A user-friendly and streamlined protocol for CRISPR/Cas9 genome editing in budding yeast

Summary CRISPR/Cas9 technology allows accurate, marker-less genome editing. We report a detailed, robust, and streamlined protocol for CRISPR/Cas9 genome editing in Saccharomyces cerevisiae, based on the widely used MoClo-Yeast Toolkit (https://www.addgene.org/kits/moclo-ytk/). This step-by-step protocol guides the reader from sgRNA design to verification of the desired genome editing event and provides preassembled plasmids for cloning the sgRNA(s), making this technology easily accessible to any yeast research group. For complete details on the use and execution of this protocol, please refer to Novarina et al. (2021).


Highlights
Complete and userfriendly protocol for CRISPR/Cas9 yeast genome editing sgRNA(s) and Cas9 cloning via Golden Gate Assembly with the MoClo-Yeast Toolkit Detailed description of repair fragment design for several genome editing applications Yeast one-step cotransformation with sgRNA(s)+Cas9 plasmid and repair fragment Figure 1. Schematic of the sgRNA(s) and Cas9 cloning through Golden Gate assembly The sgRNA and the Cas9 gene are cloned in a yeast expression vector through three consecutive Golden Gate assembly reactions. In the first Golden Gate assembly (GG1), partially overlapping annealed oligos containing the sgRNA sequence are cloned in the BsmBI-digested sgRNA dropout vector, yielding an sgRNA part plasmid (white colonies growing on chloramphenicol plates). In the second Golden Gate assembly (GG2), the sgRNA part is assembled with appropriate connectors in a second vector (AmpR-ColE1) after BsaI digestion, yielding an sgRNA cassette plasmid (white colonies growing on ampicillin plates). In the third Golden Gate assembly (GG3), the sgRNA cassette is assembled together with the Cas9 cassette and a spacer sequence (or a second sgRNA cassette) in a yeast expression vector (multi-gene dropout vector), yielding the Cas9+sgRNA(s) multi-gene plasmid (white colonies growing on kanamycin plates). Green dashed lines represent BsmBI restriction sites and blue dashed lines represent BsaI restriction sites.
1. Detect the possible sgRNA sequences in the E-CRISP ''Design'' section, using the parameters described below. a. Select organism: ''Saccharomyces cerevisiae R64-1-1''. b. Select target region by gene symbol or sequence: insert the name of the target gene or paste a specific genomic sequence in FASTA format. i. For gene knockout insert the gene name or ORF symbol.
ii. For N-terminal or C-terminal gene tagging, insert a sequence encompassing the 60 nucleotides upstream and the 60 nucleotides downstream the ATG starting codon (for N-terminal tagging) or the STOP codon (for C-terminal tagging). iii. For integration of a sequence at a specific genomic site, insert a 120-nucleotide sequence centered on the desired insertion site (or a longer sequence if the exact position of the integration site is not relevant). iv. For replacement of a genomic sequence, insert the sequence to be replaced. v. For point mutations, insert a $150 nucleotide sequence centered on the desired mutation site.
Note: Increasing the length of the input sequence increases the chance of finding a good sgRNA sequence (i.e., with high cleavage efficiency and no off-targets). However, for some genome editing applications (such as N-and C-terminal tagging or introduction of point mutations) cleavage should occur in close proximity of the genome editing site, in order to ensure efficient editing. For this reason, we suggested different lengths of the input sequence for different genome editing applications.
c. In the ''Start application'' section, select ''medium''. d. Click on the ''Display advanced options'' to adjust the design parameters. e. In the ''Design purpose'' section adjust the parameters as follows to optimize the analysis for S. cerevisiae (if not specified, leave the default setting). i. 5 0 preceding Base requirement: any (the default ''G'' comes from the cloning system used for human cells). ii. 3 0 PAM: NGG. iii. exclude targets with poly T motif (it is a transcription termination signal for Pol III, that might hinder transcription of the sgRNA in vivo). f. In the ''Gene annotation filtering'' section deselect all options. g. In the ''Off-target analysis'' section set the following parameters (for all the genome editing applications listed in step 1b). i. Exclude designs with more than X off-targets: 20. ii. Number of 5 0 mismatch positions: 6.
iii. Tolerated edit distance to the target sequence: 2. (C) Output of sgRNA evaluation. In the example, the evaluation of the same sgRNA shown in Figure 2A is shown. In the Matchstring column, a green ''M'' represents a match, a red ''X'' represents a mismatch, and a black ''n'' represents a position not taken into account during the evaluation algorithm. Note that in the E-CRISP evaluation interface, the sequence of the putative off-target and the sequence representing the position of the mismatches are not aligned with respect to each other. In this example, the correct target (the SCH9 gene) is shown, as well as 7 putative off-targets. Note that in the evaluation page more putative off-targets appear compared to the design page, since also the NAG sequence is permitted as PAM. Each putative offtarget should be manually evaluated, taking into account the number and position of the mismatches with respect to the target sequence (see step 3 in section choose the sgRNA sequence for details).
(D) Three examples of manual evaluation of putative off-targets. The PAM is highlighted in gray, while the mismatches are highlighted in red. In Example 1, two mismatches are located in close proximity to the PAM, and 5 other mismatches are located distal to the PAM: this is not an off-target, and can be ignored. In Example 2, one mismatch is located within the 10 PAM-proximal nucleotides, and three other mismatches are located distal to the PAM: this is also not an off-target, and can be ignored. In Example 3, one mismatch is located at the 10 th position distal to the PAM, and one other mismatch is located further distal to the PAM: since this could be an off-target in vivo, it is recommended to discard the corresponding sgRNA.
