Brewing Coral Terpenes – A Yeast Based Approach to Soft Coral Terpene Cyclases

Abstract

Coral terpenes are important molecules with numerous applications. Here, we describe a robust and simple method to produce coral terpene scaffolds at scale. As an example of the approach, here we discover, express, and characterize further klysimplexin R synthases, expanding the known enzymology of soft coral terpene cyclases. We hope that the underlying method described will enable widespread basic research into the functions of coral terpenes and their biosynthetic genes, as well as the commercial development of biomedically and technologically important molecules.

1. Introduction

Octocorals are among the most prolific producers of sesquiterpenoids and diterpenoids known. Many are biomedically relevant, with specific examples including the cytotoxin eleutherobin, the anti-inflammatory pseudopterosin A and the neurotoxin lophotoxin (Fenical et al., 1981; Lindel et al., 1997; Look et al., 1986). There are over 2,500 coral terpenes, many with unique or complex structures and some with hydrocarbon backbones that are specific to coral (MarinLit database, Department of Chemistry, University of Canterbury). Often, corals contain only small amounts of the most bioactive terpenes, and even when production levels are high, it can be difficult to access coral species for compound development. With ocean biodiversity facing environmental threats, a biotechnological approach to discovering and producing these molecules is desirable.

With the advent of next generation sequencing, numerous coral genomes and transcriptomes have become available (Guzman et al., 2018; Ip et al., 2023; Jeon et al., 2019; Jiang et al., 2019; Rivera-García et al., 2019). Recently, we were fortunate to be part of the initial discovery that type I terpene cyclase genes are encoded in coral genomes (Burkhardt et al., 2022; Scesa et al., 2022). For many years, discovery of coral terpene cyclases was prevented by low sequence identity with known cyclases. It is now known that hundreds, or even thousands, of these genes are collectively found in the numerous octocoral species and are an untapped source of chemical diversity for the discovery of new medicines, materials, and biofuels. Terpenoids and terpene cyclase genes are found only in the chemically-defended octocorals, and not hexacorals with hard skeletons, forming a phylogenetically distinct group of proteins. Most likely, these have evolved as a chemical defense against predation (Fenical & Pawlik, 1991).

A particularly interesting group of coral terpene cyclases are the klysimplexin R synthases, which likely produce the eunicellane core found in klysimplexin R from Klyxum simplex, palmonine A from Eunicella verricuosa, and cladiellisin from Virgularia gustaviana (Figure 1 Panel A) (B.-W. Chen et al., 2011; S.-P. Chen et al., 2001; Ortega et al., 1993). Here, we produced a phylogenetic tree to show that klysimplexin R synthases form a unique group of proteins (Figure 1 Panel B). The proteins in these trees include previously reported terpene synthases (Burkhardt et al., 2022; Scesa et al., 2022) and several previously undescribed proteins that we are reporting here. Searching the NCBI databases revealed further novel terpene cyclases from known eunicellane producers, E. verricuosa and V. gustaviana (S.-P. Chen et al., 2001; Ortega et al., 1993). Additionally, in house transcriptomic sequencing, based on reported methods (Scesa et al., 2022), provided two previously undescribed terpene cyclases from the coral Diodogorgia nodulifera, which is not known to make terpenes. Incorporating putative terpene cyclase sequences from E. verricuosa, V. gustaviana and D. nodulifera into the phylogenetic tree indicated that these were all likely to be klysimplexin R synthases (Figure 1). As such, BaTC2 and EcTPS1 are previously reported klysimplexin R synthases while DnTPS1, EvTPS1 and VgTPS1 are reported for the first time here.

Figure 1.

Genetic and chemical diversity of soft coral terpenoid biosynthesis. A) Examples of eunicellane diterpenoids isolated from soft corals. B) Neighbor-joining tree depicting sequence similarity of characterized coral terpene cyclases. An apparent klysimplexin R synthase clade is highlighted (blue). The scale bar represents residue substitutions per site. Unannotated sequences are those previously reported by Schmidt et al and Moore et al.

Providing experimental evidence supporting this assignment requires a robust method to produce terpene synthases, which is one reason that we developed the methods described below. Because corals produce numerous unique and potentially useful compounds using terpene cyclases, this class of proteins presents an exciting avenue for enzymology. However, some key challenges are present, including the bioinformatic identification of candidate genes, the biochemical screening of enzymes to rapidly determine substrate-product relationships, and the scale-up production using the terpene cyclases in a tractable format. The main challenge associated with bioinformatic identification is that terpene cyclases are famously sequence divergent. This technical challenge can be solved by using Hidden Markov Model searches implemented in software such as hmmer3 (Scesa et al., 2022). Prediction of protein coding sequences from coral transcriptomes or genomes followed by use of the hmmsearch function readily produces terpene cyclase hits (PFAM accessions PF19086 and PF03936). Our initial search provided over 300 unique terpene cyclase sequences from 17 different species of coral (Scesa et al., 2022).

