Exploiting the Diversity of Saccharomycotina Yeasts To Engineer Biotin-Independent Growth of Saccharomyces cerevisiae

Strains, media, and maintenance.Strains of 35 Saccharomycotina yeasts were tested for growth in the absence of biotin. The S. cerevisiae strains used and constructed in this study belong to the CEN.PK, S288C, Ethanol Red, NCYC 3233/PE-2, and CBS 8066 lineages (Table 2).

Yeast strains were grown on YP medium (10 g liter⁻¹ yeast extract, 20 g liter⁻¹ peptone) or on synthetic medium (SM) containing 3.0 g liter⁻¹ KH₂PO₄, 5.0 g liter⁻¹ (NH₄)₂SO₄, 0.5 g liter⁻¹ MgSO₄, 7·H₂O, 1 ml liter⁻¹ trace element solution (4.5 mg liter⁻¹ ZnSO₄·7H₂O, 0.3 mg liter⁻¹ CoCl₂·6H₂O, 1 mg liter⁻¹ MnCl₂·4H₂O, 0.3 mg liter⁻¹ CuSO₄·5H₂O, 4.5 mg liter⁻¹ CaCl₂·2H₂O, 3 mg liter⁻¹ FeSO₄·7H₂O, 0.4 mg liter⁻¹ NaMoO₄·2H₂O, 1 mg liter⁻¹ H₃BO₃, 0.1 mg liter⁻¹ KI, 15 mg liter⁻¹ EDTA), and 1 ml liter⁻¹ vitamin solution [0.05 g liter⁻¹ d-(+)-biotin, 1.0 g liter⁻¹ d-calcium pantothenate, 1.0 g liter⁻¹ nicotinic acid, 25 g liter⁻¹ myo-inositol, 1.0 g liter⁻¹ thiamine hydrochloride, 1.0 g liter⁻¹ pyridoxol hydrochloride, 0.2 g liter⁻¹ 4-aminobenzoic acid] (56). The pH was adjusted to 6 with 2 M KOH prior to autoclaving at 120°C for 20 min. Vitamin solutions were sterilized by filtration and added to the sterile medium. Concentrated sugar solutions were autoclaved at 110°C for 20 min and added to the sterile medium to give a final concentration of 20 g liter⁻¹ glucose (yeast extract-peptone-dextrose [YPD] and SMG). Biotin-free SM was prepared similarly, but biotin was omitted from the vitamin solution. Similarly, after autoclaving concentrated glucose solution at 110°C for 20 min, glucose was added to biotin-free SM to a final concentration of 20 g liter⁻¹ (biotin-free SMG). Solid medium contained 1.5% Bacto agar and, when indicated, acetamide for SMG acetamide (20 g liter⁻¹ glucose, 1.2 g liter⁻¹ acetamide, 3.0 g liter⁻¹ KH₂PO₄, 6.6 g liter⁻¹ K₂SO₄, 0.5 g liter⁻¹ MgSO₄ 7·H₂O, 1 ml liter⁻¹ trace element solution, and 1 ml liter⁻¹ vitamin solution) and 200 mg liter⁻¹ hygromycin for YPD hygromycin.

E. coli cells (XL1-Blue; Agilent Technologies, Santa Clara, CA) were grown in lysogeny broth (LB) medium (5.0 g liter⁻¹ yeast extract, 10 g liter⁻¹ Bacto tryptone, 5.0 g liter⁻¹ NaCl) supplemented with 25 mg liter⁻¹ chloramphenicol, 100 mg liter⁻¹ ampicillin, or 50 mg liter⁻¹ kanamycin. Solid LB medium contained 2.0% Bacto agar. For maintenance, stock cultures of yeast strains were grown on YPD, if not specified differently, and E. coli cultures on LB medium with an appropriate antibiotic marker until late exponential phase, complemented with sterile glycerol to a final concentration of 30% (vol/vol), and stored at −80°C as 1-ml aliquots until further use.

