task is performed by our app.

Figure 6: Induction of arsenic sensor in vitro. For this experiment, a cell extract that already contained the arsenic repressor was used in combination with BBa_K1758300. Error bars represent the standard deviation of three biological replicates.

Figure 7: Normalized induction of arsenic sensor in vitro. For this experiment, a cell extract that already contained the arsenic repressor was used in combination with BBa_K1758300. The data were normalized to account for the negative effect of arsenic on cell extract performance. Error bars represent the standard deviation of three biological replicates.

Compared to the in vivo results, the response to arsenic was relatively small and we measured a high background signal. We assume that this is due to the different construct we used in vitro. This construct had been optimized for our CFPS by exchanging the natural promoter for the T7 promoter and exchanging mRFP1 for our optimized sfGFP. However, we assume that the repression in the presence of ArsR was not effective enough to observe a clear induction. The reason is most likely that the distance between the T7 promoter and the arsenic operator was too large. The distance was a result of our cloning strategy and would likely be suitable for E. coli promoters. However, the T7 promoter requires the operator to be very close for an efficient repression (Karig et al. 2012).

In addition, we performed an experiment in which the arsenic repressor was not present in the reaction from the beginning, but was encoded on a second plasmid. The plasmid concentrations we used had been predicted by our model. In accordance with the aforementioned results, we observed no clear repression and addition of arsenic showed no effect. This experiment is discussed on the Modeling pages.

To summarize

The arsenic sensor is the most prominent example of heavy metal biosensors. Our results confirm that it is possible to detect the safety limit of 10 µg/L arsenic with a cell-based arsenic senosor. The results of the tests in our CFPS indicate that this approach is suitable for the detection of arsenic as well. However, we believe that the genetic construct requires further optimization.

References

Chen, Jian; Rosen, Barry P. (2014): Biosensors for inorganic and organic arsenicals. In Biosensors 4 (4), pp. 494–512. DOI: 10.3390/bios4040494.

Karig, David K.; Iyer, Sukanya; Simpson, Michael L.; Doktycz, Mitchel J. (2012): Expression optimization and synthetic gene networks in cell-free systems. In Nucleic acids research 40 (8), pp. 3763–3774. DOI: 10.1093/nar/gkr1191.

Chromium

Chromium is an essential part of the earth´s crust (Mitchell D. Cohen et al.), but most of it is produced trough industrial uses (Paustenbach et al. 2003). We built a biosensor for the detection of hexavalent (Cr^VI), because it is toxic and has cancerogenic effects on the human body. An intoxication of chromium can lead to damages of the nervous system. The World Health Organization recommends a limit of 50 µg/L in drinking water (Guidelines for drinking-water quality 2011, WHO 2003).

in vivo

The chromium sensor (BBa_K1758313) was constructed by using the basic construction we showed in Our biosensors. We work with the chromate inducible operon of Ochrobactrum tritici 5bvl1 which enables a resistance for chromium VI and superoxide. For our sensor we used the Cr^VI dependent repressor chrB which was introduced by team BIT 2013 (BBa_K1058007), and optimized this sequence for the use in E. coli . The repressor protein becomes deactivated by the binding of Cr⁶⁺-ions The associated chromium responsive promoter is chrP (introduced by BIT 2013) (BBa_K1058007). For output we used sfGFP and a 5’UTR untranslated region in front of sfGFP to optimize the expression of the reporter protein and increase its fluorescence (see figure 2).

Our sensor for chromium detection consists of ChrB the repressor and the chromate specific promoter ChrP. The promoter is regulated by ChrB, which binds Cr⁶⁺-ions. Behind the promoter is a sfGFP for detection of a fluorescence signal.

Figure 2: The concept of our in vivo chromium sensor (BBa_K1758313), which consists of the repressor under the control of a constitutive promoter (BBa_K1758310)and the operator and promoter sequence of the chromium inducible promoter. An untranslated region in front of the sfGFP, which is used for detection, enhances its expression (BBa_K1758312)

In vivo we could show that the addition of different concentrations of chromium has different effects to transcription of sfGFP.

Figure 3: Time course of the induction of a chromium biosensor with sfGFP for different chromium concentrations in vivo. The data are measured with BioLector and normalized on OD600. Error bars represent the standard deviation of two biological replicates.

Figure 4: Fluorescence levels at three different stages of cultivation. Shown are levels after 60 minutes, 150 minutes and 650 minutes. Error bars represent the standard deviation of two biological replicates.

