A teaching protocol demonstrating the use of EasyClone and CRISPR/Cas9 for metabolic engineering of Saccharomyces cerevisiae and Yarrowia lipolytica

ABSTRACT We present a teaching protocol suitable for demonstrating the use of EasyClone and CRISPR/Cas9 for metabolic engineering of industrially relevant yeasts Saccharomyces cerevisiae and Yarrowia lipolytica, using β-carotene production as a case study. The protocol details all steps required to generate DNA parts, transform and genotype yeast, and perform a phenotypic screen to determine β-carotene production. The protocol is intended to be used as an instruction manual for a two-week practical course aimed at M.Sc. and Ph.D. students. The protocol details all necessary steps for students to engineer yeast to produce β-carotene and serves as a practical introduction to the principles of metabolic engineering including the concepts of boosting native precursor supply and alleviating rate-limiting steps. It also highlights key differences in the metabolism and heterologous production capacity of two industrially relevant yeast species. The protocol is divided into daily experiments covering a two-week period and provides detailed instructions for every step meaning this protocol can be used ‘as is’ for a teaching course or as a case study for how yeast can be engineered to produce value-added molecules.


Introduction
Metabolic engineering is a rapidly emerging field, which draws inspiration from multiple engineering disciplines to design and alter metabolic pathways for useful purposes (Stephanopoulos, Aristidou, & Nielsen, 1998). The budding yeast Saccharomyces cerevisiae (S. cerevisiae) is frequently used as a host organism for the engineering of metabolic pathways, particularly for the production of valueadded molecules, with a large number of reports describing the engineering of this organism for an incredibly diverse range of molecules (Borodina & Nielsen, 2014). Yarrowia lipolytica (Y. lipolytica) is an oleaginous yeast increasingly used for the production of biofuels and chemicals (Darvishi, Ariana, Marella, & Borodina, 2018). Y. lipolytica is especially well suited for industrial production of oleochemicals due to its ability to accumulate lipids up to 70% of dry cell weight as well as high flux through tricarboxylic acid cycle intermediates and cellular precursors such as acetyl-CoA and malonyl-CoA (Beopoulos & Nicaud, 2012;Marella, Holkenbrink, Siewers, & Borodina, 2018;Markham & Alper, 2018).
A good example of a value-added molecule that was investigated in this protocol is the vitamin A precursor β-carotene. β-Carotene is increasingly used in food and feed additives, cosmetics, and health supplements (Li, Sun, Li, & Zhang, 2013), but is predominantly derived from chemical synthesis (Lange & Steinbüchel, 2011) using petro-chemically derived substrates (Ribeiro, Barreto, & Coelho, 2011). Biotechnological production of β-carotene via yeast fermentation has the potential to deliver a low-cost, environmentally friendly alternative to chemical synthesis. To this end, there has been a multitude of reports of the engineering of various microorganisms to enable β-carotene production (Do Quynh Nguyen et al., 2012;Lange & Steinbüchel, 2011;Li et al., 2013;Rodríguez-Sáiz, Sánchez-Porro, De La Fuente, Mellado, & Barredo, 2007;Ronda et al., 2015;Verwaal et al., 2007;Yang & Guo, 2014).
Here we describe a detailed protocol outlining all steps required to introduce β-carotene production in both S. cerevisiae and Y. lipolytica. The protocol details assembling DNA into integration expression cassettes using the EasyClone method Jessop-Fabre et al., 2016), and uses the red phenotype from ADE2 knockout (Ugolini & Bruschi, 1996) to demonstrate how to generate and assemble guide RNA vectors for CRISPR-Cas9 genome targeting . All necessary plasmids have been deposited at Addgene and all yeast strains have been deposited at Euroscarf.
β-carotene biosynthesis will be introduced into yeast following the schematic outline in Figure 1. Via the ergosterol pathway, β-carotene precursor molecule trans, trans-farnesyl diphosphate (FPP) is produced using native yeast metabolism. Heterologous genes sourced from Xanthophyllomyces dendrorhous (X. dendrorhous) then catalyze the conversion of FPP to β-carotene. Expression of heterologous genes XdcrtE, XdcrtYB and XdcrtI results in de novo β-carotene production from glucose. To introduce the concept of boosting precursor supply, the protocol describes modifications to increase mevalonate formation by overexpressing a truncated version of HMG1 (tHMG1) in S. cerevisiae (previously shown to increase flux through the pathway (Verwaal et al., 2007)), and a nontruncated version of HMG1 in Y.lipolytica (since the truncated variant was shown to be less-efficient ). To introduce the concept of alleviating rate-limiting steps in the metabolic pathway, the protocol also describes the introducing of an additional copy of a known rate-limiting step in β-carotene biosynthesis (XdcrtI) (Verwaal et al., 2007).  (Verwaal et al., 2007)), Red: expression of native Y. lipolytica gene, Blue; expression of heterologous genes from X. dendrorhous. HMG-CoA; 3-hydroxy-3-methyl-glutaryl-CoA, IPP; Δ3- From a genome editing perspective, one key advantage of S. cerevisiae is its inherent ability to reliably integrate heterologous pieces of DNA through homologous recombination. This is especially useful when combined with the CRISPR-Cas9 system, which promotes the homologous recombination by introducing double-strand DNA breaks. Double-strand DNA breaks also serve as a selection, where only the cells that repair the break can survive. This eliminates the need to use selection markers (Norville et al., 2013). This is demonstrated in the protocol by the simultaneous integration of expression cassettes at up to three separate loci in the S. cerevisiae genome, allowing us to take a wild-type S. cerevisiae strain and convert it into a β-carotene producer in a single transformation step.
While new metabolic engineering tools to facilitate genetic modifications in Y. lipolytica have emerged in the past years, it is still challenging to obtain multiple genome edits in a single transformation event with good efficiency . This can be due to the lower transformation efficiency of this yeast and/or due to lower efficiency of homologous recombination. Therefore, the engineering strategy used for Y. lipolytica in this protocol instead relies on single genome integration of expression cassettes. The parental Y. lipolytica strain used in this protocol (ST8889) already has the β-carotene pathway integrated in the genome and thus already produces βcarotene, and additionally, contains a KU70 deletion (ku70∆) to enhance the frequency of homologous recombination by disrupting the Ku70-Ku80 homodimer complex responsible for catalyzing nonhomologous end joining (Kretzschmar et al., 2013). To demonstrate the concept of boosting precursor supply and overcoming a rate-limiting step, three plasmids containing the genes YlHMG1 and XdcrtI are constructed and integrated separately in the parent strain. An overview of the strain construction genealogy for both S. cerevisiae and Y. lipolytica is shown in Figure 2.

