Team:UNSW Australia/Design


Team: UNSW Australia


Design

INTRODUCTION

OVERVIEW OF DESIGN ITERATIONS IN OUR PROJECT

In designing our project, we established three main aims;

  • Firstly, the development of the Assemblase scaffold, ensuring that it can conjugate with two different enzymes and improve overall reaction efficiency.
  • The improvement of the current biosynthetic pathway of Paclitaxel production, which uses the enzymes Phenylalanine Aminomutase (PAM) and Tyrocidine Synthase 1 (TycA) to biosynthetically produce the Paclitaxel side chain. Assemblase would reduce the occurrence of unwanted intermediates, improving the efficiency of the rate limiting enzymes.
  • Development of a second, more sustainable Paclitaxel production pathway which utilises the enzymes Beta-D-xylosidase (LXYL-p1-2) and 10-deacetylbaccatin III-10-O-acetyltransferase (DBAT) to convert the Paclitaxel analogue, 7-β-xylosyl-10-deacetyltaxol (XDT) into Paclitaxel.

In our efforts to achieve each aim, we integrated cycles of the Design-Build-Test-Learn workflow, a pipeline previously successfully used to optimize the efficient testing of synthetic biological systems(1)(2). The Design-Build-Test-Learn framework involves; the conceptual depiction of a biological system of interest, the construction of a sample in the laboratory, the experimental measurements gained from the physical sample, and the improvement which is done from further analysis. This is a perpetually cyclic process that allowed further incorporation of data gained from human practices.

DESIGN OF ASSEMBLASE: MEETING INCREASING DEMAND FOR PACLITAXEL

In our initial design of the Assemblase scaffold we established three main focuses;

  • We needed to design Assemblase to be simple for ease of production, assembly, and use. This would ensure that the structure could be adapted for use in manufacturing, which would have the most significant benefit for users. It would also aid researchers, making our system easier to test and incorporate into experiments. This concept of simplicity was achieved by designing the Assemblase scaffold to self-assemble, and independently conjugate to enzymes.
  • To perform reactions involved in industrial manufacturing, it needed to be ensured that enzymes were securely attached to the Assemblase scaffold. Bonds formed between enzyme and scaffold needed to be irreversible, and the scaffold needed to be stable in a large range of pH and temperature. To achieve this, we incorporated the Spy/SnoopCatcher and Tag System which form irreversible, covalent isopeptide bonds. Assemblase is also constructed from prefoldin subunits sourced from the thermophilic Archaeon, Methanobacterium thermoautotrophicum, strengthening the adaptability and stability of our system.
  • With industrial enzymatic reactions in mind, we identified the need for Assemblase and conjugated enzymes to be retained and reused. We devised two filtration methods and used Molecular Dynamics modelling to determine corresponding filter sizes. We also researched current industry practice, to determine what filtration size would be most easily integrated into the current market.

Designed experimental tests were fixated on our initial design focusses;

Self-Assembly

  • Prefoldin is a family of proteins which act as heterohexameric molecular chaperones. When subunits are individually present within solution, they will self-assemble into a hexameric structure in a set molar ratio. A complete Assemblase molecule contains 2 alpha-prefoldins and 4 beta-prefoldins. We aimed to test this concept of self-assembly through the design of a Native gel electrophoresis experiment. In utilising Native gel electrophoresis, proteins were separated based on both charge and hydrodynamic size which allowed for self-assembly and the detection of successfully formed hexameric structures.

    • Prefoldin: Hexamer Composed of 4 Beta and 2 Alpha Subunits

Conjugation of enzymes

  • In the Snoop/SpyCatcher and Tag system, ‘Catcher’ sequences on the scaffold and ‘Tag’ sequences on the enzyme form an irreversible isopeptide bond spontaneously. To test the Snoop/SpyCatcher and Tag system’s functionality, we designed an SDS-page gel experiment. Although this was not appropriate to demonstrate self-assembly due to polar interactions, the ‘catcher’ and ‘tag’ form covalent bonds when combined which do not interfere with testing. Thus, we were able to demonstrate the ability of the Assemblase scaffold to conjugate with different enzymes.