After identifying candidate sequences, expression of these proteins in a suitable host is needed for biochemical characterization. While coral terpene cyclases show little sequence identity to terpene cyclases from other organisms, including bacteria, the overall structural fold and domain architecture is quite similar to bacterial terpene cyclases (Burkhardt et al., 2022). As such, they are generally monomeric, soluble proteins which rely on magnesium ions as the only cofactor. These properties make them easy to produce using standard bacterial heterologous expression techniques. As an example, plasmid-based expression of recombinant terpene cyclases in Escherichia coli followed by purification of the over-expressed protein (usually by Ni-NTA chromatography of His-tagged proteins) and in vitro assays involving incubation with oligoprenyl pyrophosphates provides a simple method for analytical-scale screening of the genes (Burkhardt et al., 2022; Lauterbach et al., 2018; Scesa et al., 2022). A limitation is that this approach creates technical bottlenecks, especially during inevitable scale up for producing sufficient quantities of terpenes for spectroscopic analysis. Each protein must be expressed, usually in multi-liter cultures, then carefully purified to provide active protein. While this is routine for any laboratory focusing on enzymology, it can become tedious when dozens or even hundreds of genes need to be rapidly studied. Furthermore, oligoprenyl pyrophosphates from commercial vendors are expensive and pose a financial restriction, as typically quantities of 10 to 100 mg of each precursor are needed for preparative scale enzyme reactions to characterize a single gene. Because of this problem, usually these precursors are made in-house using synthetic chemical techniques. While the synthesis of terpene pyrophosphate precursors was developed decades ago and has continued to be further optimized, it still requires specialized skills and equipment that might not be found in a biochemistry setting (Dixit et al., 1981; Poulter et al., 1988).

Overall, these two factors significantly reduce the ease with which one can characterize a set of terpene cyclases. Considering that hundreds of coral terpene cyclases are currently waiting to be characterized, an approach to rapidly screen these genes using a simplified workflow is highly desirable. The simplest way to do this would be to circumvent the need to purify candidate proteins and synthesize pyrophosphate precursors. Furthermore, if such a technique were readily scalable it would allow for rapid production of material for not just spectroscopic characterization of enzymatic products but would also provide a sustainable source of valuable and unique terpenes.

Herein we describe a synthetic biology platform based on heterologous expression in baker’s yeast Saccharomyces cerevisiae to potentially discover many new terpene cyclase enzymes from corals and produce a wealth of terpene products. There is already considerable literature precedent demonstrating the utility of this approach toward other eukaryotic terpene cyclases (plant and fungal genes), which centers around expressing type I terpene cyclases in a geranylgeranyl pyrophosphate (GGPP) overproducing S. cerevisae strain (DeJong et al., 2006; He et al., 2023; Hu et al., 2020; Zhang et al., 2021). These methods are rapid and simple enough for non-specialist labs to participate in the discovery and production of coral terpenes. The key aspects of the methods include cloning and simple strain engineering, fermentation at analytical and preparative scales, and the purification and spectroscopic analysis of cyclase products. We hope this will prove to be a useful tool which will allow the natural products community to discover new biosynthetic pathways and produce useful amounts of coral terpenoid precursors.

