Fig 1.
Cartoon of gene tagging methods in yeast.
(A) Commonly used methods for tagging a gene of interest (yfg = your favorite gene): C-terminal gene tagging using a marker and 40–50 bp homology (Fig 1A-i) or ≥300 bp homology (Fig 1A–ii), and N-terminal gene tagging using the Cre-loxP system (Fig 1A-iii). Methods i and ii use a selection marker that disrupts the 3’ UTR and cannot be eliminated, which might perturb the function of the fusion. For method iii, the loxP-flanked selection marker can be excised with the Cre recombinase, which leaves behind one flippase recognition target (FRT) site “scar” in the 5’ UTR of the tagged ORF. (B) Tagging methods introduced in this study: scarless C-terminal gene tagging (Fig 1B-i), scarless N-terminal gene tagging (Fig 1B-ii), and scarless internal gene tagging (Fig 1B-iii). The scarless tagging methods require a second round of selection to eliminate the URA3 marker, which is surrounded by identical GFP sequences. Note that the resulting “GFP scar” becomes a genuine part of the full-length GFP fusion protein after recombination and that the endogenous UTRs are not altered. The scarless N-terminal tagging method can be used for tagging essential genes in haploid yeast cells because a constitutive promoter situated between the URA3 marker and the second GFP fragment drives expression of the ‘partial GFP’-tagged gene prior to excision of URA3. Integration cassettes with either a partial GFP tag (as shown here) or a full-length GFP tag (S1 Fig) can be used. (C) Detailed description of the steps for scarless C-terminal tagging of a gene of interest with GFP. The integration cassette is built using two-step PCR synthesis. The homology arms H1 and H2 are amplified in the first round of PCRs and then used as primers together with F1 and R2 in the second round of PCRs. The primer-binding sites of H1 and H2 are unique, which is important for the efficient PCR amplification of the integration cassettes. Excision of the URA3 marker is not shown in this cartoon but is identical to the depiction above (Fig 1B-i).
Fig 2.
DNA gels and Western blot confirm strain construction.
(A) PCR amplification of homology arms H1 and H2. (B) Integration cassettes for C-terminal tagging of PRE1 with mNeonGreen (mNG) and the TRP1 selection marker using either 50-bp or ≥300-bp homology. (C) Integration cassette for scarless C-terminal tagging of PRE1 with mNeonGreen using ≥300-bp homology. (D) Colony PCR confirming correct chromosomal integration of the fusions at the PRE1 gene locus. (E) Integration cassette for the scarless N-terminal tagging of YPT1 and RSP5 with mNeonGreen. (F) Colony PCR confirming correct integration of the mNG1-177-URA3-PEF1α-mNG60-236 construct at the YPT1 gene locus with primer sets for the upstream (left) and downstream flank of the integration site (middle). Growth in rich medium followed by selection on 5-FOA leads to the identification of cells that have recombined the mNeonGreen fragments and lost the URA3 marker, resulting in YPT1 tagged N-terminally with mNeonGreen (right). Colony PCR of a strain with an unmodified, wild-type YPT1 gene locus is shown as control. (G) Colony PCR confirms correct integration of the mNG1-177-URA3-PEF1α-mNG60-236 construct at the RSP5 gene locus (left and middle). Recombination results in full-length mNeonGreen-RSP5 at the endogenous gene locus (right). (H) Western blot analysis using an anti-Rsp5 antibody (upper) with S. cerevisiae strains expressing untagged Rsp5 (lane #1), mNeonGreen-Rsp5 (lane #2), mNeonGreen60-236-Rsp5 (lane #3, before URA3 excision), mCherry-Rsp5 (lane #4), Rsp5-mNeonGreen (lane #5), untagged Rsp5 and an ectopic copy of mNeonGreen-Rsp5 (lane #6), and untagged Rsp5 and an ectopic copy of Rsp5-mNeonGreen (lane #7). The mNeonGreen60-236-Rsp5 fusion displays extensive self-ubiquitination, likely because of the partial mNeonGreen tag. The C-terminal Rsp5 shows a truncation product that most likely corresponds to un-tagged Rsp5. This truncation product is not observed with the N-terminal Rps5 fusions, which appear fully functional. The molecular weights are 92 kDa for untagged, wild-type Rsp5 and ~119 kDa for the fluorescent protein-tagged Rsp5 fusions. Pgk1 is used as loading control (lower). (I) Spotting assay with the strains that were used for the Western blot of Fig 2H.