We tested our in vivo chromium sensor with sfGFP as reporter gene, to test the functionality of the system. Moreover we tested different chromium concentrations. The kinetic of our sensors response to different chromium concentrations is shown in figure 3. The first 30 hours show a strong increase in fluorescence. After that the increase in fluorescence is slower. For better visualization the kinetics of figure 3 are represented as bars in figure 4. A fluorescence level difference for 60 min, 150 min and 650 min is represented.

Our data lead to the conclusion that in a cell-based system it is possible to detect chromium. In contrast to our expectations with higher chromium concentrations we got lower fluorescence levels. These observations needed further investigation. Additionally the bar chart showed that the chromium sensor needs a long time to get different fluorescence levels at different chromium concentrations in in vivo experiments. The bar chart showed significant differences between the chromium concentrations after 650 minutes.

in vitro

For the characterization of the chromium sensor with CFPS we used parts differing from that we used in vivo characterization. For the in vitro characterization we used a cell extract produced from cells which contain the plasmid (BBa_K1758310). The plasmid contains the gene chrB under the control of a constitutive promoter, so that the cell extract is enriched with repressor molecules. In addition to that we added plasmid-DNA of the chromium specific promoter chrP with 5’UTR-sfGFP under the control of T7-promoter (BBa_K1758314 (figure 6))to the cell extract. The T7-promoter is needed to get a better fluorescence expression.

Figure 5: To produce the cell extract for in vitro characterization a construct (BBa_K1758310 ) with chromium repressor under the control of a constitutive promoter and strong RBS (BBa_K608002) is needed.

T7-chrP-UTR-sfGFP BBa_K1758314 used forin vitro characterization.

Figure 7: Influence of different chromium concentrations on our crude cell extract. Error bars represent the standard deviation of three biological replicates.

Chromium’s influence on the cell extract as shown in figure 7 is minimal for low concentrations. Higher chromium concentrations have a measurable impact on the viability of the cell extract, which is visible at concentrations of 120 µg/L and obvious at concentrations of 240 µg/L chromium.

Figure 8: Chromium specific cell extract made from E. coli cells which already expressed the repressor before cell extract production. Induction with different chromium concentrations. Error bars represent the standard deviation of three biological replicates.

The decrease of fluorescence for higher chromium concentrations in chromium specific cell extract is shown in figure 8. An increase of fluorescence at higher chromium concentrations would have been expected resulting out of the induction of the chromium sensor. A factor which should be considered is the influence of high chromium concentrations to the cell extract. The test for influence of chromium on the specific cell extract, illustrated in figure 7 showed that the influence of chromium at low concentrations is not significant. But the graphic shows that high concentrations of chromium induce fatal damages to the cell extract.

Figure 9: Chromium specific cell extract made from E. coli cells which already expressed the repressor before cell extract production. Induction with different chromium concentrations. The data are normalized on chromium’s influence to the cell extract. Error bars represent the standard deviation of three biological replicates.

Taking the influence of different chromium concentrations under consideration measured fluorescence can be normalized on chromium’s influence on the cell extract (figure 9). Normalized data suggest, that higher concentrations of chromium induce fluorescence in relevance to chromium’s influence on the cell extract.

Figure 10: Chromium sensor with alternative repressor build by team Dundee 2015, which has only the first 15 codons optimized in chromium specific cell extract under the induction with different chromium concentrations. Error bars represent the standard deviation of three biological replicates.

Figure 11: Chromium sensor with alternative repressor build by team Dundee 2015, which has only the first 15 codons optimized in chromium specific cell extract under the induction with different chromium concentrations. Data are normalized on chromium’s influence to the specific cell extract Error bars represent the standard deviation of three biological replicates.

In addition to the measurements of our chromium sensor in CFPS we measured our chromium inducible promoter with the repressor of team Dundee (figure 10, 11), which works similar to ours. In contrast to our repressor only first 15 codons of their repressor are codon-optimized. Measurements with their repressor showed tendencies similar to our measured repressor. After normalization induction with higher chromium concentrations showed a detectable fluorescence response for both measured datasets.

