From 2014.igem.org

Revision as of 18:46, 5 October 2014 by Mbennett (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)

Exeter | ERASE

The Solution: Old Yellow Enzymes

From our wider research, it was clear that there was a serious need, both from an environmental and humanitarian perspective, to develop solutions to rapid identification and long-term degradation of explosives. Thus, we initiated research into the microbial TNT and Nitroglycerin attacking enzymes that have been discovered and whether other biobricks or industrial products had been generated.

Several microbial enzymes have the ability to catalyse the break-down of nitroaromatic compounds such as TNT and NG. These enzymes fall into two main families:

Oxygen-Insensitive Nitroreductases: These enzymes sequentially perform two-electron reductions of nitro groups. They typically contain Flavin mononucleotides FMN which they use, along with NADPH as a cofactor and electron donor. Examples include the nitroreductases NfsA and NfsB from Escherichia coli¹, PnrA and PnrB from Pseudomonas putida² , and NitA and NitB from Clostridium acetobutylicum³.
Old Yellow Enzymes (OYE): The physiological function of this family of NADPH dehydrogenases is not yet well established; however, they are often associated with nitroaromatic compound reduction. Within the OYE family, two types of enzymes have been described:

Type I hydride transferases, which, like the oxygen-insensitive nitroreductases above, reduce the nitroaromatic compounds to hydroxylamine derivatives

Type II hydride transferases, which catalyse a nucleophilic attack on the aromatic ring of TNT^4,5

Of bacterial OYE family members, those that are the best characterized are XenA - XenF of P. putida KT24406 XenB from P. fluorescens⁷, PETN reductase from Enterobacter cloacae PB28, NemA reductase from E. coli⁹ and YqjM from Bacillus subtilis¹⁰.

In 2009 the Edinburgh iGEM team developed the concept of a Nitrate/Nitrite biosensor which could be used to detect TNT. They generated a biobrick (BBa_K216006) for the gene onr (organic nitrate reductase) that encodes pentaerythritol tetranitrate (PETN) reductase to function as a TNT degrader. However, neither were completely characterised.

We sought to find new, mostly uncharacterised, enzymes to function as additional solutions to offer alternative mechanisms which may be more suited to particular problems. Given that the PETN reductase was a member of the Old Yellow Enzyme family, we decided that we would target additional members of the OYE group to help expand the both the range of enzymes that can be utilised to degrade explosives and the range of UXO chemicals that could be degraded to those that contain nitroglycerin.

Our initial shortlist contained the following the proteins: XenA, XenB, NemA and YqjM''' . From these we selected two to examine over the summer. These were XenB and NemA.

XenB (BBa_K1398001)

XenB is an NADH-dependent flavoprotein (Xenobiotic Reductase B) from the soil bacterium ''Pseudomonas fluorescens''. We chose XenB because:

XenB may serve a dual purpose as a Nitroglycerin and TNT degrading enzyme (see below).
The phylogeny of OYE members¹¹ suggests that XenB is similar in sequence to the previously submitted PETN reductase and is comparatively well understood.
Heterologous expression of XenB from P. fluorescens has been previously demonstrated in E. coli DH5α¹² enhancing our chances of successful expression.

Degradation of TNT

Pak et al. (2000)¹⁴ reported the purification of the NADH-dependent flavoprotein oxidoreductase xenobiotic reductase B (XenB) from Pseudomonas fluorescens.

Those authors provide a guide to some of the degradative reactions of XenB from P. fluorescens. This indicates that XenB may catalysed the reduction of TNT, either by adding a hydride to the aromatic ring and forming a dihydride Meisenheimer complex¹⁵ or by catalysing the reduction of nitro groups directly (Figure 1). However several unidentified products and proposed intermediates indicate that this process is yet to be fully elucidated.

Figure 1: Diagram depicting the potential reductive steps catalysed by XenB. (Adapted from Pak et al. ¹⁴)

XenB from Pseudomonas fluorescens is not the only XenB enzyme that exists. A more commonly studied XenB enzyme is found in Pseudomonas putida. This XenB shares 88% identity to our chosen XenB from P. fluorescens¹⁶.

XenB from P. putida (as well as other xenobiotic reductases) demonstrate type II hydride transferase activity against TNT. Research into these xenobiotic reductases provided a greater understanding of the potential products and intermediates generated in the reactions with TNT (Figure 3). Using figure 1 as a model, in the mechanisms suggested by Pak et al. (2000) (above) it seems likely that the unknown m/z 196 compound was 4ADNT and that the proposed bridge product (m/z= 376) is instead linked by an amine group bridge (Figure 2).

Figure 2: Diagram depicting the known reductive steps catalysed by XenB. (Adapted from van Dillewijn et al. ¹⁷)

We were unable to find more recent attempts to characterise the mechanism of XenB from P. fluorescens. However, given that these mechanisms for P. putida Xenobiotic reductase proteins are similar to the mechanisms of the type II hydride transferase family¹⁸ it suggests we are more likely to find the amine bridged products in Figure 2.