S. cerevisiae, Y. lipolytica, and E. coli strains
Parental S. cerevisiae and Y. lipolytica strains required for this protocol are available from Euroscarf.
A full list of strains (including those constructed in this protocol) can be found in the supplementary materials.
• Chemically competent DH5α Escherichia coli cells for plasmid transformation and propagation.

Plasmids
All plasmids required for this protocol can be obtained from Addgene. A full list of plasmids (including those constructed in this protocol), and their respective Addgene identifiers can be found in the supplementary materials. • HPLC machine (Thermo Fisher Scientific or similar) with a Discovery HS F5 150 mm x 2.1 mm column (particle size 3 mm)

Liquid growth medium
• LB medium: For preparation of 1 L LB medium, mix 10 g Bacto tryptone, 5 g Bacto yeast extract and 10 g sodium chloride and fill up to 1 L with demineralized water. For solid medium add 2% (w/v) Bacto agar. Heat sterilize for 20 min at 121 °C. After sterilization add antibiotics as required.
• YPD (Yeast peptone dextrose) medium: For 1 L YP medium, mix 5 g Bacto yeast extract, 10 g Bacto peptone and fill up to 1 L with demineralized water. For solid medium, add 2% (w/v) Bacto agar. Heat sterilize for 20 min at 121 °C. After sterilization, add sterilized glucose solution to a final concentration of 2% (w/v) (YPD), add antibiotics as required.
• SM (synthetic medium): For 1 L of synthetic medium, start with 750 mL of demineralised water and add 5 g ammonium sulphate [(NH4)2SO4], 3 g monopotassium phosphate [KH2PO4] and 0.5 g magnesium sulphate heptahydrate [MgSO4.7·H2O] and add trace elements according to (Verduyn, Postma, Scheffers, & van Dijken, 1992) and described below. Dissolve salts and set the pH to 6.0 with 2 M potassium hydroxide [KOH], add demineralized water to reach a final volume of 1 L and heat sterilize for 20 min at 121 °C. After sterilization add vitamins according to (Verduyn et al., 1992) and described below.
• Trace metal solution for synthetic media: For 1 L of trace metal solution.

