NTH2 1271_1272delTA Gene Disruption Results in Salt Tolerance in Saccharomyces cerevisiae

: Trehalose is a common energy reservoir, and its accumulation results in osmotic protection. This sugar can accumulate through its synthesis or slow degradation of the reservoir by trehalase enzymes. Saccharomyces cerevisiae contains two neutral trehalases, NTH1 and NTH2 , responsible for 75% and 25% of the enzymatic metabolism. We were interested in the loss-of-function of both enzymes with CRISPR/Cas9. The later NTH2 was of great importance since it is responsible for minor metabolic degradation of this sugar. It was believed that losing its functionality results in limited osmotic protection. We constructed an osmotolerant superior yeast capable of growing in 0.85 M NaCl after independent nth2 1271_1272delTA mutation by CRISPR/Cas9 technology, compared with nth1 893_894insT and wild type. We suggest that this yeast model could give clues to breeding commercial yeast resulting in non-GMO salinity-tolerant strains.


Introduction
Trehalose (α-D-glucopyranosyl-α-D-glucopyranoside) is a stress-tolerance-related sugar [1]. This nonreductive disaccharide, composed of two molecules of D-glucose, is also common in the metabolism of bacteria, fungi, plants, and some invertebrate animals [2][3][4]. Trehalose protects against structural disorders during water removal. It has been suggested that the hydroxyl groups of trehalose form hydrogen bonds with polar lipids and proteins instead of water, protecting these hydroxyl groups from structural disorder during water removal [5]. Trehalose can also work indirectly by stabilizing heat shock proteins during the refolding of damaged proteins [6]. Trehalose acts as a cryoprotectant to avoid aggregation by stabilizing proteins in a chaperone-like manner [5,[7][8][9][10].
The intracellular accumulation of this disaccharide is due to the trehalose synthesis complex: TPS1, TPS2, TPS3, and TSL1 [11]. However, the NTH1, NTH2, and ATH1 genes are responsible for degrading trehalose and are also positively regulated under stress conditions, creating a metabolic cycle of carbohydrate regulation whose understanding is still controversial [12,13]. A high intracellular concentration of this molecule results in protecting the structures of biomolecules and the blockage of proteases while allowing R64-1-1.80) was designed using the CRISPRdirect Platform [28]. Recognition of the single guide insertion for the NTH2 gene (5 -TGCTATTAAAGAATATAAAG [AGG]-3 ) and NTH1 (5 -GGTTACCCTTATGCTGTTCC [TGG]-3 ) were cloned into bRA89 at Genscript, just below the RNA scaffold section, using BplI restriction sites. Sanger sequencing confirmed the insertion of the single guide section next to the 5 end of the RNA scaffold. More details are provided in the Supplementary Materials ( Figures S1 and S2).

S. cerevisiae Competent Cell Preparation and Transformation
Competent cells were prepared and transformed using the lithium acetate (LiAc) method with the following modifications [29]. An aliquot of approximately 1 × 10 7 cells/mL (16 h culture) was inoculated in 50 mL of YPD medium and incubated until reaching an OD (600 nm) of 0.5. Then, the biomass was harvested by centrifugation (4000 rpm) for 10 min at room temperature, washed twice in 10 mL of sterile distilled water, and resuspended in 1.5 mL of sterile lithium acetate buffer (1 volume of TE 10X buffer, pH 7.5; 1 volume of 1 M 10X LiAc, pH 7.5 and 8 volumes of distilled water). Freshly prepared competent yeast cells (200 µL) were mixed with 1µg of the final plasmid and combined with 200 µg of denatured Salmon Sperm DNA (Thermo Fisher, Carlsbad, CA, USA) and 1 mL of fresh PEG buffer (8 volumes of PEG 3350 50% and 1 volume of TE 10X buffer, pH 7.5; 1 volume of 1 M LiAc 10X, pH 7.5). The solution was incubated at 150 rpm for 30 min at 25 • C and then heated to 42 • C for exactly 15 min. Subsequently, the solution was resuspended in 200 µL of YPAD medium and incubated for another 45 min at 25 • C. Finally, the cells were cultured on YPD agar plates supplemented with hygromycin B Phytotechnology Laboratories ® (Shawnee Mission, KS, USA) at 500 µg mL −1 for 72 h at 28 • C in an incubator.