Note: The recommended parameters in the ''Off-target analysis'' section are rather loose, allowing retrieval of sgRNAs with several potential off-targets. This is not a problem, because the user will evaluate the potential off-targets later, based on the position of the mismatches with respect to the PAM sequence (see step 3 of this section), which is not taken into account by the E-CRISP algorithm. Conversely, if the chosen parameters are too stringent here, the user will miss a lot of potential good sgRNA sequences.
Optional: If you want to consider potential off-targets in the plasmid expressing Cas9 and the sgRNA, or other exogenous sequences already integrated in the target genome (i.e., fluorescent tags, markers, etc.), you can make use of the ''Select to check for secondary off-targets'' option.
h. In the ''Output'' section adjust the parameters as follows.
i. Maximum number of results per exon: select a high number (>50), otherwise some of the potential sgRNA sequences will not be displayed among the results. ii. Set the other parameters as you like; that will only influence the visualization of the results. j. An example of the output is displayed in Figure 2A. 2. Select a few candidate sgRNA sequences based on the following parameters: a. The position of the target sequence, graphically visualized at the bottom of the output page. i. For gene knockout the position is not relevant: the sgRNA can be located anywhere within the coding sequence. ii. For N-terminal tagging the sgRNA should be located across the ATG starting codon (optimal situation) or within a 120-nucleotides window centered on the ATG starting codon. iii. For C-terminal tagging the sgRNA should be located across the STOP codon (optimal situation) or within a 120-nucleotides window centered on the STOP codon. iv. For integration of an exogenous sequence at a desired genomic site, the sgRNA should be located within a 120-nucleotides window centered on the insertion site. v. For replacement of a genomic sequence, the sgRNA can be located anywhere within the sequence to be replaced (optimal), or within 60 bp outside of this sequence (upstream or downstream). vi. To introduce a point mutation at a desired genomic sequence, the sgRNA should be located across the mutation site (optimal situation) or within a 120-nucleotides window centered on the mutation site. b. The predicted cleavage efficiency, based on the Efficacy score (E-score) ( Figure 2B): the higher the E-score, the higher the chance that the candidate sgRNA will work.
Note: The Efficacy score (E-score) predicts Cas9-dependent cleavage efficiency for the given target sequence. However, there is no guarantee that a specific sgRNA works until direct in vivo validation. The Specificity score (S-score), calculated on the basis of the number of predicted off-targets should be ignored at this stage, because the potential off-targets will be later evaluated by the user. The Annotation score (A-score) can be ignored, since it is optimized on the human genome.
Note: In the output window, the candidate sgRNAs are ranked according to the S-score, so do not hesitate to go down the list in search for good E-scores, even among sgRNA with several putative off-targets.
3. Evaluate the selected sgRNA sequences in the E-CRISP ''Evaluation'' section: a. Choose the following parameters in the ''Select organism'' section. i. Saccharomyces cerevisiae R64-1-1. ii. Number of 5 0 mismatch positions ignored by the program: 6.
iii. Tolerated edit distance to the target sequence: 2. b. In the ''Enter target sequence'' section, paste the 20-nucleotide target sequence (without the PAM) in FASTA format. c. Press the ''Start'' button.
Note: The parameters chosen for the sgRNA evaluation are still very relaxed. With these settings the user will visualize all the putative ''off-targets'' with a more permissive PAM (NAG or NGG, as discussed above), with less than 3 mismatches in the 14 PAM-proximal bases and allowing any mismatch in the 6 PAM-distal bases ( Figure 2C). d. Evaluate each of the putative ''off-targets'', taking into account the number and the position of the mismatches with respect to the target sequence ( Figure 2D), based on the following empirical rules: i. More than 3 mismatches in the target sequence abrogate cleavage by Cas9, based on studies on mammalian cells (Hsu et al., 2013). ii. 1 mismatch in the 10 PAM-proximal nucleotides is sufficient to strongly reduce cleavage by Cas9, based on yeast in vivo studies (Fu et al., 2016). iii. In summary, as a safe rule for evaluation of putative ''off-targets'', if one potential ''offtarget'' sequence has at least one mismatch in the 10 PAM-proximal nucleotides, and at least two extra mismatches in the whole 20-nucleotide sequence, you can be relatively certain that it will not be recognized by the selected sgRNA (i.e., it is not a real ''off-target'').
CRITICAL: While evaluating the total number of mismatches, you should also consider the sequence of the 6 most PAM-distal nucleotides.
Note: By evaluating the putative ''off-targets'' as suggested (i.e., taking into account also the position of the mismatches with respect to the PAM), many of the putative ''off-targets'' can be neglected.
e. Discard the sgRNA sequences that have off-targets that do not pass the safe rule for evaluation described above. 4. Select the two best sgRNA sequences (in terms of predicted efficiency and position relative to the editing site).
Note: It is advised to design and clone in parallel multiple (at least 2) sgRNAs, since there is no guarantee that they will actually work in vivo. If several sgRNAs are cloned in parallel, in case the first tested sgRNA does not work, a second one will be immediately available for yeast transformation.
Note: If you still have doubts that a putative ''off-target'' might be cleaved in vivo, you can sequence that specific genomic locus after performing the CRISPR genome editing, to verify that no unintended mutations have been introduced.
Design oligos for sgRNA cloning Timing: 5 min ll

