Proof of Concept


Introduction

To provide proof of concept for our aim of producing pterostilbene for use as a therapeutic in treating Alzheimer’s disease (AD), our goal was to create a system capable of producing pterostilbene in E. coli at therapeutically relevant concentrations to effectively tackle the neuroinflammatory hallmark of the disease at its mild to moderate stages.

For this phase of our project, we focused on demonstrating the feasibility of our constructed system for producing pterostilbene rather than reinforcing its therapeutic potential as already indicated in the literature (Chang et al., 2012; Kosuru et al., 2016; Lange and Li, 2018; Arbo et al., 2020). Nevertheless, we have outlined potential future plans to conduct therapeutics-oriented assays for target identification, therapeutic efficacy, and cell viability and cytotoxicity of pterostilbene in the brain.

Lab Overview

For our first iteration, we based our expression system on previous literature on pterostilbene biosynthesis as well as on our communications with authors of these studies, whereby a strong promoter was used for the expression of all four enzymes in a synthetic operon system, assembled into a low-copy number plasmid as further suggested by our modelling (Heo et al., 2017; Yan et al., 2021; Gerngross et al., 2022). Since the linear biosynthetic pathway we incorporated into E. coli consists of four enzymes (RgTAL, At4CL1, VvSTS, and VvROMT mutants) to produce pterostilbene from the precursor tyrosine (Fig. 1), it is important to consider the potential metabolic stress imposed by this system from the depletion of precursors and cofactors, especially given that these considerations are often overlooked by current studies in the field. Therefore, we devised cost-effective approaches such as supplementation and future RNAi strategies to increase the intracellular pool of L-tyrosine, L-methionine, and malonyl-CoA. Moreover, we used dry lab modelling to predict the accumulation of potentially harmful intermediates (e.g. 4-coumarate), as well as pterostilbene yield and recombinant proteins concentrations. Specifically, we employed a multimodular pathway-balancing strategy for metabolic flux optimisation, which involved the fine-tuning of gene expression levels through varying plasmid copy numbers and promoter strengths, to minimise the metabolic stress imposed by the system while increasing pterostilbene yields (Li et al., 2022). Therefore, to facilitate future pterostilbene production iterations implementing the optimised metabolic flux system, we used type IIS assembly whereby four main parts (promoter, RBS, CDS, terminator) are individually constructed and assembled into a transcriptional unit (and further into multiple transcriptional units constructs). This would then be particularly useful for exclusively replacing the promoter to achieve an increased or decreased strength according to the desired concentrations for each enzyme. With this year’s iGEM Distribution Kit introducing a new type IIS assembly system, Joint Universal Modular Plasmids (JUMP) (Valenzuela-Ortega & French, 2021), we thoroughly investigated how we could use JUMP to construct our expression system in E. coli through our communications with Dr Valenzuela-Ortega and our detailed guide on how future iGEM teams can use the JUMP system for vast applications in synthetic biology.

Pterostilbene Biosynthetic Pathway from an L-Tyrosine Precursor
Figure 1. Pterostilbene Biosynthetic Pathway From an L-Tyrosine Precursor. TAL, tyrosine ammonia-lyase; 4CL1, 4-coumarate:CoA ligase 1; STS, stilbene synthase; ROMT, resveratrol O-methyltransferase. Adapted from Heo et al. (2017).

After assembling our synthetic operon expression system for strongly expressed enzymes as our literature-based iteration, we would then conduct high-performance liquid chromatography (HPLC) to detect and quantify the pterostilbene produced by our bacteria.

Although our modelling would provide an optimal promoter combination and plasmid copy numbers that would minimise metabolic stress and thus potentially further increase pterostilbene yield, the lateness in the arrival of the iGEM 2022 Distribution Kits meant that only our literature-based approach of our type IIS assembly for pterostilbene production was possible within the resulting timeframe.

Synthesis of Parts for JUMP Assembly

While JUMP type IIS assembly uses BsmBI, BsaI, and BsmBI for level 0, 1, and 2 assembly, respectively, we removed the BsmBI recognition and cleavage sites (used for level 0) in the prefixes and suffixes, replacing them with BioBricks prefix and suffix sequences for digestion with XbaI and SpeI in level 0 assembly; level 1 and 2 assemblies were unaffected since the BsaI recognition and cleavage sites were maintained in these prefixes and suffixes (for level 1), and BsmBI sites (for level 2) are located in the level 1 plasmid backbones. This way, our sequences are adapted for RFC10 assembly while still being compatible for assembly into level 0 JUMP vectors via BioBrick prefixes and suffixes.