Selection and Sequencing Confirmation of Mutants
Thirty (30) randomly selected strains were grown on YPD overnight media. Subsequently, the manufacturer's instructions obtained genomic DNA using the ReliaPrepTM gDNA Tissue Miniprep System (Promega, Madison, WI, USA). Specific primers flanking the target sgRNA site of the NTH2 gene were designed and named nth2f : 5 -GCAAGAGG TATGGTGGAGCA-3 and nth2r: 5 -TTCAGCTAGCTCCTCCCAGT-3 (Tm 55 • C; 539 bp); while NTH1 primers flanking the target sgRNA site were also designed and named nth1f : 5 -ACCCCCGGTTTACTAGCATTG-3 and nth1r: 5 -TAAGGTAACGCCGTGTTTCGA-3 (Tm 55 • C; 528 bp). Sanger sequencing of PCR products was performed in Macrogen at Rockville, MD, USA, to confirm the mutation and absence of off-targets. We selected two isolates with the same mutation on the NTH2-disrupted gene and intact NTH1 gene, named nth2 1271_1272delTA. Similarly, we selected two isolates with the NTH1-disrupted gene and intact NTH2 gene, named nth1 893_894insT.

nth1 893_894insT and nth2 1271_1272delTA Strain Phenotypes
The methylene blue staining technique allowed examination of the viability of the wild-type and mutated yeast strains in 0.85 M NaCl versus the control (0 M NaCl) on Potato Dextrose (PD) (20% Potato, 2% dextrose) for 48 h at 28 • C. Growth curves of the samples were generated with 0 and 0.85 M NaCl in the YPD medium. The test was carried out with a SPECTROstarNANO plate reader from BMG LABTECH at 200 rpm and 30 • C for 24 h, with four repetitions per variant. The growth curve comparisons resulted from T-test analysis with and without saline stress conditions of the wild-type and mutated strains. The comparison also consisted of seriated dilutions on agar PD plate-based comparison after two weeks of growth under stress (0.85 M and 1.2 M NaCl). Trehalose content was extracted following the protocol described by Divate et al. [22] with the following variations. The cells were grown after 0.85 M NaCl versus the control (0 M NaCl) on PD for 48 h at 28 • C, collected by centrifugation at 7000× g for 5 min and dried at 100 • C for 12 h. A pellet of 50 mg was mixed with 1 mL ethanol (99.5%) and incubated in a boiling water bath for 1 h. HPLC-RID determined the content, P-SA-MQ-006 provided by CITA-UCR, as follows: The ethanolic extract was centrifuged and filtered through a 0.20 µm regenerated cellulose micropore (17761-Q, Minisart-RC15 ® , Sartorius AG, Göttingen, Germany), the filtrate was collected in a 2 mL vial for HPLC. AC Chromatographic separation was performed using Agilent Technologies 1260 Infinity liquid chromatograph equipped with Suplecogel 8Ca high resolution column (300 mm × 7.8 mm, 8 µm, PN 59247-U), quaternary pump (G1311B), column compartment (G1316A), automatic liquid sampling module (ALS, G7129A) and refractive index detector (G1362A) (Agilent Technologies, Santa Clara, CA, USA). The mobile phase consisted of ultrapure water (type I, 0.055 µS cm −1 at 25 • C, 5 µg L −1 TOC) obtained using A10 Milli-Q Advantage and Elix 35 purification system (Merck KgaA, Darmstadt, Germany). Solvent flow, column compartment temperature, detector cell temperature, and injection volume were constant during the elution at 0.40 mL min −1 , 80 • C, 40 • C, and 10 µL, respectively. The area under the curve (AUC) of the trehalose signal in the samples was interpolated in the calibration curve of the certified reference standard (Sigma-Aldrich, PHR1344-500 mg), in a concentration range of 0.025 to 0.25 g/100mL.

Scanning Electron Microscopy (SEM)
Wild-type and mutated yeast grown in 0.00 and 0.85 M NaCl and fixed with Karnovsky fixative (2.5% w/v of glutaraldehyde, 0.1 M of paraformaldehyde in phosphate buffer pH 7.4, during 48 h at 4 • C) were mounted on carbon tape and sputtered with gold using a Denton Vacuum Desk V sputter system at 20 mA for 300 s. Images were taken using a JSM-6390LV (JEOL, Tokyo, Japan) scanning electron microscope with an accelerating voltage of 15 kV under high vacuum. Scanning electron microscopy images at different resolutions of the wild-type and NTH2-mutated yeasts and subsequent cell area calculations were analyzed with ImageJ version 1.52p.