OPEN ACCESS
Design partially overlapping complementary oligos that, after annealing, leave sticky ends compatible with BsmBI-digested pYTK050 (sgRNA dropout vector).
In silico Golden Gate assembly (optional) Timing: 1-3 h To generate maps of all plasmids obtained during the sgRNA and Cas9 cloning, and verify that the assemblies are done properly, it is recommended to perform all Golden Gate assemblies (see stepby-step method details) in silico beforehand, using the Benchling online tool (https://benchling. com/editor).
Note: To use Benchling, a free account is needed.
Note: An introductory tutorial for in silico Golden Gate Assembly with Benchling is available at the following link: https://help.benchling.com/en/articles/671283-clone-using-the-goldengate-assembly-wizard.
6. The partially overlapping complementary oligos for sgRNA cloning are designed such that, after annealing, the resulting fragment harbors sticky ends compatible with BsmBI-digested pYTK050. However, the inputs for the in silico Golden Gate assembly are either circular plasmids or linear fragments with blunt ends, carrying the BsaI or BsmBI restriction sites. It is therefore necessary to create a Benchling-compatible version of the sgRNA by adding BsmBI sites at both ends, so that the result of in silico BsmBI digestion is identical to the fragment produced by annealing the partially overlapping complementary oligos designed in step 5 of the previous section. a. Take the sequence of the forward oligo for sgRNA cloning (see section design oligos for sgRNA cloning). b. Add the sequence CGTCTCA (containing the BsmBI restriction site) upstream the oligo sequence. c. Add the sequence GTTTTGAGACG (containing the BsmBI restriction site) downstream the oligo sequence. d. Paste the obtained sequence in a new file in SnapGene Viewer (or your favorite DNA sequence editing program) and save it in the desired folder.
Example: Forward oligo: 5 0 -GACTTTGGCCTAAGAACATATGGTCG-3 0 Extended sequence for Benchling (added sequences in Italics): Double-stranded sequence for Benchling (BsmBI sites are highlighted in yellow): Note: If you need to perform several assemblies with many parts in common, you can chose the option ''Assembly'' / ''Re-open'' from the top right corner, and simply switch the parts you want to change.

