From 2011.igem.org

What we improved since the European Regional Jamboree

BPA degradation

Specificity of bisphenol A degradation with E. coli

In order to access the specificity of the bisphenol A degradation by the bisdA | bisdB fusion protein we tested how well two similar bisphenols, bisphenol F (BPF) and bisphenol S (BPS), were employed. The structure of those bisphenols differs only in the chemical group linking the two phenols from that of bisphenol A (see Figure 1).

Figure 1: Chemical structure of BPA, BPF and BPS showing the different chemical groups linking the two phenols.

BPF and BPS are used in a broad range of applications that involve the use of polycarbonates or epoxy resins and thus can often be found were BPA is also present. Accordingly their presence is a potential disruptive factor that could lead to a false positive signal with our biosensor. This is especially true for BPS that in some cases is used as a substitute for BPA in baby bottles [http://www.nytimes.com/2011/04/18/business/global/18iht-rbog-plastic-18.html]. Studies concerning the environmental pollution with BPF ([http://www.sciencedirect.com/science/article/pii/S0043135401003670 Fromme et al. (2002)]) and the acute toxicity, mutagenicity and estrogenicity of BPF and BPS ([http://onlinelibrary.wiley.com/doi/10.1002/tox.10035/abstract Chen et al. (2001)] and [http://toxsci.oxfordjournals.org/content/84/2/249 Kitamura et al. (2005)]) indicate their potential harmfulness but further research is needed to fully access their impact on human health.

E. coli KRX carrying genes for the bisdA | bisdB fusion protein behind the medium strong constitutive promoter <partinfo>J23110</partinfo> with RBS <partinfo>B0034</partinfo> was cultivated at 30 °C for 36 h with LB-Medium containing 120 mg L^-1 BPA, BPF respectively BPS. The BPF and BPS concentration where determined with a HPLC using the same method as with BPA. Figure 2 shows the degradation of the respective bisphenol after 24 h of cultivation in percent.

Figure 2: Degradation of BPA, BPF and BPS after 24 h cultivation with E.coli KRX carrying genes for the bisdA | bisdB fusion protein behind the medium strong constitutive promoter <partinfo>J23110</partinfo> with RBS <partinfo>B0034</partinfo>. Cultivations were carried out at different temperatures in LB + Amp + bisphenol medium (starting concentration 120 mg L-1 BPA, BPF or BPS respectively) for 24 h in 300 mL shaking flasks without baffles with silicon plugs. Samples were taken at the end of the cultivation. Two biological replicates were analyzed. While BPA is fully degraded only a small fraction of BPF (~7%) and BPS (~3%) was degraded.

The results of the experiment indicate that the bisdA | bisdB fusion protein has a high specificity for the degradation of BPA. In addition it is possible that the decrease in BPF and BPS concentration is due to internalization of those bisphenols or a endogenous enzyme of E. coli KRX and not the bisdA | bisdB fusion protein was responsible. It can be assumed that false positive signals because of BPF or BPS present in a sample are unlikely.

BPA degradation with constructs containing FNR

Constructs containing FNR, BisdA and BisdB polycistronic <partinfo>K525551</partinfo>, FNR and a fusion protein of BisdA and BisdB <partinfo>K525582</partinfo> and a fusion protein of FNR, BisdA and BisdB <partinfo>K525560</partinfo> were tested for their ability to degrade BPA. Cultivations were done using the same conditions as with previous constructs. The results are shown in figure 3 below.

Figure 3: Degradation of BPA after 24 h cultivation with E.coli KRX carrying genes for FNR, BisdA and BisdB polycistronic <partinfo>K525551</partinfo>, FNR and a fusion protein of BisdA and BisdB <partinfo>K525582</partinfo> and a fusion protein of FNR, BisdA and BisdB <partinfo>K525560</partinfo> behind the medium strong constitutive promoter <partinfo>J23110</partinfo> with RBS <partinfo>B0034</partinfo>. Cultivations were carried out at 30 °C temperatures in LB + Amp + 120 mg L^-1 BPA for 24 h in 300 mL shaking flasks without baffles with silicon plugs. Samples were taken every 3 hours. Two to four biological replicates were analyzed. All constructs were able to degrade BPA, with the constructs containing fusion proteins outperforming the construct where all proteins were expressed polycistronic.