Since the promoters and RBS were of short lengths individually (as is usually the case) and were below the minimum sequence length required for synthesis by IDT, we merged these two into a single part as a practical alternative to ordering longer parts and conducting PCR for the desired small parts. The promoter 5’ overhang and RBS 3’ overhang were used as the cleavage sites of this merged Promoter+RBS part, ensuring ligation is still directional and ordered as desired.

Therefore, to adapt our promoters, ribosomal binding site (RBS), coding sequences (CDS), and terminator for JUMP type IIS assembly with our modifications, we added the adequate prefixes and suffixes specific to each part (Fig. 2).

Prefixes and Suffixes Added to Each Part for JUMP Type IIS Assembly
Figure 2. Prefixes and Suffixes Added to Each Part for JUMP Type IIS Assembly. Prefix ‘AGCT’ and suffix ‘ACTG’ (gray) were added to sequence ends for efficient cleavage by EcoRI and PstI if used. Nucleotide ‘T’ (gray) is used as the natural ‘skip’ random nucleotide between the BsaI recognition and cleavage site. XbaI (yellow) and SpeI (orange) sites are used for Level 0 assembly. BsaI sites (pink) are used for Level 1 assembly.

As supported in the literature, we opted for the strong expression of each enzyme introduced into the E. coli (Heo et al., 2017; Yan et al., 2021). More specifically, we selected the strong Anderson promoter BBa_J23100 as it is part of the well characterised Anderson collection, compared for both its relative and absolute strength (Kelly et al., 2009) to various other promoters displaying a range of strengths. This is especially favourable for our modelling as the quantitative strength range in Polymerases per Second (PoPS) allows us to provide more accurate values to predict and control each proteins’ concentration. Regarding the RBS chosen, the very well characterised series BBa_B0030-B0034 was used to select the strong RBS BBa_B0030. Due to the suggested potential insolubility of the resveratrol O-methyltransferase enzyme (Kallscheuer et al., 2017) as well as the toxicity of the accumulated intermediate 4-coumarate (Rodríguez-Ochoa et al., 2022), it is important to tightly control the expression of the four enzymes. Therefore, we added a lac operator (lacO) sequence between the promoter and RBS (Fig. 3) in order to recruit endogenous lac repressor protein (lacR) to repress transcription in the absence of IPTG; this has also been used in literature on pterostilbene production (Heo et al., 2017).

BBa_K4388013: Strong Anderson Promoter (BBa_J23100) with Downstream lacO (BBa_K1624002) and Strong RBS (BBa_B0030)
Figure 3. BBa_K4388013: Strong Anderson Promoter (BBa_J23100) with Downstream lacO (BBa_K1624002) and Strong RBS (BBa_B0030). Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. The BBa_J23100 promoter used has an estimated Polymerases Per Second (PoPS) value of ~0.043 (Kelly et al., 2009). The lac operator (lacO) is bound by the lac repressor protein (lacR, encoded by the lacI gene), blocking transcription by RNA Polymerase II (Schlax et al., 1995). BBa_B0030 is a strong RBS (Weiss, 1970).

The genes incorporated into our system were those encoding mutants of enzymes required for the biosynthesis of pterostilbene from tyrosine, from the homologues that have demonstrated the highest efficiency for their purpose in our system: RgTAL, At4CL1, VvSTS, and VvROMT. The wild-type versions of these enzymes were identified from Yan et al. (2021) and their sources: RgTAL gene sequence was obtained from Zhou et al. (2016) (GenBank Accession Number: KF765779); the wild-type At4CL1 gene sequence was obtained from UniProt (Q42524) and the iGEM registry (BBa_K1033001); the wild-type VvSTS gene sequence was obtained from Lim et al. (2011) (GenBank Accession Number: DQ459351); the wild-type VvROMT gene sequence was obtained from Wang et al. (2014) (Genbank Accession Number: FM178870.1).

