Puromycin- and methotrexate-resistance cassettes and optimized Cre-recombinase expression plasmids for use in yeast

Abstract

Here we expand the set of tools for genetically manipulating Saccharomyces cerevisiae. We show that puromycin-resistance can be achieved in yeast through expression of a bacterial puromycin-resistance gene optimized to the yeast codon bias, which in turn serves as an easy-to-use dominant genetic marker suitable for gene disruption. We have constructed a similar DNA cassette expressing yeast codon-optimized mutant human dihydrofolate reductase (DHFR), which confers resistance to methotrexate and can also be used as a dominant selectable marker. Both of these drug-resistant marker cassettes are flanked by loxP sites, allowing for their excision from the genome following expression of Cre-recombinase. Finally, we have created a series of plasmids for low-level constitutive expression of Cre-recombinase in yeast that allows for efficient excision of loxP-flanked markers.

Figure 1. Puromycin-resistance as a selective marker in yeast

A) Wild-type and pdr5Δ cells were grown to mid-log phase in SC media before addition to a 2-fold serial dilution of puromycin in a 96-well plate. Each well contained a 250 μl culture of yeast cells at an OD₆₀₀ = ~0.005. The puromycin gradient ranged from 40 mM to 0.04 mM, with a final well containing no drug. Cells were incubated at 30°C for 72 hours, mixed, and then the OD₆₀₀ was measured. Growth of wild-type (blue) and mutant pdr5Δ (green) cells is indicated. B) Schematic diagram of homologous recombination strategy to replace the Kanamycin-resistance (kan^r) cassette at the UBI4 locus with that of the pac gene conferring Puromycin-resistance (puro^r), retaining the original TEF1 promoter and 3′ untranslated (UTR) terminator sequences for expression of the new marker. C) ubi4Δ::kan^r cells (also containing pdr5Δ mutation) were transformed with either a PCR product encoding the pac gene or a carrier DNA control. Cells were plated on YPD plates containing 4 mM puromycin and colonies were imaged after 48 hours. D) Puromycin-resistant clones from (C) were patched onto plates containing puromycin or G418. As a control the parental kan^r is included. E) PCR analysis of the UBI4 locus from genomic DNA isolated from wild-type yeast or yeast containing either the kan^r or puro^r cassettes at the UBI4 locus. F) Growth assay as described in (A) of wild-type cells expressing TEF1-pac from a low-copy CEN plasmid (pPL5103) (red). The sensitivity of wild-type cells to puromycin from (A) is indicated (dotted blue).

Figure 2. Mutant DHFR confers methotrexate-resistance in yeast

A) Wild-type cells were grown to mid-log phase before addition to the wells of a 96-well plate containing a 2-fold serial dilution of methotrexate. After 72 hours incubation at 30°C growth measured by absorbance at 600 nm was calculated as a percentage of yeast growth in media lacking drug. Growth across the methotrexate gradient is indicated (red), as is growth in the same gradient containing 5 mg/ml sulphanilamide (blue). B) The crystal structure of human DHFR in complex with the methotrexate ligand, (PDB accession number 1U72) is depicted. hDHFR is shown in grey cartoon format, the two mutations that reduce affinity to methotrexate (L22F and F31S) are shown in cyan and an arrow indicates the methotrexate-binding pocket. The location of an additional mutation (F180L) acquired in the selection for methotrexate-resistant yeast clones is shown in magenta. C) Schematic diagram of homologous recombination strategy to replace the his5⁺ cassette between the TEF1 machinery at the BRO1 locus with that of the mutant form of human DHFR carrying the L22R and F31S mutations (DHFR* / mtx^r). D) bro1Δ::his5⁺ cells were transformed with either a PCR product encoding DHFR* or carrier DNA control. Cells were plated on SC plates containing 20 nM methotrexate and 5 mg/ml sulphanilamide. E) Methotrexate-resistant clones from (D) and the parental bro1Δ::his5⁺ strain were patched onto plates containing methotrexate and sulphanilamide and plates lacking histidine (-His). F) Genomic DNA isolated from wild-type, bro1Δ::his5⁺, and bro1Δ::mtx^r yeast were analysed by PCR using oligonucleotides in the 5′ and 3′ untranslated regions of the BRO1 locus.

Figure 3. Puromycin and methotrexate selections on solid agar media

A) Haploid and diploid cells, both wild-type parents and strains carrying a puro^r cassette, were grown to mid-log phase before dilution and plating. Cultures were plated on rich media containing either 8 mM or 20 mM puromycin, and a control YPD plate that lacked drug. B) Wild-type cells or strains carrying gene deletions indicated were grown to mid-log phase before plating on YPD plates containing different concentrations of puromycin. C) Haploid and diploid cells, either wild-type or containjing a mtx^r cassette were grown in liquid culture, diluted and then plated on SC media. Cultures were also plated on SC media containing 5 mg/ml sulphanilamide and either 10 nM or 80 nM methotrexate.