The results indicate that the BPA degradation was improved using fusion proteins compared to the completely polycistronic construct. Also the fusion protein of all three enzymes (<partinfo>K525560</partinfo>) involved in the degradation of BPA worked as intended which is quite impressive.

Comparsion of all constructs used for BPA degradation

Figure 4 gives an overview of the constructs employed for the degradation of BPA.

Figure 4: Schematic depiction of all constructs used for BPA degradation: [http://partsregistry.org/wiki/index.php/Part:BBa_K525512 BBa_K525511], <partinfo>K525515</partinfo>, <partinfo>K525551</partinfo>, [http://partsregistry.org/wiki/index.php/Part:BBa_K525582 BBa_K525580] and <partinfo>K525560</partinfo>. Constructs are divided into those without FNR (left) and with FNR (right).

All constructs were cultivated under the same conditions as described under Cultivations. BPA concentrations were measured using HPLC. Figure 5 shows the percentage of BPA degraded in 24 h of cultivation and the specific BPA degradation rate calculated with the regular model and in the case of <partinfo>K525560</partinfo> with the model for the fusion protein between FNR, BisdA and BisdB.

Figure 5: Results of the BPA degradation experiments with [http://partsregistry.org/wiki/index.php/Part:BBa_K525512 BBa_K525511], <partinfo>K525515</partinfo>, <partinfo>K525551</partinfo>, [http://partsregistry.org/wiki/index.php/Part:BBa_K525582 BBa_K525580] and <partinfo>K525560</partinfo> and the corresponding specific BPA degradation rates calculated with the model. The construct consisting only of BisdA shows almost no degradation activity, which was expected since BisdA alone can't facilitate the degradation of BPA. In comparsion the fusion protein of BisdA and BisdB degraded BPA considerbly more efficent than the polycistronic construct, indicating that the fusing of the two proteins improved its activity. Regarding the constructs containing FNR the completely polycistronic construct lags distinctively behind the other two constructs. Again, the fusing of the BisdA and BisdB proteins improved the BPA degradation activity. Using the fusion protein between FNR, BisdA and BisdB we measured the highest specific BPA degradation rate of all constructs. Allthough it has to be taken into account, that a different model had to be used for this particular construct.

We are especially impressed that the fusion protein of FNR, BisdA and BisdB (<partinfo>K525560</partinfo>) not only was capable to degrade BPA but also was the fastest construct we employed. This also contributes to the feasibility of our approach for a cell free BPA biosensor since in a cell free enviroment fusion proteins are beneficial when compared to their non-fused counter parts.

Model for the fusion protein between FNR, BisdA and BisdB

The cultivations and BPA degradation of E. coli KRX carrying a fusion protein consisting of ferredoxin-NADP⁺ oxidoreductase, ferredoxin_bisd and cytochrome P450_bisd differ from the cultivations with the other BPA degrading BioBricks. Thus, the model for these cultivations has to be adjusted to this behaviour. First of all, no diauxic growth is observed so the growth can be modelled more easily like

The BPA degradation starts when the imaginary substrate is depleted, like observed in the other cultivations with BPA degrading BioBricks. But it seems that the BPA degradation is getting slower with longer cultivation time. So it is modelled with a Monod-like term in which the specific BPA degradation rate is dependent from the BPA concentration:

with the maximal specific BPA degradation rate q_D,max and the constant K_D.

Figures 13 shows a comparison between modelled and measured data for cultivations with BPA degrading fusion protein E. coli. In Tab. 1 the parameters for the model are given, obtained by curve fitting the model to the data.