For synthesis of the desired protein sequences, the gene sequences underwent a variety of checks and modifications:

  1. DNA sequences were translated using the ExPASy Translate tool to identify the reading frame (based on the long ORF) for subsequent modifications.
  2. Any 5’ and 3’ untranslated regions were removed.
  3. Any 3’ stop codons were removed.
  4. Mutations were inserted in the required positions.
  5. Sequences were checked for signal peptides using the SignalP 6.0 server.
  6. Double stop codons (TAATAA) were added to the 3’ end of each sequence.
  7. Sequences were codon-optimised for E. coli K12 and verified using the Rare Codon Analyser.
  8. Illegal restriction sites (RFC10: EcoRI, NotI, XbaI, SpeI, PstI; RFC1000 & JUMP: BsaI, SapI, BsmBI) were removed by altering one of the codons within the restriction sites to the next highest frequency codon, consulting the codon usage frequency table.
  9. The prefix and suffix for coding sequences were added to each sequence (Fig. 4-7).
BBa_K4388001: Rhodotorula glutinis Tyrosine Ammonia-Lyase Mutant Codon-Optimised for Expression in E. coli K12
Figure 4. BBa_K4388001: Rhodotorula glutinis Tyrosine Ammonia-Lyase Mutant Codon-Optimised for Expression in E. coli K12. Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. This mutant (S9N/A11T/E518V) has demonstrated an increase in activity, catalytic efficiency, and 4-coumarate yield over the wild-type (Zhou et al., 2016).
BBa_K4388002: Arabidopsis thaliana 4-Coumarate:CoA Ligase 1 Mutant Codon-Optimised for Expression in E. coli K12
Figure 5. BBa_K4388002: Arabidopsis thaliana 4-Coumarate:CoA Ligase 1 Mutant Codon-Optimised for Expression in E. coli K12. Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. This mutant (L57I/L460H) has demonstrated an increase in in vivo and in vitro activity over the wild-type (Yan et al., 2021).
BBa_K4388003: Vitis vinifera Stilbene Synthase Mutant Codon-Optimised for Expression in E. coli K12
Figure 6. BBa_K4388003: Vitis vinifera Stilbene Synthase Mutant Codon-Optimised for Expression in E. coli K12. Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. This mutant (T50I/V170A) has demonstrated an increase in in vivo and in vitro activity over the wild-type (Yan et al., 2021).
BBa_K4388004: Vitis vinifera Resveratrol O-Methyltransferase Mutant Codon-Optimised for Expression in E. coli K12
Figure 7. BBa_K4388004: Vitis vinifera Resveratrol O-Methyltransferase Mutant Codon-Optimised for Expression in E. coli K12. Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. This mutant (S29P) has demonstrated an increase in in vivo and in vitro activity over the wild-type (Yan et al., 2021).

The terminator selected was BBa_B0015 (Fig. 8), a strong, forward, double terminator that is the most commonly used for ensuring transcription termination.

BBa_B0015: Strong Forward Double Terminator
Figure 8. BBa_B0015: Strong Forward Double Terminator. Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. According to its registry page, this terminator has demonstrated 0.984 and 0.97 forward termination efficiency.

In order to assemble a synthetic operon whereby all four genes are transcribed into a single mRNA molecule, there should only be one promoter (upstream of the first gene) and one transcriptional terminator (downstream of the last gene). Therefore, all other 5’ UTRs should encompass a linker+RBS sequence (Fig. 9) rather than promoter+RBS, and all other 3’ UTRs should replace the transcriptional terminator with a linker sequence (Fig. 10).

BBa_K4388011: Linker (No Promoter) with Downstream Strong RBS (B0030)
Figure 9. BBa_K4388011: Linker (No Promoter) with Downstream Strong RBS (B0030). Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. The strong Anderson promoter (BBa_J23100) from the BBa_K4388013 part was replaced by a linker sequence that does not promote transcription. BBa_B0030 is a strong RBS (Weiss, 1970). Altogether, this part is needed to create an operon-like system and, as such, was inserted upstream of each gene but the first one.
BBa_K4388012: Linker (No Terminator)
Figure 10. BBa_K4388012: Linker (No Terminator). Prefix and suffix sequences were added to adapt this part for RFC10 and JUMP assembly. The strong Anderson promoter (BBa_J23100) from the BBa_K4388013 part was replaced by a linker sequence that does not promote transcription. BBa_B0030 is a strong RBS (Weiss, 1970). Altogether, this part is needed to create an operon-like system and, as such, was inserted upstream of each gene but the first one.