Chemical
Amount ( dissolve all chemicals listed above except for EDTA one-by-one while maintaining the pH at 6.0 in 900 mL H2O. Add the EDTA and gently heat the solution until completely dissolved. Adjust the final pH to 4.0 and the volume to 1 L. Heat sterilize for 20 min at 121 °C and store at 4 ºC.
Add 2 mL per 1 L of synthetic medium.
• Vitamin solution for synthetic media: For 1 L of vitamin solution. Dissolve biotin in 20 mL 0.1 M NaOH then add 900 mL water. Adjust pH to 6.5 with HCl and add the remaining vitamins. Re-adjust the pH to 6.5 just before and after adding myo-inositol. Adjust to a final volume of 1 L. Filter sterilize and store at 4 ºC. Add 1 mL of vitamin solution per 1 L of synthetic media

Solid growth medium
Prepared as above for liquid growth medium but with the addition of 20 g L -1 agar

Protocol Molecular biology preparation before starting the course
We recommend preparing the following before commencing with the experimental protocol: 1. Isolate the plasmids listed in Table 1 gRNA plasmids for targeting genomic integration sites (approx. 2 μg per person required) Episomal yeast expression plasmids (approx. 1 μg per person required) 7. Since most steps in this protocol are time demanding we also strongly recommend preparing backups of each step in the protocol should any experiment fail.

Monday (Week 1, Day1)
Today students will PCR amplify DNA biobricks and assemble them into EasyClone-MarkerFree plasmids for transformation into yeast Jensen et al., 2014). DNA parts to amplify include the heterologous β-carotene biosynthetic genes XdcrtI, XdcrtE and XdcrtYB (Verwaal et al., 2007), a truncated version of native S. cerevisiae HMG1 gene (tHMG1), reported to remove feedback regulation (Verwaal et al., 2007), the native HMG1 gene from Y. lipolytica (YlHMG1), and strong constitutive yeast promoters for expression of each gene. While plasmids containing gRNA cassettes for targeting integration fragments into the yeast genome are supplied, to demonstrate how to generate and assemble gRNA expression plasmids, students will also PCR amplify an ADE2 gRNA expression cassette and assemble it into a gRNA plasmid backbone using the EasyClone system. Table 3.) will be constructed to demonstrate how they impact β-carotene production. EasyClone-MarkerFree is a DNA assembly and genome integration method based on Uracil-Specific Excision Reagent (USER) cloning and yeast homologous recombination at pre-defined genomic landing pads. Briefly, USER cloning is a directional cloning technique where a short pre-defined (6-10 bp) homology arm starting with a single deoxyuridine (dU) residue is placed at the 5' end of each primer. After amplification of the DNA fragments to be fused with these primers, the dU residue is cleaved by a USER mix to generate complementary 3' single stranded overhangs that can be fused together (Nour-Eldin, Geu-Flores, & Halkier, 2010). The

Different gene combinations (outlined in
EasyClone system uses standardized overhangs to fuse up to 2 genes of interest to a specific promoter and terminator in a single plasmid. Furthermore, plasmid backbones contain long (500 bp) homology arms flanked by NotI restriction sites allowing the USER assembled expression cassette to integrate at pre-defined landing pads in the genome using yeasts own homologous recombination machinery (Jensen et al., 2014). These genomic landing pads have been previously validated for growth impairment and strong expression and are placed between essential genetic elements resulting in celldeath should loop-out of the inserted fragments occur Mikkelsen et al., 2012). Marker free selection is achieved using CRISPR-Cas9 where a targeted double-strand break (DSB) at the genomic landing pads creates a lethal event that can be repaired by integration of the expression cassette (Jessop-Fabre et al., 2016). A schematic overview of the cloning and genomic integration procedure is shown in Figure 3. 1. PCR amplify biobricks for EasyClone and gRNA plasmid assembly listed in Table 2. according to the scheme listed below. Up to 500 ng of each biobrick is required for EasyClone plasmid assembly so we recommend doing some biobrick amplifications 2x.    Table 2. Negative image given for clarity. Table 3. using the EasyClone method according to Jensen et al., 2014) and outlined below. We recommend including controls where only the plasmid backbone is added to account for false positives caused either by undigested plasmid or re-ligation of the plasmid.

Assemble plasmids outlined in
a. Prepare reaction mixtures as follows:  Alternate protocol.
1. If the total number of E. coli colonies on the transformation plate is low it may be due to poor chemical competency. To increase the transformation efficiency, after heat shock, incubate cells in 500 μL of LB or SOC media before plating.

Tuesday (Day 2)
Today students will confirm correct assembly of EasyClone plasmids by colony PCR (cPCR).
Students will then set up overnight E. coli cultures of clones to further confirm correct assembly by Sanger sequencing. Table 4 to confirm correct assembly of EasyClone plasmids. required. An example of a successful cPCR confirming correct EasyClone assembly is given in Figure 5.    Table 5, note that there will also be a band at ~2800 bp corresponding to the plasmid backbone. Also, note that pCfB8622 (NatMX_ADE2_gRNA) is an expression plasmid not an integration plasmid and thus does not contain NotI sites for digestion.

