A new strategy for seamless gene editing and marker recycling in Saccharomyces cerevisiae using lethal effect of Cwp1

Abstract Technologies development for seamless gene editing and marker recycling has allowed frequent genomic engineering in Saccharomyces cerevisiae for desired laboratory strains and cell factory. Alternative new approaches are still required for complicated scenarios. In this study, we report that inducible overexpression of cell wall protein 1 (Cwp1) by galactose addition confers yeast cells a robust growth inhibition. Direct repeats flanking the Gal‐CWP1:selectable marker cassette allow for its homology recombination excision and counter selection upon galactose addition, therefore enable seamless gene editing and marker recycling. We used this strategy and efficiently generated scarless Ade8 deletion mutants. Our results highlight the utility of lethal effect of Cwp1 overexpression a new counter selection strategy and a simple and efficient method for seamless gene editing and marker recycling in S. cerevisiae and potentially other fungi.


| INTRODUC TI ON
The budding yeast Saccharomyces cerevisiae is an attractive model organism for fundamental biological research and powerful cell factory for industrial application (Dikicioglu, Pir, & Oliver, 2014;Hong & Nielsen, 2012). Complex and multiple genomic engineering in S. cerevisiae therefore turns commonplace. However, genetic technologies innovation is still needed to enable simple and extensive genetic manipulations in such model organism.
As each modification such as gene deletion, insertion, or tagging retains one selectable marker in most current methods, it therefore presents a hurdle when dozens of genetic changes are required.
Especially, industrial strains hardly modified for the use of auxotrophic selection because they are typically aneuploidy or polyploidy (Querol & Bond, 2009), thus only depends on very restricted dominant markers. Second, even if it is sufficient in use, selectable markers may have deleterious effects and interfere with physiology of host cell (Gopal, Broad, & Lloyd, 1989). Additionally, not introducing heterologous adaptors and scarless modification is always the best criterion for genomic engineering, especially for strains dedicated to food and biopharmaceutical industry.
To overcome these defects, recyclable marker and seamless genetic manipulation in S. cerevisiae gets rapidly developed. Both tools are getting essential in the booming field of synthetic biology and also in biotechnology industry. Several strategies for marker recycling have been in use, among which selectable markers rescue involves counter selection of auxotrophic markers is widely used for laboratory strains. Such methods include counter selection of colonies with spontaneous mutation or loss of a gene required for a specific nutrient such as URA3, LYS2, CYH2, and MET15, taking advantage of 5-fluoroorotic acid (FOA), α-aminoadipate, cycloheximide, and methyl-mercury, respectively (Alani, Cao, & Kleckner, 1987;Brachmann et al., 1998;Chattoo et al., 1979;Käufer, Fried, Schwindinger, Jasin, & Warner, 1983;Singh & Sherman, 1974;Struhl, 1983). However, the prerequisite to use this strategy is that the original strain should be respective auxotrophic, which is hard to work for many strains, especially for prototrophic industry strains.
In the present study, we found that inducible overexpression of cell wall protein 1 (Cwp1) by galactose addition confers yeast cells a robust growth inhibition. We took advantage of the fact that short repeat sequences make homology recombination efficient in budding yeast and designed direct repeats flanking the Gal-CWP1:selectable marker cassette allows for its homology recombination excision and counter selection upon galactose. We used this strategy and efficiently generated scarless Ade8 deletion mutants validating our method for seamless gene editing and marker recycling. We therefore reported a simple and efficient method for seamless gene editing and marker recycling in S. cerevisiae using lethal effect of cell wall protein overexpression and hence potentially in other fungi.
Galactose induction medium for overexpression of Cwp1 includes 1% yeast extract, 2% peptone, and 2% glactose. For serial dilution assays, exponentially growing cultures at 30°C were spotted on the indicated plates to the same concentration using a 10-fold serial dilution as described previously .