SnoopCatcher-SnoopTag

Co-localisation of enzymes

  • To increase the overall rate of reaction, Assemblase co-localises consequent enzymes in multi-step reactions, decreasing diffusion distances and lowering the probability of unwanted intermediate production. To confirm that Assemblase successfully met our design, we planned a Förster resonance energy transfer (FRET) experiment. By fusing fluorescent proteins; mCerulean3 and mVenus to Snoop/Spy tags and adding them to scaffold in solution, we are able to confirm conjugation and co-localisation effects through their interaction. This is because a positive FRET signal will appear if mCerulean3 and mVenus proteins are within 1-10 nanometres.

GENE DESIGN

To streamline laboratory workflows, we used pET19b plasmids for all enzyme inserts. This allowed us to perform many of the same cloning and expression methods. pET19b plasmids were also preferentially used by our supervisor Dr Dominic Glover, allowing for assistance in troubleshooting.

Shared designed components of the sequences were:

  • Gibson Forward and Reverse Overhangs: Complementary overhangs are added outside protein coding regions to enable cloning into pET19b via Gibson Assembly.
  • GSG Linker: GSG linkers are included between components for additional flexibility to allow individual, unhindered protein folding of each component.
  • 6x HIS tag: These tags have a high affinity to Ni2+ enabling protein purification using Immobilised metal affinity chromatography (IMAC) via a Ni-NTA column.

DESIGN OF PATHWAY 1: MEETING INCREASING DEMAND FOR PACLITAXEL

OVERVIEW

Our first pathway aims to improve the current biosynthetic pathway of Paclitaxel production by scaffolding the rate limiting enzymes; Phenylalanine Aminomutase (PAM) and Tyrocidine Synthase 1 (TycA) which produce the Paclitaxel side chain. Assemblase was designed in order to reduce the occurrence of unwanted intermediates and increase overall reaction rate by co-localisation.

HOW WE DESIGNED OUR PROJECT TO MEET THIS AIM

The enzymes PAM and TycA were chosen as they were two succeeding enzymes in the current Paclitaxel biosynthesis pathway with two of the lowest kinetic Kcat values at 0.015/s and 0.05/s respectively (3). Their kinetic properties influenced our design, as we determined that attaching the slower enzyme to a greater amount of prefoldin subunits would be optimal. Thus, we attached the SnoopTag to PAM in order to facilitate attachment to the 4 beta-prefoldin SnoopCatcher subunits, and the SpyTag to TycA. To further increase overall reaction rate, we implemented a single amino acid mutation of Serine to Alanine at the 563 position in TycA. This mutant was shown to half the Michaelis constant (Km) of CoA (4), a vital co-substrate in the enzyme catalysed reaction.

TESTS DESIGNED TO PROVE THE VALIDITY OF OUR DESIGN

In order to determine whether conjugation and co-localisation of the designed enzymes would improve the overall reaction rate, assays were designed to determine Kcat values.

PAM:

  • PAM transfers the amino (NH2) group from the α-carbon of phenylalanine to the β-carbon. We hypothesised that the change in the position of the functional group would alter the hydrogen environments in the amino acids, which could be detected using 1H nuclear magnetic resonance (NMR) spectroscopy. During sample preparation, an internal standard could be included so that spectra can be taken at different time points and calibrated accordingly, allowing for semi-quantitative analysis.

    • Phenylalanine Aminomutase

TycA:

  • TycA ligates CoA-SH to β-phenylalanine utilizing energy from ATP hydrolysis. Free thiols can be quantified using Ellman’s Reagent (DTNB). DTNB has an oxidizing disulphide bond that is reduced when free thiols are present, releasing one molecule of 5-thio-2-nitrobenzoic acid (TNB). Thus, free thiol concentration can be determined by observing absorption at 412 nm and measuring TNB concentration. This can be used to calculate DBAT Kcat values with regards to the substrate: 10-deacetyltaxol (DT) and co-substrate: Acetyl Coenzyme-A.