Preparation of E. coli competent cells
Timing: 2-3 days 15. Prepare and freeze down at À80 C DH5a E. coli competent cells using one of these recommended methods: a. Inoue method (Green and Sambrook, 2020). b. Calcium chloride method (Sambrook and Russell, 2006).

Note:
The Inoue method is more laborious, but yields competent cells with a higher transformation efficiency.

Plasmid preparation
Timing: 1 h 16. Grow an overnight culture (12-20 h at 37 C) of E. coli strains pYTK050, pYTK003, pYTK068, pYTK095, pYTK-DN1, pYTK-DN2, pYTK-DN4, pYTK-DN5, pYTK-DN6 (and, if needed, also pYTK004, pYTK072, pYTK-DN3) in 2 mL LB medium containing the appropriate antibiotic for plasmid selection. 17. Extract the plasmid with a Miniprep kit according to the manufacturer's instructions. 18. Measure plasmid concentration with the NanoDrop. 19. Store the plasmids at À20 C. Adjust ddH 2 O before autoclaving. Add (filter-sterilized) glucose and antibiotics when the medium has cooled down to $65 C. YPD agar plates (with or without antibiotics) can be stored at 4 C for (at least) 3 months, provided that they are placed in a plastic bag to prevent them from drying out. Autoclave or filter-sterilize. If the medium is autoclaved, add antibiotics when the medium has cooled down to $65 C. It is also possible to add antibiotics just before use. LB liquid medium without antibiotics can be stored at room temperature (20 C-25 C) for (at least) 6 months. LB liquid medium with antibiotics can be stored at 4 C for (at least) 3 months. Alternatives: Instead of preparing E. coli competent cells, it is also possible to buy competent cells, for instance NEB 5-alpha Competent E. coli (High Efficiency) (New England Biolabs, C2987).

Alternatives:
We have optimized this protocol using the S288C-derived prototrophic YSBN6 yeast genetic background (Canelas et al., 2010), but any S. cerevisiae laboratory strain can be used.
Alternatives: Instead of the NanoDrop, any other equivalent spectrophotometer can be used for nucleic acid quantification.
Alternatives: Instead of E-CRISP, the CRISPOR online tool (http://crispor.org) could be used for sgRNA design and off-target evaluation, with the limitation that input sequences longer than 2,300 nucleotides are not allowed.
Alternatives: For Golden Gate Assembly and PCR, any thermocycler can be used.