To summarize

Our chromium sensor detects the presence of chromium in vivo, but the outcome differed from our expectations. We would have expected an increase in fluorescence by increasing chromium concentrations. Our in vitro data suggest that these decrease in fluorescence could be explained by chromium’s influence on E. coli which is not reflected in growth but shown by chromium´s influence on the cell extract. Before normalizing the in vitro data the same pattern as in vivo could be observed. After normalization an increase in signal is noticeable. Therefore with optimization our chromium sensor would be compatible to our cell free sensor system.

References

Guidelines for drinking-water quality (2011). 4th ed. Geneva: World Health Organization, zuletzt geprüft am 20.08.2015.

Mitchell D. Cohen; Biserka Kargacin; Catherine B. Klein; and Max Costa: Mechanisms of Chromium Carcinogenicity and Toxicity, zuletzt geprüft am 19.08.2015.

Paustenbach, Dennis J.; Finley, Brent L.; Mowat, Fionna S.; Kerger, Brent D. (2003): Human health risk and exposure assessment of chromium (VI) in tap water. In: Journal of toxicology and environmental health. Part A 66 (14), S. 1295–1339. DOI: 10.1080/15287390306388.

WHO (2003): Mercury in Drinking-water Background document for development of WHO Guidelines for Drinking-water Quality, checked 15.08.15

Copper

There are some possibilities for the contamination of drinking water with copper, for the production of pipes, valves and fittings copper is used (Guidelines for Drinking-water Quality, Fourth Edition). Copper is an essential trace element for humans, animals and plants, but an overdose can lead to anemia, liver and brain damages (US EPA ORD NCEA Integrated Risk Information System (IRIS) 2014). Additionally, high input of copper is associated with aging diseases as Atherosclerosis and Alzheimer’s disease (Brewer 2012). These damages can finally cause death. The World Health Organization recommends a limit of 2 mg/L in drinking water (Guidelines for Drinking-water Quality, Fourth Edition).

in vivo

Our sensor for copper detection consists of CueR a MerR like activator and the copper specific promoter copAP. The promoter is regulated by CueR, which binds Cu ²⁺ ions. We also used a sfGFP downstream the promoter for detection through a fluorescence signal.

For our copper sensor we used the native operator of cooper homeostasis from E.coli K12. We constructed a part (BBa_K1758324) using the basic genetic structur shown in our biosensors.The operator sequence, which includes the promoter (copAP), is regulated by the activator CueR. In BBa_K1758324 we combined a codon optimized version of cueR (BBa_K1758320) under the control of a constitutive promoter with sfGFP under the control of the corresponding promoter copAP (BBa_K1758321)(figure 2). Through the addition of a 5’ UTR upstream of the sfGFP we optimized the expression of sfGFP and increased fluorescence.

Figure 2: The concept of our in vivo copper sensor (BBa_K1758324), which consists of the activator under the control of a constitutive promoter (BBa_K1758320) and the operator and promoter sequence of the copper inducible promoter. An untranslated region in front of the sfGFP, which is used for detection, enhances its expression (BBa_K1758323).

Figure 3: Time course of the induction of a copper biosensor with sfGFP for different copper concentrations in vivo. The data are measured with BioLector and normalized on OD₆₀₀. Error bars represent the standard deviation of two biological replicates.

Figure 4: Fluorescence levels at three different stages of cultivation. Shown are levels after 60 minutes, 150 minutes and 650 minutes.

We tested our in vivo copper sensor with sfGFP as reporter gene, to test the functionality of the system. Moreover, we tested different copper concentrations. The kinetic of our sensors response to different copper concentrations is shown in figure 3. The first ten hours show a strong increase in fluorescence. After that the increase in fluorescence is slower. For better visualization the kinetics of figure 3 are represented as bars in figure 4. A fluorescence level difference for 60 min, 150 min and 650 min is represented.

In vivo we could show that the adding different concentrations of copper has effects on the transcription levels of sfGFP.

The shown data suggest that sensing copper with our device is possible even if the detectable concentrations are higher than the desireble sensitivity limits. Therfore we tested the copper sensor in our in vitro transcription translation approach.

in vitro

For the characterization of the copper sensor with CFPS we used parts differing from that we used in vivo characterization. For the in vitro characterization we used a cell extract out of cells which contain the plasmid (BBa_K1758320) (figure 5), so that the resulting extract is enriched with the activator CueR. To this extract we added plasmid-DNA of the copper specific promoter copAP with 5’-UTR-sfGFP under the control of T7-promoter (BBa_K1758325) to the cell extract. The T7-promoter is needed to get a better fluorescence expression.