    • Tyrocidine Synthetase

Design of Pathway 2: Increasing sustainability of Paclitaxel production

OVERVIEW

The second pathway of our project aims to scaffold enzymes that can produce Paclitaxel from more sustainable pre-cursors. This would allow us to make full usage of the Yew Tree’s natural products, reduce logging of the tree, and re-place toxic solvents or reagents.

HOW WE DESIGNED OUR PROJECT TO MEET THIS AIM

Upon researching further into Paclitaxel production and Yew Tree farming, we realised that the Yew Tree contains many other precursors and natural products that we can extract and use, which are wasted in current production procedures. In Lentinula edodes, Beta-D-xylosidase (LXYL) naturally catalyses the reaction of 7-beta-xylosyl-10-deacetylbaccatin III (XDB) to 10-deacetylbaccatin III (DB). We decided to use a mutant of LXYL (LXYL-p1-2), which has been engineered to catalyse the removal of the xylose group from 7-beta-xylosyl-10-deacetyltaxol (XDT) to produce the intermediate 10-deacetyltaxol (DT). We then chose the succeeding enzyme of 10-deacetylbaccatin III-10-O-acetyltransferase (DBAT) which catalyses the acetylation of 10-deacetyltaxol (DT) to Paclitaxel. However, we designed our sequence to have two amino acid mutations of G38R and F301V, which have been documented to have a catalytic efficiency six times higher than the wildtype. Once again we designed the Tag sequences to correspond to enzyme Kcat values, with the slower enzyme of DBAT (Kcat of 0.001499/s) to fuse to the SnoopTag, and LXYL-p1-2 (Kcat of 1.92/s) to fuse to the SpyTag (5)

TESTS DESIGNED TO PROVE THE VALIDITY OF OUR DESIGN

In order to determine whether conjugation and co-localisation of the designed enzymes would improve the overall reaction rate, assays were designed to determine Kcat values.

DBAT:

  • DBAT converts the co-substrate Acetyl Coenzyme-A to Coenzyme-A-SH. Free thiols can be quantified using Ellman’s Reagent (DTNB). DTNB has an oxidizing disulphide bond that is reduced when free thiols are present, releasing one molecule of 5-thio-2-nitrobenzoic acid (TNB). Thus, free thiol concentration can be determined by observing absorption at 412 nm and measuring TNB concentration. This can be used to calculate DBAT Kcat values with regards to the substrate: 10-deacetyltaxol (DT) and co-substrate: Acetyl Coenzyme-A.

10-deacetylbaccatin III-10-O-acetyltransferase

References

  1. Veggiani G, Nakamura T, Brenner MD, Gayet RV, Yan J, Robinson CV, Howarth M. Programmable polyproteams built using twin peptide superglues. Proceedings of the National Academy of Sciences. 2016 Feb 2;113(5):1202-7.
  2. Carbonell P, Jervis AJ, Robinson CJ, Yan C, Dunstan M, Swainston N, Vinaixa M, Hollywood KA, Currin A, Rattray NJ, Taylor S. An automated design-build-test-learn pipeline for enhanced microbial production of fine chemicals. Communications biology. 2018 Jun 8;1(1):66.
  3. Walker, K. D., Klettke, K., Akiyama, T., & Croteau, R. (2004). Cloning, Heterologous Expression and Characterization of a Phenylalanine Aminomutase Involved in Taxol Biosynthesis. Journal of Biological Chemistry, 53947-53954.
  4. Bennett, B. D., Kimball, E. H., Gao, M., Osterhout, R., Dien, S. J., & Rabinowitz, J. D. (2009). Absolute metabolite concentrations and implied enzyme active site occupancy in Escherichia coli. Nature Chemical Biology, 593-599.
  5. Li, B. J., Wang, H., Gong, T., Chen, J. J., Chen, T. J., Yang, J. L., & Zhu, P. (2017). Improving 10-deacetylbaccatin III-10-β-O-acetyltransferase catalytic fitness for Taxol production. Nature communications, 8, 15544. doi:10.1038/ncomms15544