Shake flask cultivation conditions.Cultivation experiments for the determination of biotin requirements of yeasts were performed as follows. A 1-ml aliquot of a stock culture was inoculated in 100 ml SMG in a 500-ml shake flask and incubated for 20 h at 30°C. A second 100-ml SMG culture was started by inoculating 2 ml of the first shake flask culture. When the second culture reached mid-exponential phase (optical density at 660 nm [OD₆₆₀] of 3 to 5), it was used to inoculate a third culture at an OD₆₆₀ of 0.1 to 0.3. Similarly, a 1-ml aliquot of a thawed stock culture was inoculated in 100 ml biotin-free SMG in a 500-ml shake flask and incubated for 20 h at 30°C. A second 100-ml biotin-free SMG culture was started by inoculating 2 ml of the first shake flask culture. If the second culture reached mid-exponential phase (OD₆₆₀ of 3 to 5), it was used to inoculate a third culture at an OD₆₆₀ of 0.1 to 0.3. Shake flasks were incubated as biological duplicates at 30°C and 200 rpm in an Innova incubator (Brunswick Scientific, Edison, NJ). Strains IMX585 and CEN.PK113-7D, which consistently failed to grow on biotin-free SMG in the third culture, were used as a negative control in all growth experiments.

Growth of cultures was monitored by OD₆₆₀ of an appropriate dilution of the third shake flask culture, measured with a Jenway 7200 Spectrophotometer (Cole-Palmer, Stone, United Kingdom). Specific growth rates were calculated from a minimum of six data points during exponential growth, covering 3 to 4 doublings of the OD₆₆₀. Growth rate was calculated using the equation X = X₀e^μt, in which μ indicates the exponential growth rate. All aerobic shake flask experiments were carried out in an Innova shaker incubator (New Brunswick Scientific, Edison, NJ) set at 30°C and 200 rpm.

For growth profiling under anaerobic conditions, first and second cultures were grown in 40 ml SMG or biotin-free SMG in a 50-ml shake flask, as described previously, under aerobic conditions. Two hundred-microliter samples of mid-exponentially growing cells from the second culture were transferred to an anaerobic workstation (Bactron; Sheldon Manufacturing, Cornelius, OR) at 30°C and used to inoculate the third culture. Oxygen entry through the airlock of the anaerobic workstation was minimized with the use of a regenerated Pd catalyst for H₂-dependent oxygen removal that was introduced into the chamber. Concentrated solutions of Tween 80 and ergosterol were added to the medium, aiming for final concentrations of 420 mg liter⁻¹ and 10 mg liter⁻¹, respectively. To deplete all of the nutrients from the aerobic growth phase, a fourth culture was inoculated from exponentially growing cells. The OD₆₆₀ of the fourth culture was measured with an Ultrospec 10 cell density meter (Biochrom; Harvard Bioscience, Cambridge, United Kingdom), which was placed inside the anaerobic workstation. All anaerobic experiments were carried out in biological duplicates as described before (57).

In the case of cocultivations, the strains were inoculated in 100 ml SMG and biotin-free SMG in a 500-ml shake flask by addition of a 1-ml aliquot of a stock culture and incubated for 20 h at 30°C. A second 100-ml SMG culture was started by inoculating 2 ml of the first shake flask culture. When the second culture reached mid-exponential phase (OD₆₆₀ of 3 to 5), it was used to inoculate a third culture at an OD₆₆₀ of 0.1 to 0.3 and an approximate 60:40 mix of the two strains from the same medium (biotin supplemented or biotin free). After reaching late exponential phase, cocultures were transferred into fresh medium to an OD₆₆₀ of 0.2. The fifth and sixth consecutive cultures were started similarly by transferring stationary-phase cultures from the previous batch. During the third, fourth, fifth, and sixth cultures, samples were taken to analyze the OD₆₆₀ and population distribution by flow cytometry.