Wednesday (Day 3)
Today students will purify PCR confirmed EasyClone plasmids and prepare them for transformation into yeast. Students will also prepare plates for transformation and inoculate agar plates with the parental yeast strains.  3. Prepare yeast parental strains for transformation by restreaking ST7574 onto YPDG and ST8889 onto YPD agar plates. Incubate plates at 30 °C for 2 days.
Alternate protocol 1. If time and resources allows you can additionally confirm correct plasmid assembly by DNA sequencing the plasmids using the same primers listed in Table 4. Follow the protocol given by your chosen sequencing provider. Map the sequencing reads to the plasmid maps provided in the supplementary materials.

Thursday (Day 4)
Today students can make sure everything is ready for transformation of the DNA constructs into yeast. The day can also be used to repeat any experiments that may have failed in the previous days.

Friday (Day 5)
Today students will transform the DNA constructs they prepared during the week into yeast and incubate over the weekend to allow single colonies to form. Expression cassettes will be introduced according to the scheme outlined in Table 6.  e. Add DNA parts for transformation into S. cerevisiae strain ST7574 as outlined in  e. Transfer the required volume for each transformation to a sterile Eppendorf tube, centrifuge for 5 min at 3000 g at room temperature and remove the supernatant.
f. Add 1000 ng of linearized integration vector DNA and 500 ng of plasmid DNA to the cell pellet according to the scheme outlined in Table 6.
g. Gently resuspend the cell pellet in the following transformation mix.
i. Spin down the cells for 5 min at 3000 g at room temperature and resuspend in 500 µl YPD. Incubate at 30 °C for at least 2 h with gentle shaking.
j. Spin down the cells for 5 min at 3000 g at room temperature, remove 450 µl of YPD media and resuspend the cell pellet in the remaining 50 µl. Plate cells on YPDN.
k. Incubate at 30 °C for approx. 3 days until single colonies are visible.

Monday (Week 2, Day 8)
Over the weekend, single colonies should appear on the transformation plates. For the introduction of genes involved in β-carotene synthesis, successfully integrated colonies should appear shades of yellow, orange or red due to the buildup of β-carotene and its intermediates (Verwaal et al., 2007), an example of a transformation plate is shown in Figure 6. For the ADE2 knockout transformation, successful clones should appear red due to the buildup of a red pigment from the adenine biosynthesis pathway (Ugolini & Bruschi, 1996). Today students will do colony PCR (cPCR) to confirm the genotype of their transformants. Because some transformations introduced DNA at two or three different locations in the genome, multiple PCR reactions will need to be performed for each clone tested. Once students have confirmed which clones have the correct genome modifications, they can inoculate pre-culture media with the correct clones to start a production assay the following day to determine how the different engineering strategies impact β-carotene production. Diagnostic primers used for genotyping bind outside the genomic integration site (Out Fwd and Out Rev) and bind in the integration fragment (In Rev). Pathway.
a. Prepare cPCR reactions according to the outline in Table 7. e. Analyze the samples on 1% agarose gel (for corresponding PCR product size see Table 7). Include a GeneRuler 1 Kb DNA ladder. A gel image showing a successful cPCR and correct integration of each cassette is given in Figure 7.
a. Inoculate 1 mL YPD pre-culture media with at least one correct clone from each transformation in a 10 mL pre-culture tube (Greiner) and incubate overnight 30 °C with shaking at 200 RPM.  Table 7 with each integration site given below.
Negative image given for clarity. 2. Run a PCR to confirm correct genome integration described above.
3. Since both parental strains are prototrophic both the pre-culture and subsequent production assay can be performed in synthetic media (SM). While the strains will likely not grow as fast, the clear media will allow the carotenoid pigments to be more visible.

Tuesday (Day 9)
By today, you should have yeast pre-cultures of various colors of yellow, orange and red. Today you will start a production assay to measure how much β-carotene is produced by each strain and use these results to assess the relative success of each genomic integration. 2. If 24-deep well plates are not available you can cultivate the strains in e.g. shake-flask or any other system with a minimum volume of 2.5 mL.

Wednesday (Day 10)
No experimental work is scheduled on these days. We suggest spending the time showing students the in silico design of gRNA plasmids using online tools such as CRISPy (http://staff.biosustain.dtu.dk/laeb/crispy_cenpk/) . This protocol only scratches the surface of possible engineering targets to boost flux towards β-carotene, a further suggestion would be to ask students to perform a literature search to identify additional metabolic engineering strategies and make a workflow for creating these additionally boosted strains.