Comparison between modelled and measured data

Figure 9: Comparison between modelled (lines) and measured (dots) data for cultivations of E. coli KRX carrying BPA degrading BioBrick <partinfo>K525512</partinfo>. The BioBrick <partinfo>K525512</partinfo> (polycistronic bisdAB genes behind medium strong promoter) was cultivated seven times in E. coli KRX in LB + Amp + BPA medium at 30 °C, using 300 mL shaking flasks without baffles with silicon plugs. The BPA concentration (closed dots) and the cell density (open dots) is plotted against the cultivation time.

Figure 10: Comparison between modelled (lines) and measured (dots) data for cultivations of E. coli KRX carrying BPA degrading BioBrick <partinfo>K525517</partinfo>. The BioBrick <partinfo>K525517</partinfo> (fusion protein between BisdA and BisdB behind medium strong promoter) was cultivated five times in E. coli KRX in LB + Amp + BPA medium at 30 °C, using 300 mL shaking flasks without baffles with silicon plugs. The BPA concentration (closed dots) and the cell density (open dots) is plotted against the cultivation time.

Figure 11: Comparison between modelled (lines) and measured (dots) data for cultivations of E. coli KRX carrying BPA degrading BioBrick <partinfo>K525552</partinfo>. The BioBrick <partinfo>K525552</partinfo> (polycistronic fnr : bisdA : bisdB genes behind medium strong promoter) was cultivated four times in E. coli KRX in LB + Amp + BPA medium at 30 °C, using 300 mL shaking flasks without baffles with silicon plugs. The BPA concentration (closed dots) and the cell density (open dots) is plotted against the cultivation time.

Figure 12: Comparison between modelled (lines) and measured (dots) data for cultivations of E. coli KRX carrying BPA degrading BioBrick <partinfo>K525582</partinfo>. The BioBrick <partinfo>K525582</partinfo> (polycistronic fnr : BisdAB (fusion protein) genes behind medium strong promoter) was cultivated four times in E. coli KRX in LB + Amp + BPA medium at 30 °C, using 300 mL shaking flasks without baffles with silicon plugs. The BPA concentration (closed dots) and the cell density (open dots) is plotted against the cultivation time.

Figure 13: Comparison between modelled (lines) and measured (dots) data for cultivations of E. coli KRX carrying BPA degrading BioBrick <partinfo>K525562</partinfo>. The BioBrick <partinfo>K525562</partinfo> (fusion protein between FNR, BisdA and BisdB behind medium strong promoter) was cultivated four times in E. coli KRX in LB + Amp + BPA medium at 30 °C, using 300 mL shaking flasks without baffles with silicon plugs. The BPA concentration (closed dots) and the cell density (open dots) is plotted against the cultivation time.

The specific BPA degradation rate per cell q_D is about 50 % higher when using the fusion protein compared to the polycistronic bisdAB gene. This results in an average 9 hours faster, complete BPA degradation by E. coli carrying <partinfo>K525517</partinfo> compared to <partinfo>K525512</partinfo> as observed during our cultivations. The fusion protein between BisdA and BisdB improves the BPA degradation by E. coli. Introducing a polycistronic ferredoxin-NADP⁺ reductase gene into these systems does not lead to a higher specific BPA degradation rate. The rates are a bit lower, though. The effect that the BisdA | BisdB fusion protein is more efficient than BisdB when expressed polycistronically with BisdA can still be observed in this setup. The cultivations with the expression of the fusion protein FNR | BisdA | BisdB differ from the expression of the other BPA degrading BioBricks. The BPA degradation rate is concentration dependent which is typical for enzymatic reactions but was not observed in the other cultivations. In addition, the growth was faster and not diauxic. The maximal specific BPA degradation rate of this BioBrick is higher than the observed specific BPA degradation rates in the other cultivations. But due to the dependence of the BPA degradation from the BPA concentration the BPA degradation with this BioBrick is not as efficient and not complete. Anyway, the fact that this BioBrick is working is impressive.