Figure 4. Expression of cre-recombinase from a constitutive promoter

A) Schematic diagram depicting the homologous recombination strategy used to create TEF1*-cre plasmid (pPL5071) using the GAL1-cre parent plasmid, pSH47 (Güldener, et al., 1996). B) Confirmation of the replacement of GAL1 promoter with that of TEF1* by PCR using oligonucleotides that anneal in common regions either side of the promoter sequences. The slightly smaller PCR product generated from the TEF1* template migrates faster. HpaI digestion, a restriction site unique to the TEF1* promoter, was used to confirm the GAL1 promoter has been exchanged. C) Yeast transformed with the GAL1-cre plasmid were grown to mid-log phase in SC-Ura media supplemented with either galactose (Gal), raffinose (Raf) or glucose (Glu). Yeast retaining the TEF1*-cre plasmid were also grown in glucose containing SC-Ura media. Cells were harvested and treated with 0.2 N NaOH for 3 minutes before lysates were generated by addition of Laemmli sample buffer containing 8 M urea. Lysates were resolved by SDS-PAGE followed by transfer to nitrocellulose and probing with polyclonal antibodies raised against Cre and CPY. D) Quantitation of Cre levels driven from either the GAL1 promoter in different carbon sources or the TEF1* promoter. Densitometry of Cre immunoblots was carried out using Fiji image analysis software and the mean and SD shown from 3 experiments are shown. E) bro1Δ::loxP-his5⁺-loxP cells were transformed with TEF1*-cre (white) or GAL1-cre (black) plasmids were selected for on glucose containing SC-Ura plates. Single colony transformants (n = 48 each) were patched on SC-His plates and the percentage of cells that had successfully excised the loxP-his5⁺-loxP marker are shown. Transformants were also washed 3x in SC-Ura galactose media and then grown for 1 hour or 8 hours in galactose media before being washed 3 times in SC-Ura glucose media to repress GAL1 induction of Cre. Cultures were struck out on SC-Ura glucose plates and the percentage of single colony transformants that had excised the his5⁺ marker were identified by patching on SC-His plates. F) The triple null ade2Δ met15Δ sna3Δ strain containing loxP flanked puro^r, mtx^r and his5⁺ markers, respectively, was grown on YPD plates, YPD 20 mM puromycin, SC + 50 nM methotrexate, or SC-His plates. As a control the parental wild-type strain By4742 was included. The excision of loxP markers from ade2Δ met15Δ sna3Δ cells transformed with either TEF1*-cre (white) or GAL1-cre (black) is shown. Cre efficiency was calculated by PCR analysis of the ADE2, MET15 and SNA3 loci from genomic DNA isolated from single colony transformants of each plasmid (24 each), shown in supplemental Figure S3. G) Plasmid map of pPL5071 encoding TEF1*-cre from a low-copy URA3 marked CEN plasmid. H) Table of TEF1*-cre plasmids including the size of each vector, the positive selection marker, and the means to remove plasmid.

Figure 5. Use of new dominant markers and TEF1*-cre plasmids

A) The ADE2 gene of wild-type yeast (BY4742) was deleted using the loxP-puro^r-loxP cassette. The puro^r cassette was removed using pPL5071, a URA3 marked plasmid expressing cre-recombinase from a mutant TEF1* promoter. pPL5071 was removed from ade2Δ::loxP cells by growth on 5-FOA, single colonies were selected and plated on 5-FOA and -Ura media to confirm plasmid loss. B) As with (A) the ADE2 gene of BY4742 was deleted, this time using a URA3 deletion cassette generated from PCR template pUG72 (Gueldener, et al., 2002). The URA3 marker was excised by cre expression from the ADE2 marked version (pPL5606), which was subsequently removed by growth on rich media and screening for red single colonies. Cells were plated on SC minimal media and incubated for an extra 24 hours to develop red colour for clarity. C) The MET15 gene of BY4742 was deleted using the loxP-mtx^r-loxP dominant marker cassette. The mtx^r cassette was removed by expression of cre-recombinase from pPL5627, which carries a MET15 marker. This plasmid was removed by growth on rich media, and cells lacking the plasmid could be easily identified on modified YPD media containing Pb²⁺ ions. Cells were incubated for an extra 48 hours at 4°C to develop colour difference for documentation in this image. D) The TRP1 locus of BY4742 was deleted using a his5⁺ cassette amplified from pUG27 (Gueldener, et al., 2002). The marker was removed using the TRP1 marked cre expression plasmid (pPL5608), which was later removed by selective growth on SC media containing 5-FAA.