These are just a couple of examples on how we had to conceptualize the plasmid making process differently on the LAS than how we typically do on the benchtop.After the construction of mutation plasmids both manually and utilizing the Living Biofoundry, we heterologously expressed the mutated Ma_OvA plasmids with Ma_OvaB and Ma_OvaC in Aspergillus nidulans. Out of the many mutation plasmids we constructed, we identified two that produced olivetolic acid analogs. Heterologous expression of first plasmid, containing the F418A and Y420A mutations, with Ma_OvaB and Ma_OvaC produced the nonyl and undecyl variants of olivetolic acid, both previously mentioned as having antibacterial activity. Heterologous expression of the second plasmid, containing the T318W and S347W mutations, with Ma_OvaB and Ma_OvaC produced orsellinic acid and divarinic acid which were confirmed by analytical standard and heterologous expression also produced what we propose to be ethyl variant based on mass and UV, although at low quantities so NMR was not taken. We also further utilized genome mining to elucidate new clusters producing olivetolic acid analogs. As previously detailed, we had identified three clusters homologous to the Metarhizium anisopliae cluster containing the Ma_OvaA, Ma_OvaB, and Ma_OvaC genes. As described, we identified the Tolypocladium inflatum and Metarhizium rileyi clusters in which heterologous expression produced the same product profile as Metarhizium anisopliae albeit at lower titers and the Talaromyces islandicus cluster which selectively produced olivetolic acid.
From the percent identity comparisons between the enzymes,2×4 flood tray we determined that clusters harboring close to 50% or less homology to the Metarhizium rileyi cluster would produce the greatest variety in products different from the product profile of the Metarhizium anisopliae cluster. Equipped with this knowledge, we utilized the Targeted Genome Mining Information Finder program, a MATLAB based program develop by Dr. Nicholas Liu, an alumnus of the Tang lab. Employing MATLAB’s Bioinformatics Toolbox which includes the ability to use Basic Local Alignment Search Tool to analyze FASTA formatted sequences, Dr. Liu developed a program to elucidate possible bio-synthetic gene clusters based on a target queried for. Although this program was developed to query for target resistance gene clusters, it can be used for a variety of different purposes. We employed the program to query for tandem polyketide synthases and used the Tang lab’s in-house fungal strain list as the database. We elucidated a cluster in Penicillium thomii containing 48%, 41%, and 36% homology to Ma_OvaA, Ma_OvaB, and Ma_OvaC, respectively. We heterologously expressed this cluster in Aspergillus nidulans and produced the nonyl olivetolic acid variant with a diene at the C1 and C3 positions of the alkyl chain as well as the heptyl variant unsaturated at the C3 position of the alkyl chain and a hydroxy group at the C2 position, both of which are non-native to the Cannabis sativa plant and therefore can be further processed to new to nature cannabinoids.We have detailed ways in which we were able to be diversify our product profile to produce rare olivetolic acid analogs. There are still more mutations near the active site of the KS domain as well as the AT domain that can be made to produce even more olivetolic acid analogs with different alkyl chain lengths.
With regards to homologous cluster expression, we only searched through the100+ sequenced fungal strains that we have. Therefore, there is a large number of fungal strains that can be queried to search for homologous clusters that can producer are olivetolic acid analogs. Testing of these variants and the ones produced utilizing the Living Biofoundry can lead to potentially promising results with regards to biological activities. As previously detailed, microbial production of naturally occurring and novel cannabinoids has potential to be a disruptive technology to the ~$10 billion global cannabis industry. From olivetolic acid, the next step in the cannabinoid bio-synthetic pathway is the geranylation of olivetolic acid to produce cannabigerolic acid , known as “the mother of all cannabinoids”CBGA can be decarboxylated to form cannabigerol , a cannabinoid with intriguing therapeutic potential. In the Cannabis sativa plant, CBG is produced in larger quantities in the early stage of the plant but in minute quantities in the mature stage of the plant. Preliminary research has indicated that CBG is non-psychoactive but has anti-oxidant, antimicrobial, anti-inflammatory, anticancer, photoprotective, and appetite-enhancing properties.Studies on the effect of CBG on the cannabinoid receptors have shown that CBG is a partial agonist for the CB2 receptor but cannot bind to the CB1 receptor, hence its nonpsychoactive properties.As previously described, the CB2 receptors are primarily located in the nervous system and agonists of the receptors provide anti-inflammatory and anti-oxidant effects. CBG was first isolated in 1970 and was fully characterized shortly after.We therefore sought to produce this cannabinoid microbially. To do so, we had to identify the prenyltransferase responsibly for geranylating the C2 carbon position of olivetolic acid.
