Additionally, Chaichanan et al. measured an 8-fold increase in zofimarin production by the endophytic fungus, Xylaria sp. Acra L38 utilizing a complete, from screening to optimization, design of experiments approach.Therefore, we sought to implement this RSM approach to optimize the media for increased titer of olivetolic acid and its analogs. To do so, similar to the screening design, we had to determine which RSM design to use. JMP software offers three methods of performing a response surface design experiment: custom design, central composite design, and Box-Benhten design. Box-Behnken designs are the second of the two RSM designs provided by JMP. BoxBehnken designs only contain three levels: low, midpoint, and high. There are no axial points. Due to its lack of axial points, Box-Behnken designs typically have higher prediction variance than central composite designs.We ultimate decided to proceed with the central composite design due to its ability to incorporate these axial points. Since we did not know where the optimum of the factor values lied,flood tray we hypothesized that these axial points could prove closer to the optimum than the initially two-level factor values and so it was best to include them. We ran the assays in sets of 5, once again utilizing CD-ST as our control media for each run. We augmented the design, added 12 more runs .
We recorded the titers for each of the runs and graphed the data. From the data, we observed that in 6 media formulations, our A.nidulans cultures producing olivetolic acid and its analogs produced these metabolites at higher titers than when cultured in CD-ST. Both the screening design runs and the central composite design RSM set of experiments were done in 10 ml cultures in 50 ml falcon tubes. From there, we chose the three media formulations that outputted the greatest titers and cultured the A. nidulans strain in 25 ml of the respective medias in a 125 ml flask to test whether these results would hold.Therefore, RSM run #6 is the best option for improvement of olivetolic acid titer and analog. For further direction and next steps, we would need to determine if this new optimized titer is a local optimum or a global optimum. To do this, we would need to model this data using the JMP software and utilize the software’s predictive modeling tool to predict whether or not we have reached the global optimum and if not, which values outside the 5 levels previously chosen that we need to input for the next series of runs. To increase titer outside of optimization of culture media, we sought to increase malonyl coenzyme production in the Aspergillus nidulans fungal body. With regards to our platform, one unit of acetyl coenzyme A is used as a building block along with three units of malonyl-CoA to form hexanoyl-thioester, hexenyl-thioester, octanoyl-thioester, or octenyl-thioester, generated by the HRPKS of Metarhizium anisopliae. Similarly, one unit of acetyl-CoA along with three units of malonyl-CoA are utilized by the HRPKS of Talaromyces islandicus to produce hexanoyl-thioester.
Perusing through the literature, we found that overexpression of the enzyme acetyl-CoA carboxylase was found to increase the malonyl-CoA production.Acetyl-CoA carboxylase is responsible for converting one unit of acetyl-CoA to one unit of malonyl-CoA. The enzyme catalyzes the carboxylation of acetyl-CoA to malonyl-CoA. This reaction is ATP-dependent. Malonyl-CoA is a regulator of fatty acid oxidation and is the primary substrate for fatty acid synthase. In fact, in fungi, inhibition of acetyl-CoA carboxylase rapidly leads to cell death through membrane dysfunction due to fatty acid depletion caused by lack of the malonyl-CoA building block.Therefore, we hypothesized that increase of the malonyl-CoA production through overexpression of acetyl-CoA carboxylase would lead to increased titers of olivetolic acid and its analogs. We mined for the acetyl-CoA carboxylase enzyme in A. nidulans and overexpressed the enzyme in combination with heterologous expression of Ti_OvaA, Ti_OvaB, and Ti_OvaC and measured the titers. From the data, overexpression of acetyl-CoA carboxylase with Ti_OvA, Ti_OvAB, and Ti_OvaC in A. nidulans consistently led to improved olivetolic titers over the A. nidulans control strain expressing Ti_OvaA, Ti_OvaB, and Ti_OvaC with the acetyl-CoA carboxylase overexpression colonies producing more than 2.5-fold more olivetolic acid. Therefore, through optimization of production media utilizing DOE and increased malonyl-CoA production through overexpression of acetyl-CoA carboxylase, we were able to increase the titers of our platform’s compounds. This is necessary as a step to generate an industrial relevant strain producing cannabinoids.