Molecular biology techniques.DNA fragments were amplified by PCR amplification with Phusion Hot Start II high-fidelity polymerase (Thermo Fisher Scientific, Landsmeer, Netherlands) and desalted or PAGE-purified oligonucleotide primers (Sigma-Aldrich, St. Louis, MO) (Table 3) according to the manufacturers’ instructions. For diagnostic PCR analysis of constructed plasmids and strains, DreamTaq polymerase (Thermo Fisher Scientific) was used according to the manufacturer’s recommendations. PCR products were separated by electrophoresis on 1% (wt/vol) agarose gels in Tris-acetate-EDTA (TAE) buffer (Thermo Fisher Scientific) with SERVA DNA stain G (1:100,000) (Serva, Heidelberg, Germany) and, if required, purified with a Zymoclean gel DNA recovery kit (Zymo Research, Irvine, CA) or GenElute PCR clean-up kit (Sigma-Aldrich). Yeast strains of the CEN.PK lineage, S288C, and NCYC 3233/PE-2 were transformed by the lithium acetate (LiAc) method. Yeast strains Ethanol Red and CBS 8066 were transformed using electroporation, as previously described (58), with a 2-mm cuvette (Bio-Rad, Hercules, CA) using a Gene Pulser XCell electroporation system (Bio-Rad). Yeast genomic DNA was isolated using the YeaStar genomic DNA kit (Zymo Research) or using the SDS-LiAc protocol (59). E. coli cells were chemically transformed and plated on selective LB agar medium. Plasmids from selected clones were isolated from E. coli with a Sigma GenElute plasmid kit (Sigma-Aldrich).

Plasmid construction.(i) Construction of part plasmids using yeast toolkit principles. Coding sequences of putative ScBIO1 orthologs from Y. lipolytica W29, P. kudriavzevii CBS 5147, W. ciferrii CBS 111, C. fabianii CBS 5481, L. kluyveri CBS 3082, and T. delbrueckii CBS 813 were obtained by PCR with primer combinations 14925/14926, 14892/14893, 15104/15105, 13963/13964, 13291/13039, 13290/13038, and 12991/12992 using genomic DNA of the respective yeast as a template. In the case of S. cerevisiae CEN.PK113-7D BIO1, the plasmid pUDE450 (Table 4) (15) was isolated from E. coli cultures and used as a template for PCR with primer pair 11614/11615. The ScBIO1 terminator (ScBIO1t) was similarly PCR amplified using primer pair 11618/11619 and plasmid pUDE450 as the template. The DNA fragments containing BIO1 coding sequences from Y. lipolytica, T. delbrueckii, L. kluyveri, and CEN.PK113-7D, as well as ScBIO1t, were in vitro assembled with entry vector pUD565 using BsmBI-T4 ligase-directed Golden Gate cloning (60), resulting in yeast toolkit type 3 plasmids pGGkp243, pGGkp169, pGGkp178, and pGGkp080, respectively, and a yeast toolkit type 4 plasmid with ScBIO1t pGGkp078 (Table 4). To remove a BsaI restriction recognition site in the coding sequence of LkBIO1, the open reading frame was PCR amplified from L. kluyveri CBS 3082 genomic DNA using two primer pairs, 13291/13039 and 13290/13038, with overhangs allowing for BsmBI-T4 ligase-directed Golden Gate cloning of the two DNA fragments into entry vector pUD565 (61), leaving behind a sequence without a BsaI site and resulting in the yeast toolkit type 3 plasmid pGGkp178. After in vitro assembly, plasmids were transformed into E. coli and plated on LB chloramphenicol for selection. The yeast toolkit type plasmids pGGkp080, pGGkp169 pGGkp078, and pGGkp178 were confirmed by diagnostic PCR with primer pair 12616/4892, 12616/13287, 12616/10235, and 12616/13290, respectively. Yeast toolkit type plasmid pGGkp243 was confirmed by restriction analysis with restriction enzymes PvuII and DraI. The promoter ScPYK1p was synthesized by GeneArt (Thermo Fisher Scientific) and cloned as plasmid pGGkp117. The yeast toolkit type plasmids were stored in transformed E. coli cultures.