This prenyltransferase activity was first demonstrated in the Cannabis plant and was proposed to beattributed to cannabis sativa prenyltransferase 1 and was soon patented. However, when Luo et al heterologously expressed CsPT1 in Saccharomyces cerevisiae, they observed no activity. Therefore, they mined for prenyltransferases with predicted activity similar to CsPT1 and heterologously expressed those candidate prenyltransferases. They demonstrated that the Cannabis sativa prenyltransferase 4 was able to successfully prenylate olivetolic acid to produce CBGA. CsPT4 was found in the Cannabis sativa plant and is part of the UbiAmembrane bound family of prenyltransferases, predicted to contain eight transmembrane helices. When expressed in Saccharomyces cerevisiae, and with the plastid targeting amino acid sequence removed, the CsPT4 enzyme was found to be located in the microsomal fractions of the yeast strain. Luo et al performed in vitro assays with the microsomal fraction harboring CsPT4 and demonstrated that the enzyme displayed Michaelis Menten behavior when olivetolic acid concentration varied whilst GPP concentration stayed constant but deviated from Michaelis Menten behavior when olivetolic acid concentration was held constant while GPP concentration varied.Aromatic prenyltransferases capable of geranylating olivetolic acid to produce CBGA outside of the Cannabis sativa plant have also been discovered. There are three classes of aromatic prenyltransferases: ABBA-type prenyltransferases, UbiA-type prenyltransferases, and dimethylallyl tryptophan synthase -type prenyltransferases. ABBA-type and DMATStype prenyltransferases are found in bacteria and fungi and UbiA-type prenyltransferases are found in fungi, plants, and bacteria. These aromatic prenyltransferases catalyze formation of carbon nitrogen, carbon oxygen, and carbon-carbon bonds between the prenyl donor’s carbonand the aromatic substrate.ABBA-type and DMATS-type have been elucidated as soluble aromatic prenyltransferases while UbiA prenyltransferases are membrane bound aromatic prenyltransferases. UbiA-type prenyltransferases are membrane bound prenyltransferases found in a variety of organisms such as bacteria, fungi, plant, human, etc. These prenyltransferases have been observed to be involved in menaquinone and ubiquinone biosynthesis as well as fungal meroterpenoid biosynthesis, archaeal membrane lipid biosynthesis, and prenylated aromatic secondary metabolites biosynthesis in plants, among other biosynthesis reactions. These UbiAtype prenyltransferases typically contain eight to nine transmembrane helices. Regarding their structure, enzymes in the family contain two conserved aspartate rich motifs with the first used for Mg2+ binding in order to catalyze the reaction; therefore, these prenyltransferases are metal dependent.DMATS-type prenyltransferases have been elucidated in fungal and bacterial species. These aromatic prenyltransferases are metal independent although addition of metal ions like Ca2+ and Mg2+ have been reported to have enhanced the catalytic activities of several of these prenyltransferases.DMATS-type prenyltransferases primarily act upon indole derivatives such as tryptophan,food and drain table indole terpenoids, and cyclic dipeptides that contain tryptophan by prenylating these compounds. Reports have demonstrated that DMATS-type prenyltransferases have the ability to prenylate all positions of the indole ring and characterization of its structure have revealed that these prenyltransferase also have the α-β-β-α prenyltransferase folds that ABBA-type prenyltransferases have.Similarly to ABBA-type prenyltransferases, DMATS prenyltransferases show selectivity in prenyl donor, with most enzymes in the family utilizing dimethylallyl pyrophosphate for prenylation, but have great flexibility with regards to prenyl acceptor, capable of prenylating not only those indole derivative previously mentioned but also xanthones, tricyclic and tetracyclic aromatic compounds, and tyrosine.ABBA-type prenyltransferases are found in both fungi and bacteria. They primarily utilize DMAPP and geranyl pyrophosphate as the prenyl donor. All the members of the ABBA-type family of prenyltransferases, except for NphB, do not need metal to assist in catalyzing the reaction.