Figure 5: To produce the cell extract for in vitro characterization a construct (BBa_K1758320) with copper activator under the control of a constitutive promoter and strong RBS (BBa_K608002) is needed.

Figure 6: T7-copAP-UTR-sfGFP BBa_K1758325 used for in vitro characterization.

The results presented in figure 7 illustrate the influences of different copper concentrations on the cell extract.

Figure 7: Influence of different copper concentrations on our crude cell extract. Error bars represent the standard deviation of three biological replicates.

As shown in figure 7 copper has no negative influence on the functionality of our cell extract. Therefore, a relatively stable system for copper sensing is provided. First tests with specific cell extract and different copper concentrations lead to further tests and normalizations, illustrated in figure 8.

Figure 8: Copper specific cell extract made from E. coli cells which have already expressed the activator before cell extract production. Induction of copper inducible promoter without T7 upstream of the operator site with different copper concentrations. Error bars represent the standard deviation of three biological replicates.

Figure 9: Copper specific cell extract made from E. coli cells which have already expressed the activator before cell extract production. Induction of copper inducible promoter without T7 in front of the operator site with different copper concentrations. Error bars represent the standard deviation of three biological replicates. Data are normalized on coppers influence to the cell extract.

In addition,we measured the operator device under the control of T7 promoter as described before.

Fluorescence was normalized to influence of copper on the the cell extract (figure 10 and figure 11).

Figure 10: Copper specific cell extract made from E. coli cells which have already expressed the activator before cell extract production. Induction with different copper concentrations. Error bars represent the standard deviation of three biological replicates.

Figure 11: Copper specific cell extract made from E. coli cells which have already expressed the activator before cell extract production. Induction of copper inducible promoter with different copper concentrations. Error bars represent the standard deviation of three biological replicates. Data are normalized on coppers influence to the cell extract.

Compared to the former fluorescence levels the T7 reporter device showed higher levels. Therefore, a reporter device under the control of T7 promoter is more suitable for our CFPS.

After normalizing on coppers influence to the cell extract these differences were even more obvious.

To summarize

Our copper sensors in vivo data show that detection of different copper concentrations is possible. The fluorescence levels differ clearly between different induction concentrations. As shown above higher copper concentration, higher the fluorescence signal. Therefore the concept of our sensor is functional even if the concentration needed for induction are to high to reach sensitively concerning the WHO guidelines for copper. Our sensor has been tested in vitro as well. For copper we tested our original CopAP construct without a T7 promoter in front of the inducible at first. After realizing that the sensor shows the right tendencies but the general fluorescence is quite low we created an inducible promoter under the control of a T7 promoter to use in CFPS. Fluorescence levels of this device showed the same tendencies as the one without but were higher fluorescence’s, which helps detecting it.

References

Background document for development of WHO Guidelines for Drinking-water Quality, checked on 9/9/2015. Copper excess, zinc deficiency, and cognition loss in Alzheimer's disease - Brewer - 2012 - BioFactors - Wiley Online Library. Available online at http://onlinelibrary.wiley.com/doi/10.1002/biof.1005/abstract, checked on 8/28/2015.

US EPA ORD NCEA Integrated Risk Information System (IRIS) (2014): Copper (CASRN 7440-50-8) | IRIS | US EPA. Available online at http://www.epa.gov/iris/subst/0368.htm, updated on 10/31/2014, checked on 9/2/2015.

Lead

Lead is one of the most frequently used metals. Therefore, lead is found in many different parts of the environment (World Health Organization (WHO): Fact sheet number 379, Lead poisoning and health). The contamination of drinking water is often caused by obstructed pipes. Long time absorption leads to adverse health effects in most organs in the body (EPA Health Effects: How Lead Affects the body). The main targets are the nervous system, brain and liver. These damages can finally cause death. The World Health Organization recommends a limit of 10 µg of lead in 1 litre drinking water (WHO: Guidelines for Drinking-water Quality, fourth edition (2015)).

in vivo characterization

In addition to the other heavy metal sensors, we constructed a sensor for lead detection. It consists of the repressor PbrR which binds at the operator box downstream of the pbrAP promoter. The binding of the repressor is reversible in the presence of Pb²⁺ Ions. Those ions can weakened the repressors binding and hence, all genes downstream of the pbrAP promoter can be expressed. Like the former sensors this one encloses a sfGFP for detection via fluorescence. So if no lead is present in the media, the repressor binds to the operator box and the pbrAP promoter is blocked meaning that the transcription of sfGFP is prevented. No fluorescence signal is the results. By supplementation of lead, the repressor is separated from the operator box and the genes downstream of the promoter can be expressed.