| Constructs generation and deletion cassette construction
pRS306-GAL-CWP1 and pFA6a-GAL-CWP1-KanMX constructs were assembled by inserting GALl/GALl0 promoter and CWP1-13MYC fragments into pRS306 and pFA6a-KanMX plasmids, respectively, using AFEAP cloning method . Deletion cassettes were generated by PCR using Q5 hot start high-fidelity DNA polymerase (NEB) and following manufacturer recommendations. Primers used for deletion cassette construction were designed as follows. The forward primers (5'-3') contain a ~55 bp sequence homologous to the region upstream the start of the fragment to delete, a followed ~55 bp sequence homologous to the region downstream the end of the fragment, and a ~20 bp sequence annealing to the upstream of plasmid template region. The reverse primers contain a sequence homologous to the last ~55 bp of the fragment to delete and a ~20 bp sequence annealing to the downstream of plasmid template region.
PCR products as the deletion cassette were transformed into the host strains using LiAc chemical transformation. For ADE8 deletion cassette, primer pairs oYX7 and oYX8 using pRS306-GAL-CWP1 as template and primer pairs oYX13 and oYX14 using pFA6a-GAL-CWP1-KanMX as template were used, respectively. Primer pairs oZC19 and oZC20, oligo nucleotides complementary to the upstream and downstream sequences of the deletion part. The primer sequences were listed in Table 1.

| Western blotting
Whole cell extracts from indicated cultures were prepared by glass beads beating in trichloroacetic acid, then resolved by SDS-PAGE as TA B L E 1 Oigos used in this study previously described (Ren et al., 2016). The primary antibodies used in this study were anti-Myc (9E10, monoclonal mouse hybridoma supernatant).

| Calcofluor white staining and fluorescence microscopy
For cell wall observation, the indicated cultures were harvested and fixed with 70% ethanol. The fixed cells were washed and stained with a specific chitin stain calcofluor white 0.1% (Sigma-Aldrich) for 15 min at room temperature. Images were taken using a Delta Vision Elite microscope (Applied Precision Inc., Mississauga, ON, Canada) with Volocity software.

| Overexpression of Cwp1 causes cell wall division defects and lethal
Cell wall proteins are good candidates to be counter selection markers in yeast. Because induced overexpression of such proteins is promising to block yeast division due to cell wall division defects without cytotoxicity. We therefore screened a collection of cell wall proteins and found out Cwp1 as a good target. As shown in

| Use of inducible lethal effect of Cwp1 for seamless gene deletion and marker recycling
To make Cwp1 overexpression system attractive for yeast genetic manipulation, we took advantage of a strategy that applies homologous recombination between repeat sequences flanking a counter-selectable Cwp1 cassette for its excision in the genome. The proposed methodology is described as shown in Figure 2a. In all, this strategy makes GAL-CWP1 a nice seamless gene deletion and marker recycling system.