2. Overview

2.1. Strain engineering

Like corals and other animals, yeast cells express the mevalonate pathway (Figure 2, Panel A) at a level sufficient for terpene production (Ro et al., 2006a). For this reason, in many cases coral terpenoids can be produced at detectable levels in yeast by expression of only the cyclase (for sesquiterpene synthases) or co-expression of a GGPP synthase (GGPS) with a cyclase (for diterpene cyclases). This circumvents the need to chemically synthesize oligoprenyl pyrophosphate precursors. Our design of a yeast strain for expression of coral terpene cyclases is based on the use of dual galactose promoters to co-express the GGPS XdCrtE along with the coral terpene cyclase in a standard laboratory yeast strain (DeJong et al., 2006; Milne et al., 2020). Numerous plasmid backbones are available for galactose inducible expression of genes in yeast, including pYES2 (Invitrogen) for expression of one gene or pESC-URA (Agilent) for co-expression of two genes (DeJong et al., 2006). In our initial work, we chose to use the pESC-leu2d plasmid with the leu2d marker which was designed by the Keasling group and is available from Addgene (#20120) (Paddon et al., 2013). Terpene cyclase genes can be amplified from RNA by RT-PCR or, because many terpene cyclases in corals do not have introns, genomic DNA (Scesa et al., 2022). Alternatively, genes can be synthesized, which is preferable because it allows codon optimization (we have not systematically explored the effects of codon optimization). PCR products or synthetic gene fragments can be incorporated into vector backbones using restriction site cloning or Gibson assembly, but usually we elect to order clonal plasmids containing the gene of interest from a commercial vendor. Cloning from synthetic fragments will be described as an example. Terpene cyclase genes are inserted downstream of a galactose inducible promoter (usually P_Gal1) and terminator such as T_CYC1 to control transcription, while the Kozak sequence AAAACA is added just upstream of the start codon to enhance translation (Figure 2, Panels B and C) (Paddon et al., 2013). We have found that sesquiterpene cyclases from corals can be expressed in standard laboratory yeast strains alone without overexpression of any mevalonate pathway genes, overexpression of farnesyl pyrophosphate (FPP) synthase, or repression/knock-out of ergosterol biosynthetic genes. Native FPP levels alone supply sufficient FPP for production of sesquiterpenes in the range of 1 to 100 mg/L. Often, a 100 mL culture will be sufficient for purification and NMR characterization. On the other hand, native GGPP levels are insufficient to produce diterpenes, and overexpression of a GGPS is necessary. In theory, a variety of GGPSs can be used with varying efficacy, but we have elected to use XdCrtE from the red yeast Xanthophyllomyces dendrorhous carotenoid pathway (Alcaíno et al., 2014). This gene was PCR amplified from plasmid pCfB8748 designed by the Borodina group and supplied by AddGene (#126907). The gene is then subcloned by standard restriction digestion and ligation protocol into pESC-leu2d (already containing a coral terpene cyclase) at the PacI/SpeI sites (Milne et al., 2020; Ro et al., 2006b). This site is under the control of P_Gal10 and T_ADH1; the Kozak sequence AAAACA was placed upstream of the start codon. Plasmids were transformed into yeast using a lithium acetate based method (Milne et al., 2020). Co-expression of XdCrtE and a coral terpene cyclase produced up to 98 mg/L in the case of klysimplexin R using EcTPS1 (Scesa & Schmidt, 2023).

Figure 2.

Metabolic engineering in yeast. A) Intercepting the native mevalonate pathway in yeast with heterologous GGPP synthases and terpene cyclases allows the production of coral terpenes in vivo. B) GGPP synthase/terpene cyclase co-expression vector schematic. The vector contains the GGPPS XdCrtE and a coral terpene cyclase and is based on the pESC-leu2d backbone. C) Positioning of genetic parts along the plasmid backbone. A Kozak sequence is placed almost immediately downstream of the GAL1 promoter and upstream of the start site. The CDS is terminated by a stop codon immediately upstream of the CYC1 terminator.

As our chassis, we usually use standard lab strains without additional MVA refactoring including haploid BJ5464 (MATalpha ura3-52 trp1 leu2-delta1 his3-delta200 pep4::HIS3 prb1-delta1.6R can1 GAL), diploid InvSc1 (MATa his3D1 leu2 trp1-289 ura3-52 MAT his3D1 leu2 trp1-289 ura3-52) or haploid YPH499 (MATa ura3-52 lys2-801_amber ade2-101_ochre trp1-Δ63 his3-Δ200 leu2-Δ1). Conveniently, strains can be produced for this method using a pESC plasmid kit from Agilent, which contains the YPH499 strain and a vector that can co-express two genes under galactose control with selection by uracil, leucine, tryptophan or histidine auxotrophic markers. This simple design is useful for initial enzyme characterization by spectroscopic methods and even semisynthesis, but further metabolic engineering of yeast strains has the potential to allow high-level production of coral terpenes in yeast.

2.2. Fermentation, purification and analysis

We use a variety of fermentation formats including initial screens in shake flasks or microtiter plates. Shake flasks are usually sufficient for producing enough terpene product for NMR analysis. Scale-up fermentation is usually tested in a 1-L bioreactor before moving to a 10-L bioreactor. Production in bioreactors usually provides hundreds of milligrams of material for semisynthesis, but will not be described here as plates and flasks are readily accessible in most laboratories, while specialized equipment such as bioreactors are not. Using a bioreactor considerably enhances yields, but using 125-mL shake flasks or 48-deep well plates allows simultaneous screening of dozens or even hundreds of terpene cyclases, respectively. Fermentation conditions are standard in the field, save a few key parameters. Complex media containing yeast extract and peptone as nitrogen sources work best, while as a carbon source a blend of 2 g/L dextrose, 20 mL/L ethanol, 20 mL/L glycerol and 40 g/L galactose provides effective induction and robust growth. Generally, seed cultures are grown overnight in synthetic complete media (with the appropriate nutrient dropped out) before being seeded into complex production media (Brunson John K. et al., 2018; Pompon et al., 1996). Cells can be seeded at an initial low density (OD₆₀₀ of approximately 0.1) then grown to early-log phase (OD₆₀₀ of approximately 1.0) and induced by addition of galactose (with dextrose, glycerol and ethanol serving as the initial carbon source). Alternatively, an early induction method (Nowrouzi et al., 2020) can be used by adding seed cultures directly to production media containing full strength galactose. The latter method is usually more convenient, especially for small scale screens, but often produces a lower terpene yield. A key parameter is temperature, where it is important to express the protein below the standard temperature used for S. cerevisiae. An expression temperature range of 20 to 24 °C is effective, with 22 °C being most commonly used in our work (Nowrouzi et al., 2020). Presumably, higher temperature inhibits stability or proper folding of coral terpene cyclases and is detrimental to production. The addition of polymeric resins such as HP20 or XAD-4 at 2% v/v enhances terpene production considerably (Ignea et al., 2019). In the case of klysimplexin R synthases, addition of HP20 increases production by up to 2-fold. Most likely, this is due to reduction of metabolic negative feedback or enzyme substrate inhibition, as the terpenes appear to be secreted entirely from the cell and adsorbed (strains grown over resin do not contain leftover terpenes in the cell pellet or media after removal of resin, Figure 3). Once the culture has been induced with galactose at the proper temperature and resin added, the culture is grown for about five days with detectable levels of terpenes after two to three days. A side-by-side comparison of yeast cells harboring coral terpene cyclase bearing plasmid or the corresponding empty vector shows reduced growth rates on agar plates, but we find growth of these strains is still sufficiently robust for production.