The pbrAP promoter, the operator box and the PbrR repressor are parts of the chromosomal lead operon of Cupriavidus metallidurans (figure 2). This was cloned and transformed into E.coli KRX. This operon includes now the promoter pbrAP (BBa_K1758332), which is regulated by the repressor PbrR. The PbrR belongs to the MerR family, of metal-sensing regulatory proteins, and is Pb²⁺-inducible. Our sensor system comprises pbrR (BBa_K1758330), which is under the control of a constitutive Promoter and pbrAP and a 5’ untranslated region, which controls the transcription of a sfGFP and increases the fluorescence. Fluorescence implemented by sfGFP protein is the measured output signal (figure 3 and figure 4).

Figure 2: The concept of our in vivo lead sensor (BBa_K1758332) which consists of the repressor under the control of a constitutive promoter (BBa_K17583230) and the operator and promoter sequence of the lead inducible promoter. An untranslated region in front of the sfGFP, which is used for detection, enhances its expression (BBa_K1758332).

Figure 3: Time course of the induction of a lead biosensor with sfGFP for different lead concentrations in vivo. The data are measured with BioLector and normalized to the OD₆₀₀. Error bars represent the standard deviation of two biological replicates.

Figure 4: Fluorescence levels at three different stages of cultivation. Shown are levels after 60 minutes, 150 minutes and 650 minutes. Error bars represent the standard deviation of two biological replicates.

We tested our lead sensor with sfGFP as reporter gene to verify the functionality of the system. Subsequently, we tested different lead concentrations. The kinetic of our sensors response to different lead concentrations is shown in figure 3. The first 40 hours show a strong increase in fluorescence. After that the increase in fluorescence reaches a plateau. For better visualization the kinetics of figure 3 are represented as bars in figure 4. A fluorescence level difference for 60 min, 150 min and 650 min is represented.

The results of the lead sensor show no significant differences between the different concentrations (figure 3). This might be due to the pbrAP’s weak promoter strength in E. coli. Further reasons are most likely in the weak repressor binding to its operator. So, we suggest for the usage of this sensor, it has to be optimized. Moreover we were lacking time for further in vivo characterizations and different experimental setups. Hence, we did not use this sensor in further experiments regarding Cell-free-Protein-synthesis (CFPS). . In the future a characterization in the CFPS systems would be desirable.

To summarize

Our lead sensor was characterized in vivo only. The differences between inductions with various lead concentrations are really slight therefore this sensor needs further optimization which was not possible in this limited time. But as there is a fluorescence response to lead this sensor has the potential work as expected. In the future a characterization in CFPS systems would be interesting.

References

EPA Health Effects: How Lead Affects the body, checked on 2015-09-17.

WHO Guidelines for Drinking-water Quality fourth edition, checked on 2015-09-09.

WHO lead poisoning and health, fact sheet number 379, reviewed August 2015, checked on 2015-09-17.

Mercury

The main sources of mercury exposure are generated through humans. For example mercury contamination can be caused by medical waste (damaged measurement instruments), Fluorescent-lamps, Chloralkali plants and thermal power plants. In the environment, mercury is one of the most toxic elements . Acute effects of a mercury intoxication can range from diseases of the liver, kidney, gastrointestinal tract, to neuromuscular and neurological problems. A chronic intoxication of mercury results in kidney changes, changes in the central nervous system and other effects like cancer. The World Health Organization recommends a limit of 6 µg/L in drinking water.

in vivo

One of the already existing sensors we used for our system is the mercury sensor consisting of MerR the activator and the mercury specific promoter pmerT. The promoter is regulated by the MerR, which binds Hg²⁺-ions. Similar to the former sensors we added a sfGFP for detection via fluorescence.

For our mercury sensor we used parts of the mercury sensor constructed by iGEM team Peking 2010. These parts consist of the mercury dependent mer operon from Shigella flexneri R100 plasmid Tn21. The expression of the genes in the mer operon depends on the regulation by MerR its activator and promoter PmerT. For our sensor we used the codon optimized activator (BBa_K1758340), under control of a constitutive promoter,(BBa_K346001). Additionally to this activator we designed and constructed the specific promoter PmerT(BBa_K346002)(figure 2). For our sensor we added a 5’-UTR downstreamd of this promoter, which increased the fluorscence of the used reporter protein sfGFP.