Furthermore, high production of these intermediates in the cannabinoid bio-synthetic pathway is important due to the high costs of these intermediates in the market due to their low availability. High production and isolation of these compounds can be used in combination with in vitro enzymatic methods to produce the final elaborated cannabinoids. These cannabinoids, especially the non-common ones which are not characterized, can be subjected to biological activity assays and thereby characterized, further advancing the knowledge of the field and possibly providing therapeutic options. Even more so, as detailed in the next chapter, these intermediates can themselves have intriguing biological activities, making high production of these intermediates important. As previously described in Chapter 2, heterologous expression of the Ma_OvaA, Ma_OvaB, and Ma_OvaC genes from Metarhizium anisopliae in Aspergillus nidulans led to the production of four products: olivetolic acid, an unsaturated analog of olivetolic acid, sphaerophorolcarboxylic acid, and an unsaturated analog of sphaerophorolcarboxylic acid at high titers, thereby making this an attractive platform to further develop downstream in order to access rare or new to nature cannabinoids that have biological potential. Furthermore, even olivetolic acid and olivetolic acid analogs are proposed to have biological activity such as antibacterial, antifungal, cytotoxic, and photoprotective properties. One study published had demonstrated that olivetolic acid had shown an anticonvulsant effect in a mouse model of Dravet syndrome, similar in effectives to CBD.Recently, a study chronicling the antibacterial effects of OA and a few of its analogs was published. Lee et al. chemically synthesized olivetolic acid as well as the propyl, heptyl, nonyl, undecyl, and tridecyl analogs of olivetolic acid. They tested these compounds for antibacterial activity against the bacteria Bacillus subtilis and Staphylococcus aureus. Although OA and the propyl variant showed very little antibacterial activity against both Bacillus subtilis and Staphylococcus aureus, the heptyl, nonyl, undecyl, and tridecyl variants did, with a partial trend of increasing activity based on length. Both the undecyl and tridecyl analogs had a minimum inhibitory concentration of 2.5 µM against Bacillus subtilis and 6.25 µM against Staphylococcus aureus demonstrating that not only do cannabinoids have potent biological activity/therapeutic potential further to be explored but also the intermediates in the cannabinoid bio-synthetic pathway also have biological activity and are worth further exploring. Not only have the olivetolic acid variants shown promising activity, but also analogs of ∆ 9 -THC have also demonstrated potent biological activity. A report in 2019 was published in which the authors extracted the heptyl analog of ∆9 -THC, known as ∆9 -tetrahydrocannbiphorol from the cannabis plant and performed assays to measure its biological activity. This rare cannabinoid displayed almost 30 times greater binding affinity to CB1 than ∆9 -THC did and also six times greater binding affinity to CB2 than ∆9 -THC. The pharmacological activity of ∆9 – THC is ascribed to its binding activity to the CB1 receptor, with the length of the alkyl chain being directly correlated to the binding activity.Therefore, a rare or new to nature cannabinoid that has greater binding affinity to the CB1 receptor may potentially offer greater medicinal effects than what is currently known for ∆9 -THC. Therefore, although we already produced, 4×8 grow tray through our platform, four compounds with potential activity, we sought to further diversity our product profile and ultimately access the final elaborated cannabinoids. To achieve this, the ketosynthase domain of the HRPKS was mutated. The domain facilitates decarboxylative Claisen condensation, catalyzing the formation of a carbon-carbon bond, extending the growing acyl chain.