Figure 3.

Analysis of crude yeast extracts and purified product. A) Thin layer chromatography of purified klysimplexin R (left lane), HP20 extract (middle lane) and cell pellet extract generated after removal of HP20 from cell suspension (right lane) with iodine stain. B) ¹H-NMR spectra of crude HP20 extract (top) and purified klysimplexin R (bottom). Klysimplexin R is clearly observable by TLC (R_f 1.9; hexane/diethyl ether 1:1) and ¹H NMR (olefin resonances near δ_H 5.25 and methyl resonances near δ_H 1.50 and 1.70) in crude extracts of HP20 (CDCl₃, 500 MHz). There is no detectable klysimplexin R in the cells.

Cultures are either extracted with diethyl ether (plate method) or grown over resin (flask/fermenter method) and the resin extracted with acetone. Initial screening in shake flasks or deep-well plates allows analysis by standard GCMS, LCMS, HPLC or TLC, methods that have been exceptionally widely used in the terpene field and for which detailed descriptions are readily available. While GCMS is probably the most informative analytical method for initial screening, we have found TLC to be very convenient and sufficiently sensitive for rapidly determining whether terpenes are being produced. Often the identity of common coral terpene cyclase products can be readily inferred from TLC R_f values, F₂₅₄ activity or stain coloration. Growth in shake flasks at 50–250 mL scale with resin present usually produces sufficient quantities (1–10 mg) of terpenes for spectroscopic analysis by NMR after purification using column or flash chromatography over silica gel. We use standard 2D NMR spectroscopic techniques for structural characterization, primarily HSQC, HMBC, COSY and NOESY spectra along with accompanying 1D ¹H and ¹³C spectra. Fermentation in a 10-L bioreactor has allowed us to produce compounds at a useful scale for semisynthesis, allowing for production of value-added terpenoids (Scesa & Schmidt, 2023). Overall, these methods demonstrate a broad range of utility from initial enzymological screening and substrate/product characterization to a useful lab-scale production method of desirable terpene feedstocks.

3. Protocols

3.1. Strain construction

3.1.1. EQUIPMENT

• Static incubator set to 30 °C.

• Shaking incubator set to 30 °C.

• Water bath set to 42 °C.

• Thermocycler

• Benchtop or floor centrifuge with a rotor that can fit multiwell plates

• Pipettes and sterile tips (multichannel pipettes are best for working in plates)

• Serological pipettes for transferring larger volumes (10 and 50 mL)

• Sterile reagent reservoirs

• Sterile multiwell plates (96-well, 48-well or 24-well) and film seal

• 12-well cell culture plates

3.1.2. MATERIALS

• YPH499 S. cerevisae cells (Agilent, included with various part numbers)

• Chemically competent E. coli DH10β cells with SOC medium (ThermoFisher #EC0113)

• Plasmid pESC-leu2d (Addgene #20120)

• Plasmid pCfB8380 (Addgene #126907)

• Synthetic oligonucleotides encoding terpene cyclase of interest (Twist, Genewiz, etc.)

• Restriction enzymes: BamHI, PacI, SpeI and XhoI and rCutsmart buffer (NEB # R0136, # R0547, #R3133 and #R0146, respectively)

• T4 DNA ligase and buffer (NEB # M0202)

• High fidelity DNA polymerase, such as Q5 polymerase (NEB #M0491), and dNTPs (NEB #N0447)

• Primers (see text for sequence) used in this work were synthesized at the University of Utah Health Science Peptide and DNA Synthesis core

• Plasmid miniprep kit (Qiagen #27115)

• Sterilized LB liquid medium (tryptone 10 g/L, NaCl 10 g/L, yeast extract 5 g/L [premixed, Fisher]) with 25 mg/L carbenicillin (Fisher) made with house deionized water