Degradation of Nitroglycerin

XenB also displays an apparent capacity to denitrify Nitroglycerin (NG). It has been demonstrated that XenB catalyses the NADH-dependent cleavage of nitro groups from NG, releasing nitrite (Figure 3). P. fluorescens XenB also exhibits five-fold regioselectivity for removal of the central nitro group from NG compared to P. putida XenB.

Figure 3: Simplistic diagram depicting the chemistry involved in Glycerol Trinitrate transformation by XenB19.

As nitrate is released from the degradation of Nitroglycerin there is a suggestion that construct could be built upon to work simultaneously as an explosive degrader and soil enricher, thus taking an environmental toxicant and converting it into an environmental benefit.

The XenB Construct: (BBa_K1398001)

The coding sequence for P. fluorescens XenB has been deposited within GenBank (accession no. AF154062)²⁰ and this forms the functional unit of our construct (BBa_K1398001) (Figure 4):

Figure 4: The XenB construct depicting the constituent parts of the construct.

The construct contains the coding sequence for XenB (BBa_K1398000). Expression of XenB is driven by a Lactose-inducible promoter (BBa_R0010) coupled with a strong RBS (BBa_B0034). The construct is terminated using a double terminator made up of BBa_B0010 and BBa_B0012.

NemA (BBa_K1398002)

The second of our enzymes is the detoxification enzyme N-Ethylmaleimide (NEM) reductase from Escherichia coli encoded by the gene nemA²¹). Often referred to simply as NemA, the flavin-dependent NEM Reductase was chosen for several reasons:

Firstly, like XenB, NemA has a dual capacity to degrade both TNT and Nitroglycerin.
NemA has a high degree of homology (87% identical) to pentaerythritol tetranitrate (PETN) reductase (BBa_K216006) from Enterobacter cloacae²². The E. cloacae PETN reductases and E. coli NEM reductase showed broadly similar activity profiles, with high activity against nitrate esters²³, However, NemA was shown to have a higher substrate preference for PETN and Nitroglycerin than PETN reductase, the uncharacterised gene of which was submitted by the iGEM team from Edinburgh in 2009 William’s et al. (2004)²⁴.
Finally, NemA was found to already exist in E.coli, albeit at a low expression level. This suggested to us a capability to be expressed successfully. However, we also realised this meant that we would more likely to face natural control mechanisms that could silence the activity of the enzyme if it is overexpressed. It also meant that we would need to assess the natural tolerance and expression levels of our E.coli DH5α strain to the explosives to ensure our engineered enzyme provided enhanced tolerance to TNT and NG compared to the wild-type.

Through a comparative study of PETN, NemA, Morphinone reductase and the yeast OYE, the TNT mechanism for the similar PETN reductase was detailed. Due to the sequence similarity, we are using this as a model for the pathways involved in anaerobic degradation of TNT by NemA (Figure 5).

Figure 5: Diagram depicting the potential mechanistic steps in the transformation of TNT by PETN reductase. (Adapted from Rylott et al. ²⁵ and Williams et al. ²⁶)

In addition to degradation of TNT and NG, NemA may also function as an efficient chromate reductase to remediate hexavalent chromium, a serious and widespread environmental pollutant²⁷, and has been suggested to have a role in bleach detoxification.

The NemA construct (BBa_K1398003)

With the nemA DNA sequence identified(28) we were able to design our construct to evaluate NemA as follows (Figure 6):

Figure 6: The NemA construct depicting the constituent parts of the construct.

The construct contains the coding sequence for NemA (BBa_K1398002), an enzyme involved in the degradation of toxic compounds for their reuse in nitrogen metabolism. The construct also contains a Lactose-inducible promoter (BBa_R0010), a strong RBS (BBa_B0034) and a double terminator made up of BBa_B0010 and BBa_B0012). The protein has been codon-optimised for expression in E. coli.

The NemR recognition promoter (BBa_K1398008)

One of the major unsolved problems left by the 2009 Edinburgh iGEM team concerning the design of a TNT degrader using PETN reductase was the issue of how the TNT was to be detected to generate a gene expression response.

https://2009.igem.org/Team:Edinburgh/biology%28results%29

https://2009.igem.org/Team:Edinburgh/biology%28biobricks%29

While their design and parts demonstrated a staggering amount of work and understanding, few of their submitted parts were fully characterised and their proposed construct was complicated; relying on several uncharacterised potential parts working together in unison.

Figure 7: TNT sensing pathway developed by the Edinburgh 2009 iGEM team. https://2009.igem.org/Team:Edinburgh/biology%28tntsensing%29 TNT binds to TNT.R1 in the periplasm. The TNT-TNT.R1 complex induces a conformational change in the Trg-EnvZ (Trz) fusion protein. Trg-EnvZ autophosphorylates and subsequently phosphorylates ompR. Phosphorylated ompR activates transcription.