(ii) Construction of transcriptional modules. The control S. cerevisiae CEN.PK113-7D BIO1 transcriptional module was constructed by Golden Gate cloning combining DNA fragments with compatible overhangs from plasmids pGGkd017, pGGkp117, pGGkp080, and pGGkp078, yielding plasmid pUDE718. The entry plasmid pGGkd017 was also constructed by Golden Gate cloning combining DNA fragments with compatible overhangs from pYTK002, pYTK047, pYTK072, pYTK074, pYTK082, and pYTK083 (60). The T. delbrueckii BIO1 transcriptional module was constructed by Golden Gate cloning, combining DNA fragments with compatible overhangs from plasmids pGGkd015, pGGkp117, pGGkp169, and pGGkp078, yielding plasmid pUD788. The entry plasmid pGGkd015 was obtained by Golden Gate cloning with yeast toolkit type plasmids pYTK002, pYTK047, pYTK067, and pYTK095. The L. kluyveri BIO1 transcriptional module was constructed by Golden Gate cloning, combining DNA fragments with compatible overhangs from plasmids pGGkd015, pGGkp117, pGGkp178, and pGGkp078, yielding plasmid pUD789. The Y. lipolytica BIO1 transcriptional module was constructed by Golden Gate cloning, combining DNA fragment with compatible overhangs from plasmids pGGkd015, pGGkp117, pGGkp243, and pGGkp078, yielding plasmid pUD989. The transcriptional modules of P. kudriavzevii, W. ciferrii, and C. fabianii BIO1 genes were constructed by Gibson assembly (62) (New England Biolabs, Ipswich, MA) using pUDE718 as the plasmid backbone, which includes the ScPYK1p and ScBIO1t sequences. The BIO1 genes from P. kudriavzevii CBS 5147, W. ciferrii CBS 111, and C. fabianii CBS 5481 were amplified by primer pairs 14892/14893, 15104/15105, and 13963/13964, respectively, using genomic DNA of the respective yeast as a template. DNA fragments were assembled with linearized pUDE718 backbone using primer pair 7428/14891, yielding plasmids pUD988, pUD990, and pUD790, respectively. The assembly mixes were transformed into E. coli and plated on LB-ampicillin for selection. The transcriptional module plasmids were confirmed by diagnostic PCR with the forward primer 10320 and the following gene-specific reverse primers: 13287 for TdBIO1, 13293 for LkBIO1, 14928 for YlBIO1, 4892 for ScBIO1, 14909 for PkBIO1, 14907 for WcBIO1, and 14162 for CfBIO1. The transcriptional module for the expression of fluorophore mRuby2 was constructed by Golden Gate cloning, combining DNA fragments with compatible overhangs from plasmids pGGKd005, pYTK011, pYTK046, and pYTK054. The entry plasmid pGGkd005 was constructed by Golden Gate cloning, combining DNA fragments with compatible overhangs from pYTK002, pYTK047, pYTK067, pYTK079, pYTK083, and pYTK081. The assembly mixes were transformed into E. coli and plated on LB ampicillin for selection. The mRuby2 transcriptional module plasmid was confirmed by diagnostic PCR with primer pair 10320/10325 and stored as pUDE480 in transformed E. coli cultures.

(iii) Spycas9-expressing plasmids. The gRNA_ScSGA1-expressing plasmid pUDR244 was constructed by in vitro Gibson assembly. The linearized pROS11 plasmid, obtained by PCR with 6005/6006, was assembled together with a PCR-amplified fragment using primer 14139 and pROS11 as a template as previously described (26). The plasmid DNA was isolated from E. coli and the correct assembly of plasmid pUDR244 confirmed by diagnostic PCR with primers 3841/14167/5941. Similarly, plasmid pUDR376 was assembled with linearized pROS11 and the PCR-amplified DNA fragment but using primer 10866. Assembly of pUDP145 was performed in vitro by BsaI-T4 DNA ligase-directed Golden Gate cloning with the gRNA entry plasmid pUDP002 (63) and a de novo-synthesized DNA fragment (GeneArt, Thermo Fisher Scientific) encoding a gRNA targeting ScSGA1.