JUMP Type IIS Assembly

Level 1 Assembly

JUMP type IIS Level 1 assembly consisted of constructing each of the four transcriptional units (TU) via a one-pot digestion reaction with BsaI and ligation into pJUMP29-1A, pJUMP29-1B, pJUMP29-1C, or pJUMP29-1D. Successful ligation and transformation were selected for by identifying non-fluorescent colonies on kanamycin agar plates (Fig. 12).

Construction of Each Transcriptional Unit into Level 1 pJUMP29-1A/B/C/D
Figure 12. Construction of Each Transcriptional Unit (TU) into Level 1 pJUMP29-1A/B/C/D. Successful ligation into composite parts in respective plasmids: (a) BBa_K4388005 (BBa_K4388013+BBa_K4388001+BBa_K4388012) in pJUMP29-1A, (b) BBa_K4388006 (BBa_K4388011+BBa_K4388002+BBa_K4388012) in pJUMP29-1B, (c) BBa_K4388007 (BBa_K4388011+BBa_K4388003+BBa_K4388012) in pJUMP29-1C, and (d) BBa_K4388008 (BBa_K4388011+BBa_K4388004+BBa_B0015) in pJUMP29-1D. White (sfGPF-excised) colonies indicate successful ligation of TU into plasmid, and successful transformation of plasmid conferring kanamycin resistance. Final level 1 constructs were designed using SnapGene.

Due to the extreme time-constraints imposed by the lateness in arrival of the 2022 iGEM Distribution Kit, we were not able to execute level 2 JUMP assembly and subsequent bacterial culturing for pterostilbene production. Nevertheless, we have extensively prepared these future steps to continue our proof of concept. We have presented all our findings on our results page.

Level 2 Assembly

JUMP type IIS Level 2 assembly would take each TU in pJUMP29-1A, pJUMP29-1B, pJUMP29-1C, and pJUMP29-1D, for a one-pot digestion reaction with BsmBI and ligation into the final vector pJUMP49-2A. Successful ligation and transformation would be selected for by identifying non-fluorescent colonies on spectinomycin agar plates (Fig. 13).

Construction of Multiple Transcriptional Units (MTU) in Level 2 Vector pJUMP49-2A
Figure 13. Construction of Multiple Transcriptional Units (MTU) in Level 2 Vector pJUMP49-2A. Ligation together of transcriptional units BBa_K4388005, BBa_K4388006, BBa_K4388007, and BBa_K4388008 would create the multiple transcriptional units composite part BBa_K4388010, assembled into the final level 2 construct. The final plasmid vector (BBa_4388010 in pJUMP49-2A) was designed using SnapGene. Successful ligation and transformation would be observed by white (sfGFP-excised) colonies and colony survival on LB agar media with spectinomycin.

Pterostilbene Detection and Quantification

In order to detect and quantify the pterostilbene that would be produced by our system, we would culture the E. coli transformed with the final plasmid containing our synthetic operon, and would conduct reverse-phase high-performance liquid chromatography (HPLC) of the cell lysates. Firstly, a standard curve with different pterostilbene concentrations against the area under the HPLC peak would be constructed to then determine pterostilbene concentration from the cell lysates. The area under the HPLC peak at each pterostilbene concentration would be determined using appropriate chromatography software, then allowing for linear regression analysis to determine the degree of linearity.

For the analysis of pterostilbene produced in our synthetic operon system, HPLC would be performed considering varying IPTG concentrations for operon expression, as well as at different time intervals (e.g. 24, 48 hours); this information would importantly inform our entrepreneurship and proposed implementation in regards to large-scale production costs.