Transmission Electron Microscopy (TEM)
Wild-type and mutated yeast grown with and without NaCl were treated with Karnovsky fixative for six days. The yeast was then centrifuged at 3000 rpm for 5 min and rinsed with 0.1 M phosphate buffer solution (pH 7.2) for three 15 min wash/centrifugation cycles. The yeast samples were then stirred for one hour in 2% w/v OsO 4 in 0.1 M phosphate buffer solution, followed by three wash cycles as per the previous step but with type 3 water instead of buffer. The yeasts were finally isolated via centrifugation at 3000 rpm for 10 min and subsequently embedded in a solution of agar/agarose 4% w/v in a hot water bath (50 • C). After cooling to 37 • C, one drop of the solution was added to an Eppendorf tube containing 100 µL of the yeast sediment and mixed thoroughly. After cooling, the solids were removed and cut into 3 mm 3 segments. Dehydration was performed by rinsing in an ascending gradient of acetone (30% to 100% v/v) followed by infiltration with Spurr resin: acetone 50:50 overnight and three successive infiltrations with pure Spurr resin for two hours each. The resulting solids were transferred to BEEM embedding capsules. Polymerization was achieved in an oven at 70 • C for 24 h. Ultrafine segments (approximately 70 nm) were cut with a Leica EM UC7 ultramicrotome with a diamond knife and supported on Cu TEM grids (200 mesh). The sections were stained with 4% m/v uranyl acetate in 50% v/v ethanol for 15 min and washed five times in DI water, followed by soaking for 10 min in Reynold's stain and rinsing five times in water. Once dry, the grids were mounted in a JEOL JEM 2011 TEM and observed at 120 kV at magnifications of 3000-20,000×.

Confirmation of NTH1-NTH2 Gene-Mutated Cells
NTH1 and NTH2 disruption were confirmed in the randomly selected strains. We selected two isolates of NTH2 mutant strains named nth2 1271_1272delTA containing the same mutation, a deletion of two bases, TA, located at positions 1271-1272, three bases downstream from the PAM site expected. A PCR of the target sequence ( Figure 1A) and Sanger sequencing confirmed that the deletion was present in only the strains treated with the CRISPR/Cas9 technique directed through the single guide RNA complementary to section 1258 to 1276 of the NTH2 compared with the chromosome 2 sequence. Sequencing also confirmed that the open reading frame disruption results from forming the TAA triplet in the respective open reading frame, creating a stop codon at amino acid 424 ( Figure 1B). Additionally, the integrity of the NTH1 gene was verified in the NTH2-mutated strain by PCR and sequencing of the specific mutation zone. The S. cerevisiae NTH1 gene was not altered due to CRISPR/Cas9 genome editing in the strain nth2 1271_1272delTA.

Confirmation of NTH1-NTH2 Gene-Mutated Cells
NTH1 and NTH2 disruption were confirmed in the randomly selected strains. We selected two isolates of NTH2 mutant strains named nth2 1271_1272delTA containing the same mutation, a deletion of two bases, TA, located at positions 1271-1272, three bases downstream from the PAM site expected. A PCR of the target sequence ( Figure 1A) and Sanger sequencing confirmed that the deletion was present in only the strains treated with the CRISPR/Cas9 technique directed through the single guide RNA complementary to section 1258 to 1276 of the NTH2 compared with the chromosome 2 sequence. Sequencing also confirmed that the open reading frame disruption results from forming the TAA triplet in the respective open reading frame, creating a stop codon at amino acid 424 ( Figure  1B). Additionally, the integrity of the NTH1 gene was verified in the NTH2-mutated strain by PCR and sequencing of the specific mutation zone. The S. cerevisiae NTH1 gene was not altered due to CRISPR/Cas9 genome editing in the strain nth2 1271_1272delTA.  Mutation on the NTH1 and NTH2 genes showed disruption and absence of the active sites for a putative first ORF protein translated, specifically amino acid residues D478, E674 for NTH1, and D507 and E703 for NTH2. The nth2 1271_1272delTA may produce a small truncated 423 amino acid protein instead of the complete enzyme, lacking biochemical function. Similarly, the insertion in NTH1 results in a knockout of three triplets downstream of TAG, resulting in a truncated 301 AA with no disaccharide binding site. The truncated sequence of mutants nth1 893_894insT and nth2 1271_1272delTA may still contain the Ca 2+ binding domain if the three-dimensional conformation is unchanged.
When comparing the resulting putative Open Reading Frames, we noted that phosphorylation activation of the enzymes remains in the amino terminal that corresponds to the first disrupted ORF of both nth1 893_894insT and nth2 1271_1272delTA. A putative, predicted second ORF corresponding to the carboxyterminal end is different for each mutant. In the case of nth1 893_894insT, the second ORF may have all the binding sites (R302, N346, E424, R473, G476), active sites (D507, E703), and substrate binding sites (338-339WD, 384-386RSQ) in the second ORF. In the case of nth2 1271_1272delTA, the sites may be spliced by having one binding site (G505) and the active sites in the second ORF (D478, E674), but most binding sites (R331, N375, R384) and all substrate binding sites (309-310WD, 355-357RSQ) in the first ORF (Table 1). Mutation on the NTH1 and NTH2 genes showed disruption and absence of the active sites for a putative first ORF protein translated, specifically amino acid residues D478, E674 for NTH1, and D507 and E703 for NTH2. The nth2 1271_1272delTA may produce a small truncated 423 amino acid protein instead of the complete enzyme, lacking biochemical function. Similarly, the insertion in NTH1 results in a knockout of three triplets downstream of TAG, resulting in a truncated 301 AA with no disaccharide binding site. The truncated sequence of mutants nth1 893_894insT and nth2 1271_1272delTA may still contain the Ca2+ binding domain if the three-dimensional conformation is unchanged.
When comparing the resulting putative Open Reading Frames, we noted that phosphorylation activation of the enzymes remains in the amino terminal that corresponds to the first disrupted ORF of both nth1 893_894insT and nth2 1271_1272delTA. A putative, predicted second ORF corresponding to the carboxyterminal end is different for each mutant. In the case of nth1 893_894insT, the second ORF may have all the binding sites (R302, N346, E424, R473, G476), active sites (D507, E703), and substrate binding sites (338-339WD, 384-386RSQ) in the second ORF. In the case of nth2 1271_1272delTA, the sites may be spliced by having one binding site (G505) and the active sites in the second ORF (D478, E674), but most binding sites (R331, N375, R384) and all substrate binding sites (309-310WD, 355-357RSQ) in the first ORF (Table 1).

Behavior of the nth2 1271_1272delTA Strain under Salinity Stress
The nth2 1271_1272delTA strain has increased tolerance and can survive in high concentrations of NaCl (0.85 M NaCl). We noted that the nth2 1271_1272delTA strains were slightly smaller than the control under the light microscopy, although we were not able to detect any statistical difference. We validated that the cells remained the same and had no statistical differences from the control under high osmolarity conditions and also with the scanning electron microscopy. Yeast dimensions were determined from the scanning electron microscopy images ( Figure 3A). The sizes of the cells remained statistically and phenotypically identical to the wild-type strain and were not collapsed, although we noted that they were slightly smaller ( Figure 3B).
The nth2 1271_1272delTA strain cells were also not different under transmission electron microscopy analysis (data not shown). Organelles and structures such as vacuoles, nucleus, mitochondrion, cell membrane, and cell wall had no differences compared to the wild-type CEN.PK2-1C strain. We noted no organelle disruption nor structural changes in the tolerant strain (for more details check the Supplementary Materials).
The nth2 1271_1272delTA strain was viable in a maximum concentration of 0.85 M NaCl and gave a standard growth curve and an average growth rate of 0.2327 ± 0.0057 h −1 (Figure 4). When statistically analyzing the specific growth rates (p < 0.05), the nth2 1271_1272delTA strain under stress conditions had a growth rate of 0.2179 ± 0.0061 h −1 and behaved in the same way as the wild-type strain under normal conditions 0.2255 ± 0.0037 h −1 . The behavior of the wild-type strain in NaCl solution presented a significant decrease of 0.1580 ± 0.0009 h −1 .
We expected a slight salt tolerance because of the neglected reported activity of NTH2. The latter was reasonable because the strain still had functional NTH1 enzymes that metabolize trehalose. However, the data obtained showed that the nth2 1271_1272delTA strains were superiorly tolerant. The nth2 1271_1272delTA strains had an average growth curve in a 0.85 M NaCl liquid medium, with two isolates having the same mutation and behavior. Instead, the wild type had a slower growth curve (Figure 4).
We mutated the homolog gene NTH1 resulting in nth1 893_894insT strains to compare it with nth2 1271_1272delTA. The nth1 893_894insT strains showed no tolerance to 0.85 and 1.2 M NaCl when grown on agar plate-based comparison after two weeks ( Figure 5). Growth on 0.85 M and 1.2 M was detected after three days for nth2 1271_1272delTA, mutants but it took a week for the control and nth1 893_894insT mutants. We could not detect nth1 893_894insT salt tolerance when growing the mutant in 0.85 M NaCl liquid media (data not presented).   We noted no difference in the intracellular content of trehalose in nth2 1271_1272delTA in comparison with the control in the stationary phase after 48h of growth with or without stress. However, in nth1 893_894insT, the trehalose content was low with or without NaCl. (Table 2).

Discussion
The salinity tolerance of the S. cerevisiae strain CENPK2 increased when the NTH2 gene was disrupted using CRISPR/Cas9-mediated genome editing compared with wild type and NTH1 disruption. The nth2 1271_1272delTA yeast strains grew in 0.85 M NaCl with no detectable changes in behavior other than stress tolerance. Although deletion of NTH1 and NTH1-NTH2 together was known to result in stress tolerance, to our knowledge, this is the first report of using the CRISPR/Cas9 technique to disrupt NTH2 alone that results in remarkable stress tolerance, as confirmed by an automated measurement system. The result is similar to a predictive model suggesting such tolerance for NTH2 kanMX4 deletion and a neglectable tolerance for NTH1-disrupted strains [30].
The use of CRISPR/Cas9 resulted in the expected specific mutations of the S. cerevisiae of both NTH2 and NTH1 independent genes, three bases downstream of the PAM section (NGG) with the S. pyogenes Cas9 enzyme [31,32]. In the case of NTH2, the double-strand break resulted in the deletion of two nucleotides after nonhomologous end joining (NHEJ), introducing the "TAA" stop codon in the respective open reading frame (Figure 1). The nth2 1271_1272delTA strain had a deletion of two nucleotides, TA, located at positions 1271-1272, which resulted in a stop codon Y424X mutation of the NTH2 gene; notably, no changes in the NTH1 gene were observed. The anticipated tridimensional structure of the putative 423 amino acid Open Reading Frame (ORF) of nth2 1271_1272delTA indicated that this modified enzyme should be inactive due to the absence of the active residues ASP507 and GLU703. Similarly, the NTH1 mutant, nth1 893_894insT contained an insertion of one T base three bases downstream of the PAM site as expected and resulted in an early stop codon three triplets downstream.
In this study, nth2 1271_1272delTA strains remained viable and had better tolerance to 0.85 M NaCl than the wild-type and nth1 893_894insT strains. The behavior and size of S. cerevisiae nth2 1271_1272delTA did not change compared to the wild type and had normal variability depending on the generations and growth stage. Yeast tends to shrinkage and collapse in NaCl osmotic stress without plasmolysis, but its primary difference is in mitochondrial fragmentation [33]. We validate that nth2 1271_1272delTA did not collapse under stress and was identical to the control with no stress and had no fragmentation of its organelles using transmission electron microscopy.
The nth2 1271_1272delTA strain showed exponential growth under NaCl stress conditions (0.2179 ± 0.0061 h −1 ) very similar to the wild-type strain without the presence of the osmotic agent (0.2255 ± 0.0037 h −1 ). The results indicate that this mutation provides the yeast with a greater tolerance to saline conditions without significantly affecting its specific growth rate than the wild-type strain under salinity stress (0.1580 ± 0.0009 h −1 ).
Our data differ from previous reports that have proposed that eliminating the NTH2 gene resulted in no salt tolerance [13]. However, the results are not comparable due to methodological differences of complete deletion of the gene versus a point mutation. We used 0.85 M NaCl stress from the beginning, an automatic growing system under constant salt stress sampling every 15 min beginning at time 0 ( Figure 4) and validated the tolerance in the semisolid plate ( Figure 5). The growth rate is also not comparable with our control strain CENPK2 having half the growth rate in non-stress conditions of 0.2255 ± 0.0037 h −1 versus 0.1580 ± 0.0009 h −1 in stress, and nth2 1271_1272delTA 0.2327 ± 0.0057 h −1 versus 0.2179 ± 0.0061 h −1 in stress. We believe that these automatic results and visual colony growth can capture the behavior of the mutation while reducing human error [33].
We foresee the disruption of NTH2 to provide stress tolerance in an industrial strain, because undisrupted NTH1 can provide the metabolic equilibrium as described next. NTH1 and NTH2 are required and regulated for fueling growth. NTH1 is phosphorylated by Cdk1(S66) and PKA1 (S20, S21, S60, S83) to be activated, and is required for fueling biosynthesis during S, G2, and M [34]. NTH2 contains an N terminal phosphorylation region (R49, S52, R109, S112) and is expressed at a high level in the stationary phase after glucose exhaustion [15]. NTH2 and NTH1 are downregulated at the exponential phase and have a higher expression at the stationary phase [19,35]. The presence of salinity stress causes trehalose accumulation in S. cerevisiae and higher ethanol osmotolerance [13,25]. Heat stress (40C), CuSO 4 , NaAsO 2 , H 2 O 2 , cycloheximide (CHX) but not NaCl (1.5 M) trigger the expression of NTH1 and, in practice, its disruption is unrequired for salt tolerance [36]. It is also known that strains with NTH2 disruption, previously named YBR0106, grow normally in YEP glycerol and were associated with increased sensitivity against heat shock at 50 • C [14], while ∆nth1 grows poorly in YEP glycerol and cannot mobilize endogenous trehalose [24]. NTH1 disruption may provide some stress tolerance but not related to NaCl, as previously reported [22]. However, disruption of NTH1 may not be useful for industry because these mutants cannot hydrolyze trehalose after returning from a heat stress temperature of 40 • C to an average growth temperature of 30 • C [15].
NTH2-disrupted mutants can mobilize and use trehalose. Its mutation results in an increased acid trehalase activity [19], meaning that the metabolic stability of the strain is not compromised. In addition, no significant change in intracellular trehalose occurs when NTH2 is eliminated under osmotic NaCl stress such as in our results [12,22]. The latter also means that trehalose negatively affects growth, for overaccumulation is unfeasible [16].
Interestingly, in Cryptococcus neoformans, the disruption of NTH2 alone increased the survival ability of the yeast, but the deletion of NTH1-NTH2 was negative for the microorganism [40]. Similarly, a database of yeast mutants growth modeling completed with kanMX4 interrupting NTH2 in haploid BY4741 background predicts tolerance to salt stress, such as our results [30].
In our NTH2 mutation model, an alternative explanation is that nth2 1271_1272delTA translates the gene into two ORFs considering that yeast produces alternative ORFs [41][42][43]. In that case, the ORFs from nth2 1271_1272delTA may not be active but may be able to bind the substrate. The protein fragments could transitorily protect trehalose from catabolism. The first ORF, containing binding sites R331, N375, R384, 309-10WD, and 355-357RSQ but not the active sites, could bind to the trehalose and protect the molecule from the enzymatic activity of NTH1.
Breeding industrial yeast can result in cost-effectiveness or reductions in fermentation. In the case of stress tolerance traits, yeast is constantly exposed to ethanol toxicity, oxidative stress, temperature stress, and osmotic stress, diminishing its capacity to produce ethanol [44,45]. Our data also show that the osmotic tolerance of the nth2 1271_1272delTA disruption strain mediated by CRISPR is superior and could represent a solution for the fermentation industry without compromising its metabolism, phenotype, or behavior [46,47].

Conclusions
The S. cerevisiae NTH2 gene was disrupted with the CRISPR/Cas9 technique, resulting in a nth2 1271_1272delTA phenotypically normal strain that could grow under osmotic stress (0.85 M sodium chloride).