• Sterilized LB solid agar medium (tryptone 10 g/L, NaCl 10 g/L, yeast extract 5 g/L [premixed, Fisher], agar 10 g/L [Bacto]) with 25 mg/L carbenicillin (Fisher) made with house deionized water

• Sterilized transformation buffer (0.1 M LiOAc [Fisher], 10 mM Tris-HCl [Fisher], 1mM EDTA [Fisher], pH 7.5) made with ultrapure water

• Sterilized PEG solution (40% PEG 3350 [Fisher], 0.1 M LiOAc [Fisher], 10 mM Tris-HCl [Fisher], 1mM EDTA [Fisher], pH 7.5) made with ultrapure water

• Sterilized YPD agar plates (peptone 20 g/L [Fisher], yeast extract 10 g/L [Bacto], dextrose 20 g/L [Sigma], agar 15 g/L [Bacto]) made with house deionized water

• Sterilized YPD liquid medium (peptone 20 g/L [Fisher], yeast extract 10 g/L [Bacto], dextrose 20 g/L [Sigma]) made with house deionized water

• Sterilized synthetic complete solid agar medium with appropriate nutrient dropped out; in this example, leucine (2 g/L yeast amino acid supplement [Sigma], 6.4 g/L yeast nitrogen base [Difco], 20 g/L dextrose [Sigma], 15 g/L agar [Bacto]) made with ultrapure water

3.1.3. Procedure

1. Prepare all buffers and media beforehand and sterilize by autoclaving.

2. Day 0: Generate template and vector plasmids. E. coli harboring pESC-leu2d and pCfB8380 are grown to produce purified plasmid. Stabs (shipped from Addgene) can be used to collect cells which are directly inoculated into LB with carbenicillin (5 ml) and grown overnight at 37 °C.

3. Day 1: Generate sticky end linear vector and insert. Cells are collected by centrifugation at 4,000 rcf and plasmids are purified in the usual fashion using Qiagen spin column purification kit according to the manufacturer’s protocol.

4. XdCrtE is PCR amplified from pCfB8380 using the appropriate primers with inclusion of SpeI and PacI restriction sites (forward primer TAAGCAACTAGTAAAACAATGGATTACGCGAACATCCTCAC; reverse primer TGCTTATTAATTAATCACAGAGGGATATCGGCTAGCTT). PCR is performed using Q5 polymerase based on the manufacturer’s protocol (NEB).

5. The PCR product and pESC-leu2d plasmid are double digested with SpeI and PacI in rCutsmart buffer (30 minute incubation at 37 °C followed by heat denaturation using the NEB protocol).

6. Generate circularized plasmid. The XdCrtE gene is ligated into the vector backbone with T4 ligase. The ligation reaction is performed at 20 μl scale with about 100 ng vector and about 300 ng insert according to the NEB protocol. The reaction is incubated overnight at room temperature.

7. Day 2: Grow out plasmid clones. The plasmid produced is transformed into chemically competent E. coli DH10β cells. This is done by adding ligation reaction mixture (2 μl) to cells (50 μl) on ice and gently mixing by swirling with the pipette tip (do not mix by pipetting). The cells are incubated for 15 minutes on ice, heat shocked at 42 °C for 45 seconds and conditioned by adding SOC medium (250 μl) without antibiotics and incubating for 1 hour at 30 °C. The entire transformation is plated on LB agar containing 25 mg/L carbenicillin and grown overnight at 37 °C.

8. Day 3: Screen plasmid clones. Colonies are picked and seeded into 24-well plates containing LB medium with 25 mg/L carbenicillin and grown overnight at 37 °C. At least six colonies should be screened to ensure positive hits.

   Day 4: Cultures are spun down and plasmids purified by spin column method (Qiagen protocol). The plasmids are screened by by double digest with SpeI and PacI (NEB protocol, 10 μl scale) and analyzed by agarose gel electrophoresis (1% gel with ethidium bromide stain added). At this point, plasmids can be frozen and stored as long as needed to prepare for step 9.

9. Days 5 to 8. Generate TPS vectors. Terpene cyclase genes are synthesized (commercial vendor such at Twist or Genewiz) as a single oligonucleotide fragment and this fragment along with pESC-leu2d harboring XdCrtE are double digested using BamHI and XhoI followed by ligation with T4 ligase, transformation in E. coli and restriction enzyme double digest screening as described above. The sequence of the plasmid is confirmed by whole plasmid sequencing (usually this is done by a third party vendor, such as Genewiz or Plasmidsaurus). At this point, plasmids can be frozen and stored as long as needed to prepare for step 10.

10. Day 9: Generate heterologous yeast strains. A modified lithium acetate method is used for transformation of a large number of yeast strains. To start, YPH499 is streaked onto YPD agar medium and grown for 2 days at 30 °C.

