Results

Future Applications

Upon expressing pterostilbene, our team decided to investigate future avenues our project could take once pterostilbene has been successfully synthesised. For future applications, our team investigated the use of malonyl-CoA, in silico modelling as well as future protocols we can utilise to close the loop and bring pterostilbene into the forefront of the Alzheimer’s disease (AD) drug market.

Malonyl-CoA and sRNA

As previously mentioned, L-tyrosine was at first the most obvious limiting factor for our project. Many papers that try to produce pterostilbene in E. coli implement L-tyrosine overproducing strains, to increase pterostilbene yields (Heo et al., 2017; Yan et al., 2021). Upon further research into literature, however, we have identified that there are other compounds to consider to increase pterostilbene yield. One important compound is malonyl-CoA. To create one mole of resveratrol, three moles of malonyl-CoA are required (Lim, Fowler, Hueller, Schaffer & Koffas, 2011). This is a large task for the E. coli when one considers the relatively low abundance of malonyl-CoA within an E. coli cell at any given time (Wu et al., 2017). Notably, neither the paper from which we obtained the plasmid system, nor the paper from which we obtained our mutated enzymes use E. coli that have increased malonyl-CoA levels. Through literature we identified a series of methods to counteract malonyl-CoA scarcity within E. coli.

Cerulenin is an antifungal drug that has been shown to inhibit malonyl-CoA consumption via inhibiting the FabF and FabH genes involved in fatty acid synthesis (Lim et al., 2011). Notably, the fatty acid cycle is the only pathway in E. coli that actually consumes malonyl-CoA which makes the Fab genes ideal targets for increasing intracellular malonyl-CoA levels (Wu et al., 2017). One paper identified a concentration of 200 uM as sufficient to increase yields of pinosylvin by 18-20-fold in E. coli (Van Summeren-Wesenhagen & Marienhagen, 2015). However, one issue with cerulenin is its high cost which has been remarked on by researchers in past papers (Lim et al., 2011). Additionally, one paper notes a potential risk in E. coli toxicity which makes this not an ideal candidate for pterostilbene production (Wu, Yu, Du, Zhou, & Chen, 2014).

Due to these risks we identified in cerulenin, we investigated a potentially better means of inhibiting the Fab genes using small regulatory RNAs (sRNA). From our search, we found several papers that attempt to increase malonyl-CoA concentrations through the use of RNAi (Yang et al., 2015; Sun et al., 2018; Liang et al., 2021). From this literature we identified one gene that we desired to inhibit to increase malonyl-CoA concentration – the FabD gene. FabD encodes a malonyl-CoA-ACP transacylase which takes malonyl-CoA from the cytoplasm and funnels it into the fatty-acid cycle. Since it is the first enzyme in the fatty acid cycle, and is the only one directly interacting with malonyl-CoA, we identified it as an ideal candidate for RNAi. We decided to use the MicC scaffold as part of our RNAi system. MicC is a non-coding RNA that targets the OmpC gene which encodes for “outer membrane protein C”. The MicC scaffold has been designed by researchers for inhibition of other genes with over 90% RNA inhibition (Yoo et al., 2013). It works by targeting a 24-bp region starting from the start codon (ATG) of a sequence, competing for binding of the ribosome and thereby preventing translation. The MicC scaffold has been used before in iGEM by Team UT-Tokyo in 2013. We constructed our own version of the MicC scaffold to target FabD on SnapGene and as a future implementation of our project, we plan to use this compound to increase levels of malonyl-CoA within a cell (Figure 3).

The antiFabD-MicC sRNA would be assembled into a plasmid with an origin of replication compatible with pJUMP49-2A (pBR322 origin) which is used to express the genes catalyzing pterostilbene synthesis. One candidate is the p15A origin in pJUMP48-2A. This plasmid would have the sRNA expression cassette depicted in figure 3 cloned between the BioBrick prefix and suffix using EcoRI and PstI. This construct can then be co-transformed with our four genes to allow for production of pterostilbene with an increased malonyl-CoA pool.

Investigating Pterostilbene Targets in Activated Microglia

With pterostilbene and other structurally similar polyphenols such as resveratrol having been continuously studied for their therapeutic capabilities, many potential molecular targets have been identified (Rimando et al., 2005; Perecko et al., 2010; Chang et al., 2012; Zhang et al., 2022; Yi et al., 2022), especially considering the small size hence reduced specificity of many of these compounds. Therefore, it is important to conduct relevant assays for pterostilbene target identification not only to verify the reproducibility of these studies but also to understand the mechanism and pathways by which pterostilbene can attenuate neuroinflammation in AD.

Stable isotope labelling by amino acids in cell culture (SILAC), an easy-to-implement method providing high-quality quantitative data, can be used to identify pterostilbene microglial targets. By firstly growing cells in both "light" and "heavy" amino acid media, then adding "light" cell lysates to pterostilbene-immobilised affinity columns whereas "heavy" cell lysates are pre-treated (so any targets become pre-bound) with pterostilbene before affinity chromatography, it is possible to compare the ratio of "light" to "heavy"/control protein levels (Ong & Mann, 2007). By verifying targeted pathways regulating microglia activity, further investigation can then be conducted on therapeutic efficacy and cell viability aspects of the therapeutic.