Strain construction.The BIO1 transcriptional modules were PCR amplified by using primer pair 12086/12108, adding specific sequences for homologous recombination into the SGA1 locus in S. cerevisiae, directed by CRISPR/Cas9 (26). The transcriptional module was amplified from plasmid pUD788 for TdBIO1, from plasmid pUD789 for LkBIO1, from plasmid pUD989 for YlBIO1, from plasmid pUDE718 for ScBIO1, from plasmid pUD988 for PkBIO1, from plasmid pUD990 for WcBIO1, and from plasmid pUD790 for CfBIO1. Targeting at the ScSGA1 locus in IMX585 was directed by CRISPR/Cas9 and a target-specific gRNA-expressing plasmid (26). Each transcriptional module was cotransformed with plasmid pUDR119 expressing the gRNA to target Cas9 to ScSGA1 in strain IMX585 using the LiAc transformation protocol. The transformed cells were plated on selective SMG acetamide and incubated for 3 days at 30°C. Genomic DNA of transformants was isolated using the SDS-LiAc protocol (59). The desired genotype was confirmed using primer pair 11898/11899 and a gene-specific primer pair with 11898 as the forward primer and the following reverse primers: 13287 for TdBIO1, 13293 for LkBIO1, 14928 for YlBIO1, 4892 for ScBIO1, 14909 for PkBIO1, 14907 for WcBIO1, and 14162 for CfBIO1. The correct clone was inoculated in 20 ml nonselective YPD for plasmid removal and incubated for 24 h at 30°C. The cells were plated on YPD agar to obtain single-colony isolates. One isolate was restreaked on both SMG acetamide and YPD. When no growth was observed on SMG acetamide, the respective clone was once again confirmed by PCR with gene-specific primers. Furthermore, the genetic modification at the ScSGA1 locus was verified by Sanger sequencing (BaseClear, Leiden, Netherlands), using primers 11898/11899 to PCR amplify the modified locus and further using primers 11898, 11915, and 10235 for sequencing. The strain with the transcriptional module coding for TdBIO1 was stocked as IMX1857, LkBIO1 as IMX1858, YlBIO1 as IMX1862, ScBIO1 as IMX1511, PkBIO1 as IMX1861, WcBIO1 as IMX1863, and CfBIO1 as IMX1859 in SMG.

The deletion of the native ScBIO1 locus in strains IMX1859 and IMX585 was directed by CRISPR/Cas9 using plasmid pUDR244 (Table 4), which was cotransformed with annealed repair oligonucleotides 12223/12224 in strains IMX1859 and IMX585 using the LiAc transformation protocol with SMG acetamide as the selection marker, and the deletion was confirmed using primer pair 7469/10873.

To achieve the integration of the CfBIO1 module at the SGA1 locus of CBS 8066, NCYC 3233/PE-2, and Ethanol Red, the plasmid pUDP145 was cotransformed with a PCR-amplified DNA fragment using primer pair 12086/12108 and pUD790 as the template. In contrast to CBS 8066, NCYC 3233/PE-2, and Ethanol Red, the S. cerevisiae strain S288C is missing both the ScBIO1 and ScBIO6 genes (27); therefore, the CfBIO1 transcriptional module was amplified using primer pair 12086/14663 from pUD790, and an additional transcriptional module harboring ScBIO6 was amplified using primer pair 14661/14662 and plasmid pUDE448 as the template. These two DNA fragments harbored homologous flanks, allowing for in vivo assembly into the ScSGA1 locus after cotransformation with pUDP145. Transformants selected on YPD hygromycin were tested for the desired genotype using primer pair 11898/11899 and using a CfBIO1-specific PCR with primer pair 11898/14162. In the case of transformation into strain S288C, an additional diagnostic PCR with primer pair 8737/11899 was conducted. After counterselection, the strain with the CfBIO1 and ScBIO6 transcriptional module in S288C was stored as IMX2103. CBS 8066 expressing CfBIO1 was stored as IMX2104, Ethanol Red expressing CfBIO1 as IMX2101, and NCYC 3233/PE-2 expressing CfBIO1 as IMX2090 in SMG.