11. Day 11: Colonies from this plate are used to inoculate 10 mL of YPD liquid medium and grown overnight at 30 °C.

12. Day 12: The overnight culture is used to seed 50 mL of YPD liquid medium to an initial OD₆₀₀ of 0.2 then grown at 30 °C for 3–6 hours to an OD₆₀₀ of 1.0.

13. The cells are centrifuged and treated with 50 mL of transformation buffer. The cells are spun down again and resuspended in 5 mL of transformation buffer. The cell suspension is poured into a reagent reservoir and 50 μL transferred to each well of a 96-deep well plate using a multichannel pipette. When stored at 4 °C, these cells are competent for at least 3 days.

14. Sequenced TPS plasmids for screening are prepared in a 96-well plate in nuclease free water at a concentration of about 200 ng/μL. Using a multichannel pipette, 5 μL (about 1 μg of plasmid DNA) of plasmid solution is transferred to each well of the treated cell suspension and mixed by pipetting. Next, 350 μL of PEG solution is added and the plate is sealed. The plate is inverted several times, then incubated at 30 °C for 1 hour with mixing by inverting every 10 to 15 minutes. The plate is then moved to a 42 °C bath and incubated for 30 minutes with periodic mixing by inverting the plate. Cell suspensions are then transferred to 12-well cell culture plates containing solid synthetic complete agar medium and dried in a laminar flow hood till the cell suspension forms a film. The plates are then incubated for 3 days at 30 °C, after which colonies will form. Plates are usually viable for several months when stored at 4 °C. Glycerol stocks are made for long term storage.

Alternatives:

We generally prefer to purchase genes in clonal form from a commercial vendor. These coral genes are synthesized and inserted into the pESC-leu2d backbone containing XdCrtE. The plasmids can be delivered in 96-well plate format for rapid screening. For screening a small number of genes or targeted production of compounds, the plasmids can be transformed into yeast using the commercially available Sc EasyComp kit. Individual petri dishes can be used for growth on solid agar when a small number of strains are to be screened or targeted production of a terpene is desired using a known enzyme.

3.2. Analytical scale fermentation and screening

3.2.1. EQUIPMENT

• Shaking incubator

• Pipettes and sterile tips (multichannel pipettes are best for working in plates)

• Speed vac or 96 well evaporator/concentrator manifold (Chemglass part# OP-5020-01; see Figure 4)

• Benchtop or floor centrifuge with a rotor that can fit multiwell-plates

• Instrument for analytical chromatography (GC, HPLC, etc. coupled to appropriate detectors)

Figure 4.

Illustration of gas drying manifold used for evaporating solvent from plates.

Alternatives:

Instead of multiwell plates, cultures can be grown in small culture flasks or test tubes. Organic extracts can be evaporated using a rotary evaporator instead of drying in a plate, but with slower throughput. While GC and LC methods offer ideal throughput and sensitivity, TLC can also be used for analysis.

3.2.2. MATERIALS

• Agar plate with colonies of the desired yeast strain

• Sterilized synthetic complete liquid medium with appropriate nutrient dropped out, in this example, leucine (2 g/L yeast amino acid supplement [Sigma], 6.4 g/L yeast nitrogen base [Difco], 20 g/L dextrose [Sigma]) made with ultrapure water

• Sterilized liquid production medium with galactose (15 g/L yeast extract [Bacto], 15 g/L peptone [Fisher], 2 g/L dextrose [Sigma], 20 ml/L ethanol [200 proof, Fisher], 20 ml/L glycerol [Fisher] 40 g/L galactose [Goldbio]) made with house deionized water

• Diethyl ether, anhydrous (Fisher)

• Isopropanol ACS grade (Fisher)

• Corning^™ microplate Aluminum Sealing Tape

• Sterile reagent reservoirs

• Sterile multiwell-plates (96-well, 48-well or 24-well)

Alternatives:

Hexane can be used instead of isopropanol and ether.

3.2.3. Procedure

Screening yeast strains for biochemical characterization of coral terpene cyclase genes is performed in plates, including fermentation and extraction steps. Yeast cultures are induced by addition of seed cultures to galactose containing medium and extracted with solvent after fermentation. This example uses 48-well plates with a 4 mL volume. A strain harboring a plasmid with the same selection marker (i.e., leucine) but without a terpene cyclase gene is essential as a negative control and should be grown, extracted and analyzed alongside strains of interest.

1. Day 0: Seed cultures. Liquid synthetic complete medium containing 2% dextrose is poured into a reagent reservoir and 200 μL is transferred to each well of a 48-well plate using a 1.2 mL, 8-channel pipette.

2. Colonies of each strain to be analyzed are picked from an agar plate using a single channel pipette and transferred to liquid medium in a well. It is important when screening many strains to stay organized, adding only a single strain to each well, and recording the position of each strain using a plate map.

3. After addition of the strains to desired wells, the plate is sealed with aluminum sealing tape and incubated overnight in a shaking incubator set to 30 °C and 220 RPM. Strong shaking is essential to keep yeast cells suspended.