Effects of Pterostilbene on Cell Viability

Cell viability is a crucial consideration in the development of any therapeutic, and can be measured through a variety of assays analysing different parameters from metabolic rates to apoptosis of treated cells. As such, cell viability assays have been conducted in the context of the therapeutic potential of pterostilbene (Hou et al., 2015; Fu et al., 2016; Meng et al., 2019). Particularly, ATP assays are commonly used as cell viability measures and have been reported to be faster, more sensitive, and less prone to artefacts than many other methods (Riss et al., 2013).

Pterostilbene Therapeutic Efficacy

Particularly, we would focus on testing the therapeutic potential of pterostilbene targeting microglial production of IL-B and IL-18.

The initial phase of the project encompassed designing a genetic construct capable of producing pterostilbene in E. coli. Furthermore, upon successful production of pterostilbene, it is possible to further explore the potential therapeutic applications of pterostilbene to target different pathophysiological components of AD, such as the NLRP3 neuroinflammatory pathway.

In AD, amyloid-B induced neuroinflammation in microglia is caused by the activation of NLRP3/caspase-1 (Figure 4), a protein complex that induces the secretion of several inflammatory cytokines namely IL-B and IL-18 (Li et al., 2018) as seen in the figure below. Thus, AD is associated with a high accumulation of IL-1B around amyloid plaques (Zheng et al., 2016).

Pterostilbene has been shown to decrease the amount of activated microglial cells and IL-1B mRNA expression, an inflammation-inducing interleukin in the treatment of ischemic strokes (Liu et al., 2018). Moreover, pterostilbene can be explored as a potential therapeutic to treat AD by inhibiting microglial production of IL-1B and IL-18 in the NLRP3/caspase-1 neuroinflammatory pathway, as it has been shown to aid similar neuroinflammatory pathways involving the same interleukins IL-1B and IL-18 in other diseases such as ischemic strokes.

In order to test the therapeutic potential of pterostilbene in targeting microglial production of IL-B and IL-18, we propose using 2 colorimetric sandwich ELISA using indirect detection with a biotinylated antibody and streptavidin–Horseradish peroxidase using TMB (tetramethyl benzene) as the Horseradish peroxidase substrate (Figure 5). The ELISA sandwich assay works by using a capture antibody (blue) that binds to compound of interest in this case IL-1B and IL-18 (yellow target antigen). Simultaneously, a second antibody (blue) binds to the compound of interest and is also bound to a third antibody which will catalyse the reaction of the horseradish peroxidase, inducing a colorimetrically measurable change.

There are several steps required prior to performing the two sets of ELISA assays (Li et al, 2018):

1. The first step prior to performing the ELISA assay is culturing a BV-2 cell line – BV-2 cells are a well characterised model system of microglia. The BV-2 culture must be tested using an MTT assay to measure cell viability.
2. Once cell viability of the BV-2 cell line is confirmed, cells need to be supplied with Fibrillar amyloid-B peptides (5 um) in order to activate the BV2-cell microglial IL-1B and IL-18 secretion (Jana et al., 2009).
3. Pterostilbene (10 um) must also be added to the cell cultures in order to evaluate its effect on microglial production. Prepare additional control culture seeds without pterostilbene.
4. The cells are now incubated and cell supernatant should be extracted to begin detection of IL-1B and IL-18 using ELISA assay.
5. ELISA protocol for IL-18 (Thermo Fisher) and ELISA protocol for IL-1B (Thermo Fisher) can now be performed on cell supernatants.

Modelling

Within our project we focused on creating a model to suggest ways the bacterial expression system for the synthesis of pterostilbene can be improved by accounting for metabolic stress, altering promoter strengths and plasmid copy number.

This iteration of the model is built using the Michaelis Menten equations as its base. It consists of 15 ODEs each of which tracks the concentration of each enzyme and reagent. The model outputs final concentrations of each reagent accounting for 256 different combinations of 4 possible promoter strengths for each gene, as well as a range of plasmid copy numbers ranging from 5 to 300. This data is represented by the 3D surface model on the wiki which represents the maximum concentration of pterostilbene for each iteration. The model comprises 3 functions, one of which is an open source permutation MATLAB script sources externally, and one main script that outputs the final model. The other two functions include “Pterostilbene_rates_ver_2308”, which includes the 15 rate equations and “Pterostilbene_Production_Sim_function”, which carries out the ode solver. Included on the wiki is also the script “Pterostilbene_Production_Sim_ver3008” which only carries out one iteration at a time. It was used in the engineering process to refine the model and output concentrations of each reagent and enzyme specific to the promoter strength/copy number combination. It includes requests for input from the user such as initial concentrations, plasmid copy number and promoter strengths.

Unfortunately due to time constraints, the current model is unable to fully account for metabolic strain and therefore incorrectly recommends the highest possible plasmid copy number. We have put in place a simple assumption regarding the available L-tyrosine in place of this. However, given more time, by applying principles regarding cellular trade-off the model would be able to provide data that could be extremely useful by allowing the user to identify the best relative combination of promoter strengths and plasmid copy number. This model would be further extended to account for the ribosomal binding site, gene length and half-life of protein.