| Efficient seamless deletion of ADE8 by GAL-CWP1 system
To evaluate the proposed methodology, we chose ADE8 gene for this proof-of-principle experiment because the phenotype caused by ADE8 deletion can be visually screened. To delete ADE8 gene with GAL-CWP1 system, we amplified the GAL-CWP1:URA3 cassette from pRS306-GAL-CWP1 constructs using the primer pairs oYX7 and oYX8. Yeast cells w3031a YFL3 (WT) were transformed F I G U R E 2 Outline of Cwp1 overexpression mediated seamless deletion method. (a) Cassette design for targeted gene deletion and seamless marker removal. Three sequences (>40 bp) adjacent to the targeted gene are chosen for cassette design. Sequences I and II are adjacent to the targeted gene, upstream, and downstream, respectively. Sequences III is downstream and adjacent to II. PCR products of the designed cassette using primers and templates as shown in panel A were then transformed. First-round selection was done to get the strain with targeted gene replaced by GAL-CWP1 cassette taking advantage of the selection marker included in the cassette. Galactose was then added and used to counter select the cells without GAl-CWP1. (b) Cartoons depicting the structures of the indicated pFA6a-based and pRS306-based constructs for primers design to PCR up the GAL-CWP1 and marker cassettes. The primers contain two fragment sequences including the 5′ end sequence annealing to the template constructs and 3′ end sequence (I, II, and III) designed as the homologous arm to the targeted gene with PCR products of GAL-CWP1:URA3 cassette and selected on the synthetic media lacking uracil. Positive colonies with ADE8 replaced by GAL-CWP1:URA3 were white.
Mutants were then grown in YPD media for 6 hr, followed by galactose counter selection on YPGal plates. Colonies grown on the YPGal plates were randomly picked for following confirmation.
First, we checked the growth of the indicated mutant cells on SD-Uraplate. WT cells and all galactose selection colonies cannot grow on media lacking uracil (SD-Ura -) (Figure 3a), suggesting an efficient loss of GAL-CWP1:URA3 fragment upon galactose selection.
To confirm the efficiency of seamless deletion of ADE8 by GAL-CWP1 system, correct fragment integration was checked using a pair of primers with sequences complementary to the two sides of ADE8.
As shown in Figure 3b, WT control template produced a ~1 kb PCR products band, ade8 deletion mutants with ADE8 replaced by GAL-CWP1:URA3 produced a ~4 kb band. One hundred percentage of the colonies after galactose counter selection gave ~250 bp bands which confirmed a high efficiency of ADE8 seamless deletion in this system.
These short PCR products were confirmed by sequencing, further confirming the scarless deletion. This trial of deletion of ADE8 proved the proposed methodology using GAL-CWP1 system for seamless deletion.
As for most industry strains, the use of auxotrophic selection is difficult because they are typically aneuploidy or polyploidy (Querol & Bond, 2009). To expand the usage of this method for industry strains, we further repeated the evaluation using KanMX as the dominant selection marker instead of auxotrophic selection. As shown in Figure 3c,d, 100% of the colonies turns seamless ade8 deletion mutants using the proposed method.
To confirm the universality of GAL-CWP1 system for seamless deletion, we have taken advantage of this method and efficiently generated other gene deletion strains. Genes deleted using this method in our laboratory include SML1, a small gene with a 208 bp open reading frame size; MEC1, a large gene with ~7 kb in size and among others (will be published elsewhere). Therefore, this method is feasible and supposed to be universal for any gene deletion in yeast.
GAl-CWP1 system tends to be an ideal method for seamless gene deletion and marker recycling because of the following merits. First, CWP1 gene is derived from yeast genome and Cwp1 pro- In all, our results prove that cassette containing GAL-CWP1 and a selection marker with repeated flanking sequences can be used for yeast genomic editing without any scar left behind in the genome.
55 bp homologous sequences are long enough to ensure the efficiency of the two rounds homology recombination in this method and therefore make GAl-CWP1 system an efficient method for scarless gene deletion and marker recycling. Besides, cell wall protein genes are evolutionarily conserved among fungal species and the lethal effect might be also similar; therefore, this strategy could be used in other fungal species.

| CON CLUDING REMARK S
In this study, we found that inducible overexpression of Cwp1 by galactose addition resulted in strong lethal effect. We evaluated that

CO N FLI C T O F I NTE R E S T
The authors declare no financial or commercial conflict of interest.

AUTH O R S CO NTR I B UTI O N
FZ, ZH, and JD conceived the project and contributed to analysis and interpretation of data. FZ and YH participated in draft preparation and wrote the manuscript. YH, YJ, XZ, and ZY performed experiments. YH and YJ prepared the figures and wrote materials and methods section. All authors discussed and proofread the work and manuscript.

E TH I C S S TATEM ENT
We state that the ethics approval was not needed for this study, we still submitted our work to Ethics Committee in Hebei Agricultural University confirming no ethics issue related to our work.

DATA ACCE SS I B I LIT Y
All data from the manuscript are deposited in FigShare