S-layer

Purification???

Enzyme reaction fused to S-layer protein???

???

NAD⁺ detection

Identification of LigA

The gel bands after SDS-PAGE for His-tag purified LigA were analysed by MALDI-TOF MS and the comparison with the swissprot database clearly identified the purified protein as LigA (figure 2, table 2). As it is reported in literature the majority of LigA usually occurs in its adenylated form after extraction from E. coli. An indication for this is the double band on the right edge of figure 2. The proteins in both gel bands were identified as LigA from E. coli suggesting that gel band 3 is the adenylated form and gel band 4 is the deadenylated form.

Figure 2: SDS-PAGE analysis of LigA after His-tag purification as a preparation for MALDI-TOF MS analysis. The cleared lysate from <partinfo>BBa_K525710</partinfo> overexpressing E. coli KRX cells was loaded on a HisTrap™ FF crude column (1 mL), the protein was eluted with 100 mM imidazole and collected in 1 mL fractions. The framed and numbered fields indicate the gel bands which were analysed by MALDI-TOF MS. M: [http://www.fermentas.de/admin/images/media/labaid_pageruler_L7P.pdf prestained protein marker].

Table 2: Identification of LigA by MALDI-TOF MS. The values correspond to the framed and numbered gel bands in figure 2. The threshold for significance of the Mascot Score for MS is 63 and the one for MS/MS is 26. The MS-Coverage represents the sequence coverage of the investigated protein with the corresponding entry in the E. coli swissprot data base. The Sequence-Coverage shows the percentage of similarity to the translated BioBrick <partinfo>BBa_K525710</partinfo> (translation was perfomed with the [http://web.expasy.org/translate/ ExPASy Translate] tool).

Limit of detection

Figure 7: Initial velocity calibration curve for NAD⁺. The initial velocity, calculated from the average fluorescence enhancement rate in 200 s after NAD⁺ addition as shown in figure 5, is plotted versus the employed NAD⁺ concentration. Data is fitted with linear regression for NAD⁺ concentrations ranging from 0 to 200 nM (R² = 0.991, n = 4).

Selectivity test

The specifity of LigA for its substrate NAD⁺ is important in order to couple the NAD⁺ detection with investigated processes including NADH-dependent enzyme reactions. The verification of the selectivity is an important matter to exclude any unspecific reactions which might result in a NAD⁺-independent fluorescence enhancement. Hence, a selectivity test for LigA was performed with the analytes NADH, NADP⁺, NADPH, 3-ADAP (NAD⁺ with an exchanged functional group at the nicotinamide ring system), ATP and ADP, and the relative fluorescence enhancement rates in a NAD⁺ bioassay were compared with the one for NAD⁺ (figure 9).

Figure 9: LigA shows high selectivity for NAD⁺. The final concentration of all analytes was 100 nM. The responses were evaluated on the basis of the average fluorescence enhancement rate in a range of 200 s after addition of each analyte into the NAD⁺ bioassay. The dotted line marks the threshold indicating the intensity of background signal. All data are normalized to the NAD⁺ value (n = 2).

The negative control (H₂O) shows that there was a small background signal about 5 % of the signal that was produced by NAD⁺ when using 100 nM of analytes. This marks the threshold for fluorescence enhancement which is caused by the employed analyte. NADH, NADPH and ATP were similar to the negative control and can thereby seen as analytes that do not enable DNA ligation by LigA. Only the values for the three analytes NADP⁺, 3-APAD and ADP were above the threshold, but they were constantly below 10 % of the NAD⁺ signal. This leads to the suggestion that LigA ist highly selective for NAD⁺ even in presence of structurally similar analytes. This makes the NAD⁺ bioassay and the associated enzyme LigA suitable for investigating NADH-dependent enzyme reactions or measuring NAD⁺ in biological analyte mixtures such as cell lysates.