Thursday (Day 11)
Today students will prepare samples to measure β-carotene production. Unlike many other valueadded molecules produced by engineered yeasts, β-carotene is not exported in significant amounts from the cell and thus accumulates intracellularly. In order to quantify β-carotene production, students will first need to lyse the yeast cells to extract the intracellular product. A further complication is that β-carotene is highly insoluble in water, so extraction is performed using ethyl acetate. Note that ethyl acetate is highly volatile and flammable so care must be taken when handling. Perform all steps in a fume hood using nitrile gloves and eye protection. Read the SDS before use.
The extraction protocol used is according to  and is described below.
a. Transfer 1 mL of the cultivation broth into a pre-weighed 2 mL microtube.
b. Centrifuge at 10,000 g for 5 min. Remove the supernatant and place the tubes.
containing the biomass pellets in the incubator at 60 °C for 24 h. After 24 h weigh the tubes on an analytical scale.
b. Centrifuge the tubes at 10,000 g for 10 min and remove the supernatant.
c. Add 0.5 mL of 0.5-0.75 mm glass beads to each tube followed by the addition of 0.5 mL of ethyl acetate supplemented with 0.01% 3,5-di-tert-4-butylhydroxytoluene (BHT). BHT is added to prevent carotenoid oxidation.
d. Disrupted the cells using the Precellys R 24 homogenizer (Bertin Corp.) in four cycles of 5,500 RPM for 20 seconds. Cool the tubes by placing on ice for 1 min in between each lysis cycle. After disruption, centrifuge cells for 10 min at 10,000 g.
e. For quantification of β-carotene by HPLC, transfer 100 µL of the solvent fraction to HPLC vials.
f. Evaporate the 100 µL of ethyl acetate extract in a rotatory evaporator (SpeedVac™) for at least 45 min. Re-dissolve the dry extracts in 1 mL of 99% ethanol + 0.01% BHT.
a. Prepare an initial 30 mg L -1 stock solution in approx. 5 mL.
b. Prepare dilution standards according to the scheme below. 4. Run samples on HPLC.
a. The extracts can now be analyzed by HPLC (Thermo Fisher Scientific c, Waltham, USA). The machine is equipped with a Discovery HS F5 150 mm x 2.1 mm column (particle size 3 mm). For β-carotene analysis, the column oven temperature is set to 30 °C. The flow rate is set to 0.7 mL min -1 with an initial solvent composition of 10 mM ammonium formate 157 (pH = 3, adjusted with formic acid) (solvent A) and acetonitrile (solvent B) (3:1) until minute 2.0. Solvent composition is then changed at minute 4.0 following a linear gradient until % A = 10.0 and % B = 90.0. The solvent composition is kept until 10.5 minutes when the solvent returns to initial conditions and the column is re-equilibrated until 13.5 minutes. The injection volume is 10 µL .
b. The peaks obtained from the sample analysis can now be identified by comparison to prepared standards. Check the peaks for each sample to confirm the right integration of the peak areas. If necessary, do the integration manually. β-carotene is detected at a retention time of approx. 7.6 minutes, by measuring absorbance at 450 nm.

Alternate protocol
If access to an HPLC machine is infeasible, quantification can also be achieved using a plate-reader or cuvette based spectrophotometer.
1. Follow the protocol as described above, after re-dissolving the dry extracts in 1 mL of 99% ethanol + 0.01% BHT and preparing the β-carotene standards, measure the absorbance at 450 nm using a spectrophotometer. Use the β-carotene standards to determine the linear range of your machine and dilute samples as needed to be within this range. Use a 1 mL solution of 99% ethanol + 0.01% BHT as a blank.

Friday (Day 12)
Today students will analyze the HPLC results and determine the β-carotene yield of the different strains constructed. With these results, students should then be able to make conclusions about the relative success of each metabolic engineering strategy and come up with hypotheses to explain the results.
1. Measure the cell dry weight as outlined above.
2. Determine the amount of β-carotene produced per gram of cell dry weight.
Expected results Figure 8. shows the results obtained when the authors ran the experiment. While the results for Y.
lipolytica strains matched our expectations with an increase in XdCrtI and Hmg1 activity correlating with higher β-carotene yields, surprisingly the same trend was not observed for S. cerevisiae strains.
While (Verwaal et al., 2007) reported an increase in β-carotene levels upon overexpression of tHMG1, (Ronda et al., 2015) reported a decrease in β-carotene with tHMG1 overexpression similar to the results presented here. Since we only measure β-carotene in this experiment, the observed discrepancies could be due to an increase in the production of phytoene and lycopene which could not be further converted to β-carotene.