4. Day 1: Induction. After shaking the plate overnight, the cultures should have reached an OD₆₀₀ near 5.0. It is not necessary to check the OD₆₀₀ of the cultures spectrophotometrically, but it is good practice to manually inspect the cultures and ensure adequate growth. When the cultures are visibly opaque, indicating cell growth, the cells are induced by addition of galactose. This is done by transferring 1 mL of liquid production medium containing 4% galactose from a reagent reservoir to the culture plate using a 1.2 mL, 8-channel pipette. This is done by wiping the top of the sealed plate with 70% ethanol, then piercing the seal with the pipette tips during addition. This allows medium to be added rapidly while keeping track of which wells have been induced and providing aeration.

5. The plate with induced cell cultures is then placed in a shaking incubator set to 22 °C and 220 RPM and shaken for 3 to 5 days. We prefer to incubate for 5 days. Adequate shaking and aeration are essential for terpene production.

6. Day 5: Extraction and analysis. Cultures are extracted with an organic solvent, such as diethyl ether. Diethyl ether efficiently extracts terpenes and is easy to evaporate, but hexane can be used as well. First, take the plate from the incubator and remove the seal, as the seals often contain glue which causes contamination. Using a 1.2 mL, 8-channel pipette, 2 mL of solvent is added to each well by making two additions. When working with volatile liquids, it is important to prime the pipette by repeatedly pipetting the liquid to fill the headspace with solvent vapor and equalize the pressure and ensure accuracy.

7. The culture is mixed by carefully pipetting up and down to create an emulsion. The plate is wrapped with aluminum foil to prevent evaporation, cooled on ice for 15 minutes, then spun at about 5,000 g at 4 °C to break the emulsion. It is important to keep the work quickly and keep the sample covered and cold to minimize solvent evaporation.

8. Once the layers have been separated by centrifugation, 1.5 mL of the upper organic layer is pipetted off, in two portions, into a clean and dry well plate.

9. The solvent is evaporated under air in a fume hood using an air-drying apparatus consisting of a 96-needle gas drying manifold (Figure 4) attached to an in-house compressed air line. The needles of the manifold are directed into the wells of the plate and a gentle stream of air is applied from the needle tip. Air is passed over the sample until no liquid is present. To avoid oxidation of unstable terpenes or evaporation of volatile terpenes, do not dry longer than necessary to remove all organic solvent.

10. Once the extraction solvent has been evaporated, the samples are dissolved in an appropriate solvent for chromatographic analysis (hexane for GC and TLC, isopropanol for LC methods). GCMS utilizing electron impact (EI) ionization is the most common technique for analyzing terpene hydrocarbons, as most terpenes are well resolved by GC and can be structurally identified by well understood, characteristic fragmentation pathways. As an example, LC-electrospray mass spectrometry for analysis of the klysimplexin R synthases is used (Figure 5). The volatile reaction products of EvTPS1, VgTPS1 and DnTPS1, when analyzed by LCESIMS, reveal a chromatographic peak corresponding to m/z 273.3, representing the [M-OH]⁺ ion. Presumably, this ion forms in the ESI source due to rapid E1 elimination of water from the protonated complex [M+H]⁺. The chromatographic peak matches one observed for purified klysimplexin R and is not present in the negative control.

Figure 5.

LCMS traces of crude yeast extracts and purified klysimplexin R. Extracts were generated from either a strain expressing a klysimplexin R synthase or yeast harboring empty vector (negative control). Traces represent EIC at m/z 273.2, detecting the [M-OH]⁺ ion. EvTPS1 and empty vector traces were scaled 10x to highlight low abundance ions (none were detectable in the negative control).

Alternatives:

It is better to dry extracts under an inert gas such as nitrogen or argon instead of air, but this is not essential for short drying periods. To increase the evaporation rate, the plate can also be set in a water bath warmed to 40 °C, under a stream of air.

Safety considerations:

Diethyl ether is highly flammable and volatile. Do not allow solvent vapors to accumulate near an ignition source, such as flame or electrical equipment that makes heat or sparks. Keep solvent solutions covered to avoid release of vapor and work in a fume hood.

3.3. Preparative scale fermentation and screening

3.3.1. EQUIPMENT

• Refrigerated shaking incubator set to 30 °C (later set to 22 °C)

• Pipettes

• 2.8 L baffled Fernbach flasks

• 125 mL (sterile) and 250 mL Erlenmeyer flasks

• 20 mL beaker

• Round bottom flasks (500 and 250 mL)

• Rotary evaporator

• Glass chromatography column

• Buchner funnel

• Separatory funnel

• UV lamp

3.3.2. MATERIALS

• Agar plate with colonies of the desired yeast strain

• Sterilized synthetic complete liquid medium with appropriate nutrient dropped out, in this example, leucine (2 g/L yeast amino acid supplement [Sigma], 6.4 g/L yeast nitrogen base [Difco], 20 g/L dextrose [Sigma]) made with ultrapure water