Coupled enzyme reaction

The coupling of the LigA-including NAD⁺ detection system was performed with a NADH-dependent enzymatic reaction which was in concrete the conversion of pyruvate to L-lactate by lactic acid dehydrogenase (LDH) from E. coli. For this the LDH reaction was done with an excess of NADH and various pyruvate concentrations. Afterwards, the reaction mix was transferred to a NAD⁺ bioassay and the fluorescence intensity was monitored. In figure 10 the normalized initial fluorescence enhancement rates for the employed pyruvate concentrations are indicated. The calculated initial velocity was then plotted against the pyruvate concentration (figure 11).

Figure 10: Fluorescence enhancement rate after addition of LDH reaction mix with various pyruvate concentrations into a LigA-dependent NAD⁺ bioassay. 50 ng/µL LDH was preincubated with 100 µM NADH and different pyruvate concentrations for 2 min at 37 °C. 1 µL of the LDH reaction mix was then added to a NAD⁺ bioassay, composed of a nicked DNA substrate (250 nM molecular beacon hybridized to a split target) and 6.5 ng/µL LigA, and the fluorescence intensity was monitored. For illustration the fluorescence intensity is normalized for each measurement series to the first measured fluorescence value after LDH reaction mix addition (n = 4).

Figure 11: Initial velocity calibration curve for pyruvate. The initial velocity, calculated from the average fluorescence enhancement rate in 100 s after LDH reaction mix addition into a LigA-dependent NAD⁺ bioassay, is plotted versus the employed pyruvate concentration. Data is fitted with linear regression for pyruvate concentrations ranging from 0 to 10 µM (R² = 0.951, n = 4).

The addition of the LDH reaction mix resulted in a characteristic fluorescence enhancement rate depending on the employed pyruvate concentration. The existing correlation between both parameters seemed to be a linear. That the signal for 0 µM pyruvate was remarkably high could be the result of an pyruvate-independent transfer of electrons from NADH to the active site histidine of LDH under formation of NAD⁺. The limit of detection for pyruvate seemed to be near 1 µM pyruvate. That this value was not in nano molarity scale is firstly caused by 100-fold dilution of the LDH reaction mix after addition to the NAD⁺ bioassay and secondly by the fact that LDH does not necessarily converte 100 % of the pyruvate to L-lactate. Nevertheless, for the LigA based NAD⁺ detection system it has been proven that it can be coupled to NADH-dependent enzymatic reactions. This makes the NAD⁺ bioassay suitable for a wide range of biological studies dealing with the ubiquitous cofactors NADH/NAD⁺.

Molecular cloning

Because DNA ligase from E. coli is commercially aquirable for cloning purposes LigA (<partinfo>BBa_K525710</partinfo>) was tested whether it is also suitable for molecular cloning procedures. Hence, cloning of rfp into a pSB1C3 vector was performed with LigA. Ligation was done with NAD⁺ bioassay buffer containing 10 mM NAD⁺ at 37 °C or 22 °C and the vectors were transformed into E. coli KRX by electroporation. After growth over night at 37 °C on petri dishes the results were documented (figure 12).

Figure 12: rfp cloning without any ligase (A) or with ligation by DNA ligase from E. coli (LigA) at 37 °C (B) and 22 °C (C). The picture shows the grown colonies after transformation of a control sample (rfp and cut pSB1C3) or ligation sample (rfp and cut pSB1C3 with LigA) into E. coli KRX.

By analyzing the number of colony forming units and color of the colonies the ligation efficiency of LigA should be assessable. As shown for the sample without any ligase there were only a few clones able to grow on chloramphenicol supplemented LB medium. A much bigger number of clones was observed when ligation by LigA was performed at 22 °C. Furthermore, there were a lot of red colonies indicating positive clones. Deductively, LigA might be useful for (large-scale) molecular cloning procedures which could be beneficial because of its easy production in E. coli and simple purification.