STEP-BY-STEP METHOD DETAILS
In vitro oligos annealing for sgRNA cloning Timing: 1 h The partially overlapping complementary oligos are annealed in vitro to obtain a double-stranded Golden Gate-compatible fragment (ds-sgRNA) for cloning into the sgRNA dropout vector (pYTK050).
1. Prepare a 100 mM dilution of the oligos in ddH 2 O. 2. Mix 10 mL oligo forward and 10 mL oligo reverse in a PCR tube. 3. Anneal the oligos in a thermocycler using the following program: LB agar plates (with antibiotics)

Reagent Final concentration Amount
Yeast extract The sgRNA is cloned in the sgRNA dropout vector (pYTK050) to obtain the sgRNA part plasmid ( Figure 1).

CRITICAL:
The vectors used in all three Golden Gate assembly steps (pYTK050, pYTK095, pYTK-DN4, pYTK-DN5 and pYTK-DN6) contain a GFP gene driven by a promoter for expression in E. coli. This allows visual detection of E. coli colonies transformed with the undigested vector, since these colonies appear green on the plate. Conversely, correct Golden Gate assembly results in loss of the GFP gene, yielding white E. coli colonies ( Figure 3A). 4. Set up the Golden Gate reaction. a. Prepare a 1:500 dilution of the ds-sgRNA fragment obtained in step 3. b. Prepare the GG1 Golden mix in a PCR tube: CRITICAL: T7 ligase enzyme is used in combination with T4 ligase buffer, it is not a typo.

Perform the Golden Gate reaction in a thermocycler:
Pause point: Golden Gate products can be stored at 4 C for a few days.
6. Transform 3 mL of the Golden Gate Assembly product (sgRNA part) in 25 mL Inoue competent E. coli cells (or 5 mL sgRNA part in 50 mL CaCl 2 competent cells). a. Thaw competent cells on ice. b. Add 3-5 mL of the Golden Gate Assembly product obtained in step 5 (sgRNA part) to the competent cells. Mix gently by pipetting up and down or by flicking the tube 4-5 times. Do not vortex. c. Place the mixture on ice for 30 min. d. Heat shock at 42 C for 30 s. e. Place tubes in ice for 2 min. f. Add 950 mL of room-temperature (20 C-25 C) LB liquid medium to the tube. g. Incubate the tube at 37 C for 60 min, shaking vigorously (250 rpm). h. Plate 100 mL of the cell suspension on a LB + chloramphenicol plate. i. Incubate the plate overnight (12-20 h) at 37 C.
Note: On the transformation plate a mixture of white and green colonies will grow. Green colonies contain the undigested pYTK050 (expressing the GFP gene), while white colonies contain a plasmid that has lost the GFP gene as a consequence of the Golden Gate Assembly (Figures 1 and 3A). Optional: Test 1 or 2 plasmids by restriction digestion. If the 20-nucleotide target sequence contains a restriction site, digest with the corresponding enzyme, otherwise you can digest with PvuII (cuts once in the CamR gene).
Note: The Golden Gate reaction works virtually always, so it is not needed to test each plasmid by restriction digestion. We suggest to test only the plasmid obtained through the last Golden Gate Assembly reaction (steps 12-16). Keep all the E. coli plates at 4 C until the end of the whole cloning procedure, so that, if accidentally something went wrong, you can go back to the appropriate step and use another colony to prepare plasmid DNA.
Pause point: The sgRNA part plasmid can be stored at À20 C for several months.