Figure 2: The concept of our in vivo mercury sensor ( BBa_K1758343), which consists of the activator under the control of a constitutive promoter BBa_K1758340)and the operator and promoter sequence of the mercury inducible promoter. An untranslated region in front of the sfGFP, which is used for detection, enhances its expression ( BBa_K1758342).

Figure 3: During cultivation the sfGFP signal in reaction to different mercury concentrations was measured. The induction with mercury happened after 165 minutes. Error bars represent the standard deviation of three biological replicates.

Figure 4: Fluorescence levels at two different stages of cultivation. Shown are levels after 120 minutes and 190 minutes. Error bars represent the standard deviation of three biological replicates.

We tested our mercury sensor with sfGFP as reporter gene, to test the functionality of the system. Moreover we tested different concentrations. The kinetic of our sensors response to different mercury concentrations is shown in figure 3. A strong increase in fluorescence levels is notecible after induction with mercury after 120 min. For better visualization the kinetics of figure 3 are represented as bars in figure 4. A fluorescence level difference for 120 min and 190 min is represented.

In vivo data show a highly significant, well working sensor, which even reacts to concentrations below the threshold of the water guidelines by the WHO (Figure 3 and 4).

The mercury detection was measured during the cultivation of E. coli KRX at 37 °C (Figure 3 and 4). The strain contained the plasmid with the activator merR under the control of a constitutive promoter and the specific promoter with an operator binding site, which reacts to the activator with bound Hg ²⁺-ions. The specific promoter is located upstream of the sfGFP CDS. Therefore, the mercury in the medium is detected via formation of sfGFP. In vivothis sensor devise shows a fast answer to occurrence of his heavy metal contrary to the other sensor systems In vivo.

Therefore we tested our sensor in vitro to check if an already functioning highly optimized sensor provides required data for guideline detections

in vitro

For the characterization of the mercury sensor with CFPS we used parts differing from that we used in the in vivo characterization. For the in vitro characterization we used a cell extract out of cells, which contained the plasmid ( BBa_K1758340)(figure 5). In addition, we added plasmid DNA to the cell extract. This plasmid consisted of the mercury specific promoter pmerT with 5’-UTR-sfGFP. The entire sequence was placed under the control of of T7-promoter ( BBa_K1758344)(figure 6). The T7-promoter is needed to get a better fluorescence expression.

Figure 5: To produce the cell extract for in vitro characterization a construct (BBa_K175840) with chromium repressor under the control of a constitutive promoter and strong RBS (BBa_K608002) is needed.

T7-PmerT-UTR-sfGFP BBa_K175844 used forin vitro characterization.

Figure 7: Influence of different mercury concentrations on our crude cell extract. Error bars represent the standard deviation of three biological replicates.

Figure 8: Mercury specific cell extract made from E. coli cells, which have already expressed the activator before cell extract production. Induction of mercury inducible promoter without T7 in front of the operator site with different mercury concentrations. Error bars represent the standard deviation of three biological replicates.

Figure 9: Mercury specific cell extract made from E. coli cells, which have already expressed the activator before cell extract production. Induction of mercury inducible promoter without T7 in front of the operator site with different mercury concentrations. Error bars represent the standard deviation of three biological replicates.

In vitro this sensor showed good results. The fluorescence level was high at low concentrations. Additionally, it showed that the expression level at 6 µg/L (Guideline of WHO for Mercury) reached the maximal signal. This result indicated the potential for measurement of concentrations under 6 µg/L.To confirm this hypothesis, it takes more experiments and tests with lower concentrations. Due to the high expression of sfGFP at low concentrations and the same expression level at different concentrations, it is not possible to quantify mercury with CFPS analyses . , Our model predicted this observation. During the measurement we noticed that the heavy metals have negative influences on the cell extract. Because of this fact, we used a correction factor, which resulted from the heavy metals influence on the CFPS system. This already optimized sensor showed the high potential of optimized sensors in CFPS.

