Team:KU Leuven/Project/Glucosemodel/MeS

Methyl salicylate (or wintergreen oil) is a plant hormone. It is converted from chorismate with salicylate as an intermediate. In this page we will discuss the pathway that leads up to methyl salicylate and the biobricks that we made. We will end with characterisation of these biobricks using a smell test and an SDS-PAGE.

The conversion of Ery4P and PEP to DAHP (3-deoxy-D-arabinoheptulosonate 7-phosphate) is catalysed by DAHP synthase, an enzyme which exists in three isoforms, coded by the genes aroF, aroG and aroH. These isoforms are selectively inhibited by tyrosine, phenylalanine and tryptophan, respectively. In E. coli, about 80%, 20% and 1% of the enzyme activities are contributed by the DAHPS phenotypes of the aroG, aroF and aroH products, respectively (Ikeda et al., 2006).
This means that when there is enough phenylalanine produced by the cell, or enough present in the medium, the production of DAHP, and hence chorismate, will stop by inhibition of the aro genes. The production of phenylalanine will stop as well by inhibition of pheA, coding for chorismate mutase/prephenate dehydratase.
The conversion from chorismate to methyl salicylate starts with a conversion to isochorismate, catalysed by the enzyme isochorismate synthase, encoded by the pchA gene. This isochorismate is directly converted to salicylate by isochorismate pyruvate/lyase, encoded by the pchB gene. This conversion happens immediately, since the pchA and pchB gene form the pchBA operon which is always transcribed to one mRNA. Conversion of salicylate to methyl salicylate is catalysed by S-adenosylmethionine-dependent methyltransferase, encoded by bsmt1 (Gaille et al., 2003).

The iGEM team of MIT 2006 already constructed a Biobrick (BBa_J45700) encoding the pchBA and bsmt1 genes, so introducing this brick should induce the production of methyl salicylate, since chorismate is a common metabolite in E. coli. The MIT team discovered however that there was almost no methyl salicylate production observed, only when salicylic acid was added to the medium. Our experiments with this Biobrick confirmed this lack of methyl salicylate production (see the smell test in the MeS journal). There had to be something wrong with the conversion of chorismate to salicylate. One possibility is that there is something wrong with the enzymes produced by the pchBA genes. Another one is that there is a lack of chorismate present in the cell. We have attempted to increase the productivity of the MIT 2006 brick by using stronger ribosome binding sites. Also, we replaced the lac promoter by a TetR-repressible promoter, so the lacI produced in our system doesn't interfere with the transcription of pchBA..

Since chorismate is the precursor for three amino acids in E. coli, we believe that there is not much of the chorismate left to be converted to methyl salicylate, since the cell needs to keep producing a steady amount of essential amino acids.
To overcome this problem, we looked at a study by Sun et al. (2011), in which a synthetic pathway was introduced for the production of mandelic acid from chorismate. They achieved this by deleting different genes encoding enzymes that catalyse competing pathways, as well as by introducing a feedback-insensitive DAHP synthase mutant to increase the carbon flow down the shikimate pathway. This last method gave us inspiration to overcome our own problem.
If we look at the pathway in Figure 2, we can conclude that if the amino acids are present in the medium, the conversion of chorismate to these amino acids will be inhibited allosterically, as wel as the production of DAHP. Our plan is to mutate the aroG gene in a manner that the enzymatic function still remains, but that it is insensitive to allosteric inhibition by phenylalanine. We only mutate the aroG gene since this isoform is responsible for 80% of DAHP synthase activity (Hu et al., 2003).

To functionally test our Methyl salicylate production brick (BBa_K1060003), we prepared 68 microcentrifuge tubes with E.coli expressing the MeS construct to conduct a blind smell test. This experiment consisted of 4 sets, different in incubation time and temperature. The first set was grown at room temperature for 24 hours, the second was incubated at 37ºC for 24 hours, the third was grown at 37ºC for 8 hours and the last set was grown at room temperature for 8 hours. In each set, we inoculated our brick BBa_K1060003 with IPTG, IPTG + salicylate (concentration range from 0.01mM to 1mM), and IPTG + chorismate (concentration range from 0.01mM to 1mM), the same for MIT 2006 brick BBa_J45700. In addition, the same combinations of additives were introduced in parallel for our negative controls (non-MeS producing E. Coli).
We asked 11 people to participate in this blind test and calculated the percentage of people who smelled MeS. We can conclude that the MeS/wintergreen smell was stronger after 24 hours incubation time compared to 8 hours. Growing the bacteria at 37 degrees also produced a more pronounced smell than those at room temperature. What’s more, in the microcentrifuge tubes where 0.1mM or 1mM salicylate was added, the smell became more obvious both after 8 and 24hrs of growth. The positive effect of adding chorismate became noticable after 24 hours. The reason why chorismate requires more time to show an effect may be that chorismate is the very first precursor of the methyl salicylate production pathway, whereas the salicylate is an intermediate closer to the end product. As such, the impact of salicylate on the end product will become noticeable sooner. Most importantly, our brick worked in terms of producing methyl salicylate.
The results are shown below:

Similar to our set-up for the EBF synthase protein expression, we tested the MIT biobrick (BBa_J45700) and our novel MeS construct (BBa_K1060003) via protein expression. We transformed our constructs in an E.coli expression strain, grew them at various temperatures (room temperature and 37 degrees celcius) and induced expression with increasing amounts of IPTG. We also added salicylate or chorismate to the growth medium in an attempt to increase MeS production.
The figure shows results obtained with the BBa_J45700 brick. Our biobrick (BBa_K1060003) showed similar results (data not shown). Increasing the amount of IPTG did not influence the protein expression profile (compare lanes a-d or lanes e-h) but we do see some bands in the lanes a-d which we cannot see in lanes e-f (eg a band just above 55 kDa). Nonetheless, this observation can be verified with lower amounts of the protein extracts. Nonetheless, our smell test would suggest that the MeS brick does work. Hence, we need a more sensitive approach to identify the protein production of the MeS brick. Possible approaches would be via classic western blot experiment or a GC-MS set-up.