• Sterilized liquid production medium without galactose (15 g/L yeast extract [Bacto], 15 g/L peptone [Fisher], 2 g/L dextrose [Sigma], 20 ml/L ethanol [200 proof, Fisher], 20 ml/L glycerol [Fisher]) made with house deionized water

• Galactose solution (40 g/L galactose [Goldbio] in house deionized water)

• Sterilized aluminum foil

• Cheesecloth

• Acetone ACS grade (Fisher)

• Diethyl ether anhydrous, BHT stabilized (Fisher)

• Hexane ACS grade (Fisher)

• Isopropanol ACS grade (Fisher)

• Diaion^™ HP20, synthetic adsorbent resin (MilliporeSigma)

• Anhydrous sodium sulfate (Fisher)

• Silica gel 100-200 mesh (Fisher)

• TLC plates, silica gel 60 F₂₅₄ (MilliporeSigma)

• Iodine vapor stain (silica gel impregnated with 1% iodine crystals and 0.1% iron III chloride)

3.3.3. Procedure

For confirming the identity of terpene cyclase products deduced from phylogenetic analysis or mass spectra as well as de novo characterization of novel enzymatic products, material must be prepared in sufficient quantity for purification and spectroscopic analysis. We will describe growth at the 1 liter scale in shake flasks, since this method frequently provides sufficient amounts of compounds for further chemical analysis and uses readily available supplies and materials.

1. Day 0: Seed cultures. Single colonies of a yeast strain from an agar plate are picked and inoculated into 20 mL of synthetic complete medium in a sterile 125-mL Erlenmeyer flask and covered with sterile aluminum foil.

2. The cultures are placed in a shaking incubator set to 220 RPM and 30 °C. These cultures are grown overnight to an OD₆₀₀ of about 5.0. Flasks are preferred over culture tubes because high density cultures tend to settle in culture tubes.

3. Day 1: Grow out and induction. The entire seed culture is transferred to 1 L of production medium, providing an OD₆₀₀ of 0.1.

4. This culture is grown in a shaking incubator at 30 °C and 220 RPM for about 6 to 8 hours to an OD₆₀₀ of 1.0.

5. The culture is cooled down to 22 °C and induced by addition of 100 mL of 40% galactose solution. After induction, 20 mL of HP20 resin suspended in ethanol is added (slurry can be measured in a beaker). The culture is grown at 22 °C for five days.

6. Day 5: Extraction, purification and analysis. After 5 days, the HP20 resin is collected by filtering the cell suspension through cheesecloth. The resin beads are washed with water and dried on a Buchner funnel.

7. The HP20 beads are then suspended in 200 mL of acetone in a 250 mL Erlenmeyer flask. The extraction is mixed by periodic swirling of the flask.

8. The resin is allowed to settle to the bottom of the flask, and the acetone layer is decanted into a 500 mL round bottom flask and evaporated on a rotary evaporator at 150 Barr and 40 °C.

9. The viscous residue is partitioned between diethyl ether and water (100 mL each), using a separatory funnel to remove the organic layer. The organic layer is dried over sodium sulfate and evaporated in a 250 mL round bottom flask on a rotary evaporator at 400 Barr and 40 °C.

10. At this point the extract is checked by NMR or TLC.

11. The organic extract is purified by flash chromatography using silica gel as stationary phase and hexane as mobile phase. Terpene alcohols are purified by adding up to 25% isopropanol to the mobile phase with hexane. The size of the column depends on the product profile. High yielding strains require wider columns while more complex mixtures of products require long columns to achieve separation. In general a 30 cm x 1 cm glass column readily purifies the product from a 1 L culture, providing several milligrams of a terpene product. The column is eluted with 200 mL of mobile phase with collection of 10 mL fractions. Separation is monitored by TLC using UV or iodine vapor staining to visualize the plate (Figure 3, Pane A). Purified compounds are characterized by NMR spectroscopy, including 1D ¹H and ¹³C spectra as well as 2D COSY, HSQC, HMBC and NOESY spectra (Figure 3, Panel B). As an example, a 1 L culture each of DnTPS and EvTPS provids 36 mg and 4 mg of purified klysimplexin R after extraction and purification by silica gel chromatography.

Outlook

Episomal expression of coral terpene cyclases in yeast provides a means to rapidly screen for enzymatic activities and facilitates scaled up production of terpenes for spectroscopic characterization, as well as other purposes. Here we describe methods to overexpress klysimplexin R synthases as an example. Klysimplexin R represents the starting point for a variety of diterpenoids with medicinal potential, as its eunicellane core is the likely precursor to over 200 different diterpenoids, mostly from corals. Moreover, these and related enzymes represent a wealth of biochemical diversity. Therefore, the methods presented here have the potential to rapidly advance our search for terpene cyclases, especially terpene cyclases from animals such as soft corals.