Fig 3.
Fluorescence microscopy of S. cerevisiae cells producing mNeonGreen-tagged fusion proteins.
Yeast strains were constructed by modifying the endogenous gene locus using C-terminal gene tagging with a marker (Fig 1A-ii), scarless C-terminal gene tagging (Fig 1B-i), or scarless N-terminal gene tagging (Fig 1B-ii). PRE1 and TDH3 were tagged C-terminally with mNeonGreen (mNG) using the TRP1 selection marker or scarless. The eleven yeast Rab proteins (i.e. Sec4, Vps21, Ypt1, Ypt6, Ypt7, Ypt10, Ypt11, Ypt31, Ypt32, Ypt52, and Ypt53) are prenylated at their C-termini (S3 Fig) and were tagged N-terminally with mNeonGreen using the scarless gene tagging method. The Rab proteins localize to their expected organelles: for example Golgi (Ypt1, Ytp6), trans-Golgi network (Ypt6), vacuole and late endosomes (Ypt7), early endosomes (Vps21), recycling endosomes and post-Golgi exocytic vesicles (Ypt31), and secretory vesicles (Sec4). The ubiquitin ligase RSP5 was tagged C-terminally and scarless N-terminally with mNeonGreen. The C-terminal RSP5-mNeonGreen fusion resulted in cell size enlargement (S4 Fig) but no growth phenotype was observed (Fig 2I and S5 Fig). The large-cell phenotype was not observed with the scarless N-terminal fusions. The cells producing mCherry-Rsp5 and mCherry-Ypt1 fusions look similar to the cells producing the corresponding N-terminal mNeonGreen fusions, although the mCherry fusions also display some vacuolar localization. PRE1, YPT1, SEC4, and RSP5 are essential genes. Scale bar (white) is 5 μm.
Fig 4.
Scarless C-terminal tagging of a gene of interest in S. pombe.
(A) Homology arms (H1 and H2) targeting a gene of interest were attached to a DNA cassette comprising a linker (L), NmGFPmut3 (NGFP), the ura4 marker, and mGFPmut3 (GFP). The cassette was transformed into S. pombe cells followed by selection on a PMG-ura plate. The surviving colonies were cultured in YES medium for 3 days to allow spontaneous recombination between the GFP fragments and then plated on 5-FOA to select for cells that have yfg tagged scarless with linker-mGFPmut3. (B) The same method described in panel a, except that kanMX6 and HSV1tk were used as positive and negative selection markers, respectively. (C) After selection with G418, tdh1 was tagged with NmGFPmut3, which is non-fluorescent and hence no fluorescence was detected by microscopy. (D) After 3 days of culture in YES medium, followed by selection with FUdR, cells with the recombined full-length mGFPmut3 protein attached to yfg showed the expected GFP signal. Scale bar (white) is 5 μm.
Fig 5.
Pseudocode describing the automated primer design algorithm.
The illustration depicts the primer design for C-terminal gene tagging of a gene of interest (yfg) with a fluorescent protein. N-terminal and internal gene tagging is analogous except for the insertion site. The primers are shown as arrows (red) above and below the respective DNA sequences. GC clamps in the 500-bp upstream and downstream search regions are highlighted in red. The primer sequences for scarless N-terminal, scarless C-terminal, and C-terminal gene tagging with a marker for essentially all S. cerevisiae ORFs are provided in S4 Table. The primer sequences for scarless C-terminal tagging of all S. pombe ORFs are provided in S5 Table. The code for the automated primer design is contained in S1 File and available on GitHub (https://github.com/DXL38/scarless_gene_tagging_in_yeast). Primers F1 and R1 are used for amplifying the homology arm H1, and primers F2 and R2 are used for amplifying the homology arm H2. The homology arms H1 and H2 are ≥300 bp in size.