The Venus fluorophore transcriptional module was PCR amplified from pUDC193 by using primer pair 16792/16793. The mRuby2 fluorophore transcriptional module was PCR amplified from plasmid pUDE480 with primer pair 13596/13597. These two linear DNA fragments contained homologous flanks to the intergenic region X-2 (64) to enable integration at this site when cotransformed with pUDR376 into a Cas9-expressing strain background. The Venus fluorophore gene was integrated into IMX585 and IMK827 and the mRuby2 fluorophore into IMX1860. Genomic DNA of transformants selected on SMG acetamide was isolated using the SDS-LiAc protocol. The desired genotype was confirmed by PCR using primer pair 13662/11037 and a gene-specific primer pair with 13662 as forward primer and 5328 as reverse primer to confirm mRuby2 expression cassette integration. The primer pair 13662/11945 was used to confirm the integration of the Venus expression cassette at the X-2 intergenic site. IMK827 with Venus fluorescence was stored as IMX2240, IMX585 with Venus fluorescence was stored as IMX2212, and IMX1860 with mRuby2 fluorescence was stored as IMX2238 in SMG.

Flow-cytometric analysis.Samples from aerobic 100-ml cultures in 500-ml shake flasks were vortexed thoroughly to disrupt cell aggregates and used for flow cytometry on a BD FACSAria II SORP cell sorter (BD Biosciences, Franklin Lakes, NJ) equipped with 355-, 445-, 488-, 561-, and 640-nm lasers and a 70-μm nozzle and operated with filtered FACSFlow (BD Biosciences). Cytometer performance was evaluated prior to each experiment by running a CST cycle with CS&T beads (BD Biosciences). The fluorophore mRuby2 was excited by the 561-nm laser, and emission was detected through a 582-nm bandpass filter with a bandwidth of 15 nm. The fluorophore Venus was excited by the 488-nm laser, and emission was detected through a 545-nm bandpass filter with a bandwidth of 30 nm. For each sample, 10,000 events were analyzed, and the same gating strategy was applied to all samples from the same culture. The reference sample for no fluorescent cells was a mid-exponentially growing culture of IMX585 on SMG. The Venus and mRuby2 fluorescence reference was obtained from mid-exponential aerobic cultures on SMG of IMX2240 or IMX2212 and IMX2238, respectively. Cells without fluorescence and doublets or with Venus and mRuby2 fluorescence were selected in a Venus/mRuby2 plot.

Sequence analysis and structural modeling.Genomic DNA of Y. lipolytica W29 was sequenced in-house on a MiSeq sequencer (Illumina, San Diego, CA) to obtain a 300-cycle paired-end library with a fragment length of 550 bp using a PCR-free library preparation, yielding 4.28 million reads with a total sequence of 1.27 gigabases. De novo assembly was performed using SPAdes (version 3.9.0), producing a 20.48-megabase genome comprising 409 contigs and an N₅₀ of 181.71 kb in 36 contigs.

tBLASTn (https://blast.ncbi.nlm.nih.gov) was used for the identification of BIO1 orthologs. The amino acid sequence encoded by ScBIO1 (14, 65) was queried against the translation of whole-genome shotgun (wgs) or nucleotide collection data of single yeast species. In a reciprocal analysis, the yeast-specific best hits with a minimum coverage of 80% were aligned using tBLASTn against the S. cerevisiae CEN.PK113-7D nucleotide sequence to verify the similarity to the ScBIO1 sequence. The identified putative BIO1 amino acid sequences (Fig. 3) were pairwise aligned using ClustalΩ (scoring matrix BLOSUM62) (25) to determine amino acid sequence similarities. The BIO1 structural model was generated by homology modeling using Phyre2 (66).

Data availability.The sequencing data and assembly of the Yarrowia lipolytica strain W29 were deposited at NCBI (https://www.ncbi.nlm.nih.gov/) under BioProject accession number PRJNA601425.