To summarize

We demonstrated, that our mercury sensor works well in vivo (figure 4). There is a clearly noticeable increase in fluorescence after induction with mercury. Even the threshold concentration, which is given in the WHO guideline, can be measured. This well working sensor was tested in in vitro as well. The presented data suggested, that maximal output is reached at concentrations of 6 µg/ L,which represent the value of the former mentioned WHO guideline. Further optimization could lead to a decreasing in vitro detection threshold.

Nickel

Naturally nickel occurrences are quite low and rare. In Germany most drinking water pollutions by nickel take place in the last meters of the plumbing system. The reason for these pollutions are wrong tap ware. There may be higher concentrations in drinking-water in special cases of release from natural or industrial nickel deposits. Higher nickel concentrations in water can lead to dermatitis as itching of fingers and other parts of the body through long term skin contact. The World Health Organization (WHO) recommends a limit of 70 µg/L in drinking water.

in vivo

We aimed to construct a sensor for nickel detection. It consists of rcnR the repressor and the nickel specific promoter prcnA. The promoter is regulated by the RcnR, which binds Ni²⁺-ions. As the former sensors this one encloses a sfGFP for detection via fluorescence.

Our nickel biosensor consists of parts of the rcn-operon from E. coli, which encodes a nickel- and cobalt-efflux system. This system is highly sensitive to nickel. In absence of nickel or cobalt RcnR binds to the operator and inhibits the nickel responsive promoter. With Ni²⁺-ions present the repression of the promoter prcnA will be reversed, because the repressor RcnR binds Ni²⁺-ions and cannot attach to the DNA. For our biosensor we construct the part BBa_K1758353 by using the basic construction shown in figure 2. For this part we used the repressor RcnR under control of a constitutive promoter ( BBa_K1758350 ) and the nickel specific promoter PrcnA with a 5’UTR in front of sfGFP ( BBa_K1758352 ) as reporter protein.

Figure 2: The concept of our in vivo nickel sensor ( BBa_K1758354), which consists of the activator under the control of a constitutive promoter ( BBa_K1758350)and the operator and promoter sequence of the nickel inducible promoter. An untranslated region in front of the sfGFP, which is used for detection, enhances its expression ( BBa_K1758352)

Figure 3: Time course of the induction of a nickel biosensor with sfGFP for different nickel concentrations in vivo. The data are measured with BioLector and normalized on OD₆₀₀. Error bars represent the standard deviation of two biological replicates.

Figure 4: Fluorescence levels at three different stages of cultivation. Shown are levels after 60 minutes, 150 minutes and 650 minutes. Error bars represent the standard deviation of three biological replicates.

We tested our nickel sensor with sfGFP as reporter gene, to test the functionality of the system. Moreover we tested different concentrations. The kinetic of our sensors response to different nickel concentrations is shown in figure 3. The first five hours show a strong decrease in fluorescence. After that there is a slight increase in fluorescence. Starting levels of fluorescence are not reached. For better visualization the kinetics of figure 3 are represented as bars in figure 4. A fluorescence level difference for 60 min, 150 min and 650 min is represented.

The data for our nickel sensor show a trend that differs for that of the other sensors. There is no indication for a working sensor in vivo (Figure 3 and 4). There is a fluorescence signal, but it decreases in the first five hours. After reaching a minimum the fluorescence increases slowly. Additionally, there is no difference in fluorescence as response to various nickel concentration. Nickel could influence the cells and thereby caused a precipitation, which could result in decrease of fluorescence. With this sensor no production of sfGFP via fluorescence level change could be detected. Therefore, this sensor is not suitable for our approach. Due to this observation, no in vitro data using CFPS were taken.

To summarize

With this sensor no production of sfGFP via fluorescence level change could be detected. Therefore, this sensor is not suitable for our approach. Consequently, no in vitro tests were performed. To create a working sensor based on this concept further optimization is needed.

To summarize all

We have characterized heavy metal sensors for arsenic, chromium, copper, lead, mercury and nickel. The results for our nickel characterization indicated that the constructed nickel sensor is not suitable for our test strip. The sensors for lead and chromium showed great potential, as they showed responses to chromium or lead, but require further optimization. Copper, our new heavy metal sensor, worked as expected and was able to detect different copper concentrations. The already well-characterized sensors for arsenic and mercury were tested as well. While the arsenic sensor worked well in vivo, it requires some omptimization for the use in vitro. Mercury showed that a fully optimized sensor works very well in our in vitro system and has the potential to detect even lower concentrations than in vivo.

And what about the date rape drugs?