Although CloQ from Streptomyces roseochromogenes var. oscitans was the first member of the ABBA family of prenyltransferases to be discovered, NphB was the first in the family to have its crystal structure. The crystal structure revealed a structure containing an unique three dimensional α-β-β-α prenyltransferase fold, hence the name ABBA.156 NphB, a member of the ABBA family of prenyltransferases, from the bacteria Streptomyces sp. CL190 was discovered to have non-specific prenylation activity for the formation of CBGA.These prenyltransferases are soluble and are capable of catalyzing the transfer of dimethylallyl , geranyl , or farnesyl prenyl groups onto a diverse set of electron-rich aromatic acceptors. Genome mining for analogs to CloQ, the first gene identified as part of the ABBA family of prenyltransferases, led to the discovery of NphB in Streptomyces sp. CL190.Wildtype NphB is specific in prenyl donor, preferring the geranyl group but is promiscuous with regards to aromatic acceptor although the major substrate is 1,6- dihydroxynapthalene.Wildtype NphB was shown capable of prenylating olivetolic acid toCBGA although at low catalytic efficiencies .Wildtype Nphb was also non-specific in prenylation of OA, capable of producing not only CBGA when reacted with GPP but also 2-O-geranyl olivetolate.Therefore Valliere et al. preformed mutations on NphB to increase specificity for the production of CBGA. They docked the olivetolic acid structure to the NphB crystal structure and then utilizing Rosetta, developed a 22-construct library, constructed the library, and screened for CBGA production. They identified two amino acid mutations that greatly increased specificity to CBGA. From the initial library, then they constructed a focused library and discovered that all but one of the mutations in the focused library had 100-fold higher activity than wildtype NphB with regards to kcat value. Ultimately, they determined that their two best mutations were Y288AG266S and Y288VA232S. Both mutations selectively produced CBGA and both had kcat values 1000-fold higher than the wildtype. Valliere et al. demonstrated that they were capable of producing 1.25g/L of CBGA in a cell free manner utilizing their mutated NphB enzyme.Based on the Tang’s lab collaboration with the Bowie lab, from where Valliere et al developed the mutated NphB enzyme, and the company that he helped found, Invizyne, we were given the mutated NphB enzyme which we used in order to test its ability to prenylate the olivetolic acid analogs in vitro as well as test for functional expression in vivo. We purified the enzyme and performed in vitro assays with our olivetolic acid analogs and GPP. Based on LCMS/HPLC data, NphB was able to prenylate olivetolic acid analogs that we produced as well as other analogs bought commercially, although the final prenylated products were not confirmed by NMR. However, the CBGA product produced by the NphB reaction with olivetolic acid and GPP was confirmed by an analytical standard. Additionally, based on the masses of the in vitro assays, we observed an interesting trend: the shorter alkyl chain variants reacted with NphB and GPP generated not only the C3 geranylated product but also the C3 geranylated product with an O-geranylation. As the alkyl chain length increased, the less appearance of this double geranylated product, with the C3 geranylated product being the major product. We then tested for the functional expression of NphB in vivo in Aspergillus nidulans. We heterologously expressed NphB and the GPP synthase enzyme vrtD from Penicillium aethiopicum , along with Ma_OvaA, Ma_OvaB, and Ma_OvaC in Aspergillus nidulans, expecting to observe the CBGA and CBGA analog products. However, we did not observe any of the geranylated products. We also heterologously expressed Nphb and vrtD with Ti_OvaA, Ti_OvaB, and Ti_OvaC in Aspergillus nidulans and did observe production of CBGA but it was very minute. We additionally expressed CsPT4 as well as an aromatic prenyltransferase from Aspergillus terreus recorded to have prenylation activity in vitro with our platform. However similar to NphB, we observed no production of CBGA. These results led us to postulate that the enzymes are not being properly expressed in A. nidulans, especially engineered NphB, after all a bacterial gene; therefore, we sought to look at its transcription.