Therefore, modification of the KS domain can affectpolyketide elongation, generating olivetolic acid variants. Studying the Saccharomyces cerevisiae fatty acid synthase 2, Johannson et al. identified that amino acid M1251 was central to the KS channel.Utilizing that information, Zhu et al. engineered variants with the mutations M1251W and G1250S and noted that they saw increased C6 as well as C8 production.Additionally, Gajewski et al. performed mutations on the equivalent M1251 in the fatty acid synthase of Corynebacterium ammoniagenes as well as other mutations for directed polyketide production.They ultimately determined that mutations of M1251 or its equivalent in other fatty acid synthases promoted chain length control by forming a kinetic barrier that steers the product from further KS elongation and onto release. With that knowledge, we sought to mutate the M1251 equivalent in our HRPKS gene as well as a wide host of other mutations . Utilizing the CastP program, we also identified the active site of the Ma_OvaA KS domain containing the canonical cysteine-histidine- histidine catalytic triad and sought to mutate amino acids residues surrounding the active site. To make these mutations in the active site of the KS region of Ma_OvaA, we utilized the Living Biofoundry located in the California NanoSystems Institute at UCLA. The Living Biofoundry is an automated, high throughput platform capable of performing tasks such as polymerase chain reaction, gene assembly, plasmid assembly, heterologous expression, transformation, amplification and metabolite analysist. The Living Biofoundry consists of a ThermoFisher Laboratory Automation System ,a Fluent Tecan liquid handling system, two Illumina sequencers, bioreactors, and a ThermoFisher TSQ Altis triple quadrupole liquid chromatography-mass spectrometer. The LAS is the central technology in the Biofoundry “enabling execution of automated synthetic biology and workflows at >5000 samples per week”. The LAS is equipped with thermal cyclers, reagent dispensers, automated incubators, plate readers, racks/columns for plate storage, and a state-of-the-art Spinnaker™ robot. The LAS utilizes a Momentum application programming interfaces that seamlessly is compatible with a whole range of laboratory information systems . With regards to the Spinnaker™ robot, it itself is a SCARA 4-axis microplate mover encompassed with not only a barcode reader and integrated vision system making it possible for inventory management at real time tracking but also a gripper that has plate detection and adjustable gripping force. That is helpful in order to minimize loss of supernatant as well as remove labware handling errors. The ThermoFisher Momentum API allows for relatively simple programming of the desired actions. The Momentum API gives the open to type out commands in the tpical fashion programmers are accustomed to or utilize its graphical interface to schedule actions . Therefore, the Momentum API is quite user friendly while offering optimal performance containing inventory controls and compatibility with over 325 automation friendly instruments.Finally, the Living Biofoundry is a tool of the BioPolymers, Automated Cell Infrastructure, Flow, and Integration Chemistry Materials Innovation Platform , “a platform dedicated to scalable production of bioderived building blocks and polymers from yeast, bacteria, and fungi.”BioPacific MIP is a collaboration of researchers from UCSB and UCLA, funded by the National Science Foundation . We, therefore utilized the Living Biofoundry to construct the plasmids containing ketosynthase mutations for the purpose of diversifying our product profile. We programmed the system to be almost fully capable of automating the plasmid making process. The only steps that were not automated were the plating and the colony picking steps. . The process of implementing a process that is done on the bench top to the automated system proved initially to be difficult due to the difference in what is possible to do on the LAS vs on the bench top. For instance, typically after PCR, we perform gel electrophoresis, and purify the DNA band from the gel. However, that is not practical to do with the automated system; therefore, in order to purify the DNA from the PCR, we utilized a magnetic bead plate bound system. The magnetic bead plate was readily integrated into the Tecan Liquid Handling platform. As for another example on the difference between executing experiments on the benchtop vs on the LAS, we considered the ways we performed the E.coli transformation step. Typically, we electroporated the DNA into the E.coli competent cells using a cuvette and electroporator; however, we could not utilize electroporation on the LAS. There are 96 well plate electroporators, however, they are not equipped with the software to be integrable into the LAS. Therefore, instead of electroporation, we utilized chemically competent E.coli coli transformations and performed transformation experiments with those chemically competent cells.