With our supervisor’s constant reminders that ‘Simple is best’ we decided to see if we could discover and develop a vastly simpler system that would be easier to engineer and conceptually extrapolate to other chemicals. After the discovery that the NemA enzyme could already be expressed at a very low level in E.coli, it was hypothesised that ''E.coli'' should have a gene regulatory molecule which controlled the nemA operon enabling its expression only in environments where TNT was present. In E. coli, nemA is located downstream of a gene encoding a transcription factor known as NemR, previously entitled YdhM29. Disruption to the gene encoding NemR results in a decrease in nemA expression, both in the basal and induced states, which indicated that the nemA gene and NemR form a single, connected operon. Umezawa et al. (2008) proposed that the function of NemR is as a redox-operated transcriptional repressor of the nemA gene. NemR is proposed to remain bound to the operon and, when it binds TNT in the environment, undergo a conformational change to expose the promoter. The region of the promoter sequence to which NemR binds (the ‘NemR recognition box) has also been identified³⁰ (Figure 8).

Figure 8: Diagram depicting the sequence of the ‘NemR box’. Adapted from Umezawa et al.³¹

On the assumption that TNT was able to naturally diffuse into the E.coli cell, we hypothesised that if we designed a promoter containing this binding site gene sequence, the promoter would therefore become inducible by the addition of TNT through the use of the NemR protein which is naturally found in E.coli and would be a vastly simpler TNT-detecting gene expression system as it would rely on the success of considerably fewer engineered parts. It has been shown that NemR also responds to cysteine-modifying electrophiles, including NEM, showdomycin, and, more weakly, iodoacetamide ³³ . The repressor function of NemR was inactivated through the addition of Cysteine modifying reagents. Furthermore, based upon a predicted 3D structure (Figure 9), it has been proposed that NemR utilises reversible oxidation of a conserved Cys-106 residue as the signal for confirmation change and dissociation from the DNA strand thus activating gene expression³⁴.

Figure 9: 3D molecular structure of the E. coli NemR monomer. Cysteine residues are in orange, conserved Cys-106 in red, and DNA binding helices in green³³.

Finally, it has been indicated that the NemR protein was responsive to Hypochlorous acid (HOCl), the active component of household bleach. Addition of bleach resulted in the expression of two detoxifying enzymes for bleach: glyoxalase I and NemA³⁵.

Thus, it is likely that our vastly simplified promoter, while designed and refined by us to respond to TNT, would actually have a more ubiquitous role in responding to a range of important electrophiles and thus is ripe for characterisation not only by us but also for future iGEM teams.

The NemR promoter construct (BBa_K1398007)

The development of the NemR promoter was originally as part of the design of a kill switch we intended to implement. However, it was clear that impressive kill switches have already been widely developed and used. We decided instead to focus on a NemR switch that could be used to regulate them such that the bacterium would die in the absence of TNT. When we designed the NemR promoter, we debated where the ‘NemR box’ would be inserted into the construct sequence to enable successful gene repression and expression. We eventually opted for the NemR box to be placed in-between the -12 and -33 region of the promoter as this region appeared reasonably conserved across the Anderson promoter group available to us on the registry: http://parts.igem.org/Promoters/Catalog/Anderson. Thus we designed the NemR construct, BBa_K1398007 as follows:

Figure 10: The BBa NemR construct we designed

The construct begins with the synthetic promoter NemR, which combines a high-expression promoter with the NemR recognition box (BBa_K1398008). It is followed by a strong RBS (BBa_B0034), the fluorescent reporter iLOV (BBa_K660004), a double STOP codon and a double terminator made up of BBa_B0010 and BBa_B0012.

The addition of the iLOV reporter gene was also to help further characterise part BBa_K660004 submitted into the iGEM database in 2011 by the team from Glasgow: https://2011.igem.org/Team:Glasgow/LOV2. Given that this promoter would respond to TNT, this means we are also generating a biosensor for TNT.

While hybrid promoters have been designed before upon similar ideas, this is usually done to develop a switch with additional functionality to respond to the insertion of an additional engineered system (eg. For example, it is common that we utilise dual Lac and IPTG inducible promoters in the design of gene circuits). Our promoter construct’s incredibly simple design compared to previous teams proposes the idea that by identifying the binding site of a repressor protein, anyone could theoretically make a promoter of any level of expression strength that is responsive to any chemical stimulus found in nature, and likely many that are not.

We aim to demonstrate this future possibility through the creation of our TNT-responsive NemR-sensitive promoter. Any strong output could therefore be inhibited or, alternatively, a weakly expressed gene could have its rate of expression enhanced by identifying the correct gene sequence. Alternatively, any construct could theoretically could be made to respond to a very specific chemical signal, such as TNT. If we are able to prove that our TNT-specific promoter is successful, it will provide the proof of concept to apply this theory to millions of other chemical signals which gene circuits could be designed to be receptive to.

----

References: