Plasmid construction

In non-integrating yeast vectors, auxotrophic markers are often placed under the control of the A. gossypii P_TEF1 promoter. Thus, we first constructed an A. gossypii P_TEF1-S. cerevisiae FEX1-T_MFα1 cassette in the pITy backbone through two steps. First, primers 1 and 2 were used to append EcoRI and EagI sites to an A. gossypii P_TEF1 fragment amplified from pSVA13 (Supplementary Table 3). This fragment and pITy A₂aGH were digested with EcoRI-HF and EagI-HF and ligated to generate an intermediate pITy A. gossypii P_TEF1 A₂aGH construct. Second, primers 3 and 4 were used to amplify S. cerevisiae FEX1 from BJ5465 gDNA extracted using the protocol provided by Lõoke et al.⁴⁴. This fragment and pITy A. gossypii P_TEF1 A₂aGH were digested with EagI-HF and AflII and ligated to generate pITy A. gossypii P_TEF1-S. cerevisiae FEX1-T_MFα1. USER⁴⁵ cloning was used to subclone the A. gossypii P_TEF1-S. cerevisiae FEX1-T_MFα1 cassette, which was amplified using primers 5 and 6, into the pYC2/CT and pYES2 backbones amplified using primer pairs 7 and 8 and 8 and 9, respectively. The resulting constructs were named pCFEX1 A₂aGH and pEFEX1 A₂aGH, respectively. The A. gossypii P_TEF1 promoters were swapped with P_PGI1 and P_REV1 promoters amplified from BJ5465 gDNA using primer pairs 10 and 11 and 12 and 13, respectively. pCFEX1 A₂aGH and pEFEX1 A₂aGH backbones were amplified using primer pairs 7 and 14 and 9 and 14, respectively, to mediate promoter swapping through USER cloning. The pCFEX2 and pEFEX2 backbones carry the P_PGI1 promoter, and the pCFEX3 and pEFEX3 backbones carry the P_REV1 promoter. The integrating pIFEX A₂aGH construct was generated through USER cloning after amplifying pITy A₂aGH using primers 15 and 16 and the A. gossypii P_TEF1-S. cerevisiae FEX1-T_MFα1 cassette using primers 5 and 6. The pIFEX A₂aGH construct was designed to retain the NEO CDS to facilitate cloning in Escherichia coli using kanamycin; thus, the vector also confers G418 resistance to yeast. All constructs were sequence-verified using Sanger sequencing (Genewiz) and transformed into S. cerevisiae using the high-efficiency lithium acetate protocol⁴⁶. We found that a recovery period is necessary to obtain yeast transformants using the FEX vectors. Following resuspension in YPD, transformed yeast cells were incubated at 30 °C for at least 3 h prior to plating on YPD plates containing 210.5 µM NaF. Yeast transformed with pIFEX A₂aGH were plated on YPD supplemented with 10 mM NaF to promote increased gene dosage. The pRS315 P_TEF1-yEGFP construct was generated from pRS315 P_TEF1-yEGFP-Cln2, which was cloned for a separate study, in a series of steps. First, the pRS315 backbone was digested with EagI, blunted with Klenow fragment, and digested with SpeI. The P_TEF1-yEGFP-Cln2 cassette was amplified from YEp351 P_TEF1-yEGFP-Cln2 using universal M13 forward and reverse primers, then the amplicon was digested with SpeI to mediate directional cloning into pRS315. The backbone and insert were ligated and transformed into E. coli. Next, USER cloning mediated construction of pRS315 P_TEF1-yEGFP from the Cln2-tagged plasmids. Primers 17 and 18 were used to amplify the plasmid excluding the Cln2 tag and introducing two stop codons at the 3′-end of the yEGFP coding sequence.

Yeast strains and culturing conditions

S. cerevisiae strain BJ5465 (Mata ura3-52 trp1 leu2∆1 his∆200 pep4::HIS3 prb∆1.6R can1) (ATCC) was used to construct the biocontainment strain BJ5465 fex1::GSHU ∆fex2 using the Delitto Perfetto method⁴⁷. Primers 19 and 20 were used to integrate the GSHU cassette at the FEX1 locus, and primers 19 and 21 were used to integrate the CORE-Kp53 cassette into the FEX2 locus. Subsequently, primers 22 and 23 were used to remove the CORE-Kp53 cassette. Culture maintenance and gene expression were carried out using YPD medium at 30 °C with shaking at 225 r.p.m. Cultures harboring pCFEX and pEFEX plasmids were maintained in YPD supplemented with 2 mM NaF.

Fluorescence-activated cell sorting

Yeast cultures were diluted to an OD₆₀₀ = 1.0 in 1× phosphate buffered saline (PBS) prior to all fluorescence-activated cell sorting (FACS) analyses. Approximately 60,000 cells were analyzed from each sample using a 488-nm laser and 530/30 nm bandpass filter using the gating strategy illustrated in Supplementary Fig. 4. All analyses were conducted using a BD FACSAria I flow cytometer and FlowJo v10. To analyze yEGFP expression, WT and biocontainment strains harboring pRS315 P_TEF1-yEGFP were used to inoculate 5 mL synthetic dextrose medium supplemented with amino acids lacking leucine (SD -leu)⁴⁷. Following overnight growth, cells were resuspended in 1× PBS and analyzed using FACS as described above. To analyze A₂aR-GFP expression from pCFEX and pEFEX vector backbones, knockout strains carrying the vectors were first cultured overnight in YPD medium supplemented with 2 mM NaF at 30 °C with shaking at 225 r.p.m. Following overnight growth, each culture was subcultured into YP medium supplemented with 2% (w/v) raffinose (YPR) and 2 mM NaF at an initial OD₆₀₀ of 0.5. Following ~10 h of shaking at 30 °C, each culture was subcultured into YP medium supplemented with 2% (w/v) raffinose, 2% (w/v) galactose (YPRG), and 2 mM NaF to induce A₂aR-GFP expression. Cultures were incubated with shaking at 30 °C overnight prior to flow cytometric analysis. Analysis of A₂aR-GFP expression from the pIFEX backbone was accomplished using a similar induction scheme in the absence of NaF. Knockout strains harboring integrated pIFEX A₂aR-GFP cassettes were cultured in YPD medium overnight at 30 °C with shaking at 225 r.p.m. Subsequently, cultures were subcultured into YPR medium and incubated at 30 °C with shaking. After ~10 h, A₂aR-GFP expression was induced in each culture through subculturing into YPRG medium and incubation at 30 °C with shaking overnight.

Dilution spotting

Yeast cultures were grown to an OD₆₀₀ ~ 3 prior to dilution to an OD₆₀₀ = 2.5 in sterile YPD. Diluted cells were used to prepare serial dilutions up to 10⁻⁵ in tenfold increments. A total of 5 µL of each dilution was spotted onto solid media using a multichannel pipette. Plates were allowed to dry at room temperature prior to overnight incubation at 30 °C.

Fluoride dose response assay

WT and biocontainment strain cultures were grown in biological triplicate overnight in YPD at 30 °C with shaking at 225 r.p.m. In the morning, the cultures were used to inoculate 5 mL fresh YPD at an initial OD₆₀₀ of 0.15. Cultures were incubated with shaking at 30 °C for 7 h, reaching OD₆₀₀ values near 2, and used to inoculate 3 mL YPD in individual wells of a 24-well block (Qiagen #19583) containing serially diluted concentrations of NaF and covered with a Breathe Easier sealing membrane (Sigma-Aldrich Z763624). Following overnight shaking at 30 °C, OD₆₀₀ values were measured for cultures in each well.

Growth curves

To generate growth curves, the WT and biocontainment strains were used to inoculate 5 mL YPD cultures, which were grown overnight at 30 °C with shaking at 225 r.p.m. Cultures were used to inoculate 1 mL YPD in individual wells of a 24-well plate (Corning 3526), at an initial OD₆₀₀ = 0.02. Cell growth was monitored using a Tecan Spark microplate reader maintained at 30 °C with orbital shaking at 180 r.p.m. and 3 mm amplitude. OD₆₀₀ measurements were taken every 10 min with 50 ms settling time prior to each reading. Specific growth rates were calculated by fitting data to the logistic function⁴⁸ (Eq. (1)):

where N₀ is the initial number of cells, N_asymp is the maximal number of cells approached during stationary phase, k is the growth rate, and t_c is the time at which the growth curve exhibits an inflection point.

Escape rate determination

To determine the escape rate, the biocontainment strain was grown overnight in biological triplicate in YPD at 30 °C with shaking at 225 r.p.m. In the morning, ~50 colony forming units (CFUs) of each replicate culture were plated onto YPD media assuming a conversion factor of 10⁷ CFU/mL/OD₆₀₀. Using the same conversion factor, 10⁸ and 10⁹ CFUs of each replicate culture were plated onto YPD supplemented with 210.5 µM and 5 mM NaF. Following incubation of plates at 30 °C for 2 days, CFUs were counted. The CFU values obtained for the YPD control plates were used to correct the CFU/mL/OD₆₀₀ conversion factor and to calculate the total number of cells plated on each plate.

pH Buffering experiment

Individual colonies were used to inoculate 5 mL YPD medium. Cultures were grown overnight at 30 °C with shaking at 225 r.p.m. Following overnight growth, cells were used to inoculate YPR at an initial OD₆₀₀ = 0.2. Cultures were grown for 7–8 h at 30 °C with shaking at 225 r.p.m. and used to inoculate 5 mL YPRG supplemented with 10 mM NaF at an initial OD₆₀₀ = 0.2. One set of cultures were used to inoculate 5 mL YPRG buffered to pH = 6 using 100 mM MES (Sigma-Aldrich, St. Louis, MO, USA). Following overnight growth, cells were prepared for FACS analysis as described above.

Microbiostatic/microbiocidal assay

WT and biocontainment strain cultures were grown overnight and subcultured into 5 mL YPD at an initial OD₆₀₀ of 0.1. Cultures were incubated with shaking at 30 °C for ~10 h to an OD₆₀₀ between 2 and 4 and used to inoculate 3 mL YPD in individual wells of a 24-well block containing serially diluted concentrations of NaF and covered with a Breathe Easier sealing membrane. Following overnight incubation with shaking at 30 °C, the OD₆₀₀ values were measured for cultures in each well. The equivalent of 0.2 OD-mL of cells were spun at 3000 × g for 30 s, washed with sterile 1× PBS, and used to inoculate 3 mL fresh YPD in a sterile well of a 24-well block, which was covered with a Breathe Easier sealing membrane. Upon overnight growth, the OD₆₀₀ values were measured for cultures in each well.

Modeling cellular fluoride uptake

Fluoride uptake was approximated by modeling the transport of HF across the cell membrane. First, the concentration of HF in bulk is calculated from the exogenous NaF concentration using Eq. (2):

The above equation takes into account the pKa of HF at 25 °C, which is equal to 3.17. Next, the general transport equation (Eq. (3)) can be solved for the time-dependent concentration of NaF inside of the cell:

where C is the time-dependent concentration of fluoride and D is the diffusion coefficient of fluoride. Solving Eq. (3) for C:

where C_i(t) is the intracellular fluoride concentration, C₀ is the bulk fluoride concentration, A is the membrane surface area, P is the permeability constant, and V is the cell volume. The permeability constant used for HF in the cell membrane is 0.0002 cm/s as calculated by Gutknecht et al.⁴⁹. The cell surface area and volume were estimated from figures provided on bionumbers.hms.harvard.edu.

Now, the fluoride concentrations can be used to solve for the flux, J, of fluoride across the cell:

Now, Eq. (6) gives the flux of fluoride across the cell membrane given a bulk concentration of fluoride, which is dictated by the exogenous NaF concentration as calculated in Eq. (2).

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.