References

Arbo, B. D., André-Miral, C., Nasre-Nasser, R. G., Schimith, L. E., Santos, M. G., Costa-Silva, D., . . . Hort, M. A. (2020). Resveratrol derivatives as potential treatments for alzheimer’s and parkinson’s disease. Frontiers in Aging Neuroscience, 12. doi:10.3389/fnagi.2020.00103

Chang, J., Rimando, A., Pallas, M., Camins, A., Porquet, D., Reeves, J., . . . Casadesus, G. (2012). Low-dose pterostilbene, but not resveratrol, is a potent neuromodulator in aging and alzheimer's disease. Neurobiology of Aging, 33(9), 2062-2071. doi:10.1016/j.neurobiolaging.2011.08.015

Gerngross, D., Beerenwinkel, N., & Panke, S. (2022). Systematic investigation of synthetic operon designs enables prediction and control of expression levels of multiple proteins. BioRxiv. doi:10.1101/2022.06.10.495604

Heo, K. T., Kang, S., & Hong, Y. (2017). De novo biosynthesis of pterostilbene in an escherichia coli strain using a new resveratrol O-methyltransferase from Arabidopsis. Microbial Cell Factories, 16(1). doi:10.1186/s12934-017-0644-6

Kallscheuer, N., Vogt, M., Bott, M., & Marienhagen, J. (2017). Functional expression of plant-derived O -methyltransferase, flavanone 3-hydroxylase, and flavonol synthase in Corynebacterium glutamicum for production of pterostilbene, Kaempferol, and quercetin. Journal of Biotechnology, 258, 190-196. doi:10.1016/j.jbiotec.2017.01.006

Kelly, J. R., Rubin, A. J., Davis, J. H., Ajo-Franklin, C. M., Cumbers, J., Czar, M. J., . . . Endy, D. (2009). Measuring the activity of BioBrick promoters using an in vivo reference standard. Journal of Biological Engineering, 3(1), 4. doi:10.1186/1754-1611-3-4

Kosuru, R., Rai, U., Prakash, S., Singh, A., & Singh, S. (2016). Promising therapeutic potential of pterostilbene and its mechanistic insight based on preclinical evidence. European Journal of Pharmacology, 789, 229-243. doi:10.1016/j.ejphar.2016.07.046

Lange, K. W., & Li, S. (2017). Resveratrol, pterostilbene, and dementia. BioFactors, 44(1), 83-90. doi:10.1002/biof.1396

Li, H., Lyv, Y., Zhou, S., Yu, S., & Zhou, J. (2022). Microbial cell factories for the production of flavonoids–barriers and opportunities. Bioresource Technology, 360, 127538. doi:10.1016/j.biortech.2022.127538

Lim, C. G., Fowler, Z. L., Hueller, T., Schaffer, S., & Koffas, M. A. (2011). High-yield resveratrol production in engineered escherichia coli. Applied and Environmental Microbiology, 77(10), 3451-3460. doi:10.1128/aem.02186-10

Rodríguez-Ochoa, J. I., Fragoso-Jiménez, J. C., Flores, N., Martínez, L. M., Bolivar, F., Martinez, A., & Gosset, G. (2022). Global transcriptomic response of escherichia coli to P-coumaric acid. Microbial Cell Factories, 21(1). doi:10.1186/s12934-022-01874-6

Schlax, P. J., Capp, M. W., & Record Jr, T. M. (1995). Inhibition of transcription initiation buiacrepressor. Journal of Molecular Biology, 245(4), 331-350. doi:10.1006/jmbi.1994.0028

Valenzuela-Ortega, M., & French, C. (2021). Joint Universal Modular Plasmids (JUMP): A flexible vector platform for synthetic biology. Synthetic Biology, 6(1). doi:10.1093/synbio/ysab003

Wang, Y., Bhuiya, M. W., Zhou, R., & Yu, O. (2014). Pterostilbene production by microorganisms expressing resveratrol O-methyltransferase. Annals of Microbiology, 65(2), 817-826. doi:10.1007/s13213-014-0922-z

Weiss, R., & Knight, T. F. (2004). Cellular computation and communication using engineered Genetic Regulatory Networks. Cellular Computing. doi:10.1093/oso/9780195155396.003.0012

Yan, Z., Liang, J., Niu, F., Shen, Y., & Liu, J. (2021). Enhanced production of Pterostilbene in escherichia coli through directed evolution and host strain engineering. Frontiers in Microbiology, 12. doi:10.3389/fmicb.2021.710405

Zhou, S., Liu, P., Chen, J., Du, G., Li, H., & Zhou, J. (2016). Characterization of mutants of a tyrosine ammonia-lyase from Rhodotorula glutinis. Applied Microbiology and Biotechnology, 100(24), 10443-10452. doi:10.1007/s00253-016-7672-8