Timing: 2.5 days
The sgRNA is assembled in a transcriptional unit, to obtain the sgRNA cassette plasmid. Based on the connectors used, the sgRNA can be assembled as a TU2 or TU3 transcriptional unit ( Figure 1). b. For cloning the TU3 sgRNA cassette plasmid, prepare the GG2 Golden mix in a PCR tube as follows: CRITICAL: T7 ligase enzyme is used in combination with T4 ligase buffer, it is not a typo.
9. Perform the Golden Gate reaction in the thermocycler, using the same cycling conditions described in step 5.
Pause point: Golden Gate products can be stored at 4 C for a few days.
10. Transform 3 mL of the Golden Gate Assembly product (sgRNA cassette plasmid) in 25 mL Inoue competent E. coli cells (or 5 mL sgRNA part in 50 mL CaCl 2 competent cells) as described in step 6, with the following difference: a. Plate 100 mL of the cell suspension on a LB + ampicillin plate.
Note: On the transformation plate a mixture of white and green colonies will grow. Green colonies contain the undigested pYTK095 (expressing the GFP gene), while white colonies contain a plasmid that has lost the GFP gene as a consequence of the Golden Gate Assembly (Figures 1 and 3A). c. For cloning the Cas9+sgRNA1(TU2)+sgRNA2(TU3) into the yeast expression vector, prepare the GG3 Golden mix in a PCR tube as follows: CRITICAL: T7 ligase enzyme is used in combination with T4 ligase buffer, it is not a typo! Note: You can use the Excel sheets ''GG3 Cas9-sgRNA1-spacer'', ''GG3 Cas9-spacer-sgRNA2'' and ''GG3 Cas9-sgRNA1-sgRNA2'' from the ''CRISPR-Cas9 toolbox MoClo Assembly spreadsheet'' to calculate the mL plasmid DNA for assembly of Cas9 and the gRNA cassette(s) into the yeast expression vector.
13. Perform the Golden Gate reaction in the thermocycler, using the same cycling conditions described in step 5.
Pause point: Golden Gate products can be stored at 4 C for a few days.
14. Transform 3 mL of the Golden Gate Assembly product (Cas9+sgRNA(s) multi-gene plasmid) in 25 mL Inoue competent E. coli cells (or 5 mL sgRNA part in 50 mL CaCl 2 competent cells) as described in step 6, with the following difference: a. Plate 100 mL of the cell suspension on a LB + kanamycin plate.
Note: On the transformation plate a mixture of white and green colonies will grow. Green colonies contain the undigested yeast vector (expressing the GFP gene), while white colonies contain a plasmid that has lost the GFP gene as a consequence of the Golden Gate Assembly (Figures 1 and 3A).  Figure 3B), which cuts 2 times in the correctly assembled Cas9+sgRNA1(TU2) and Cas9+sgRNA2(TU3) multi-gene plasmids, and 3 times in the correctly assembled Cas9+sgRNA1(TU2)+sgRNA2(TU3) multi-gene plasmid (unless a third or fourth restriction site, respectively, is present in the 20-nucleotides target sequence). Optional: Verify the sgRNA insert by sequencing with primer sgRNA-seq-fwd.
Pause point: The Cas9+sgRNA(s) multi-gene plasmid can be stored at À20 C for several months.
Note: The following sections (steps 17-31) contain detailed instructions for the design and production of the repair fragment for different genome editing applications (i.e., gene knockout, N-and C-terminal tagging, sequence integration at a specific genomic location, sequence replacement, and introduction of point mutations). Please choose the appropriate section based on the desired genome editing application, then proceed with yeast transformation (step 32).
Note: In the following sections, we refer to the genomic sequence that is modified as ''genome editing site'' (can span from a single nucleotide to several kilobases, depending on the application), and to the genomic sequence bound and cleaved by Cas9 as ''target sequence''. For some applications these two sequences (partially) overlap, as in the case of gene deletion or sequence integration. In other instances, they can be located several nucleotides apart, which can happen sometimes with point mutations or N-and C-terminal tagging.
Design and production of the repair fragment for gene deletion

Timing: 2 h
The repair fragment for gene deletion is made up of the 60 nucleotides upstream the ATG start codon, followed by the 60 nucleotides downstream the STOP codon. The repair fragment is obtained by PCR amplification of partially overlapping primers (''no template PCR'', Figure 4).
Note: Studies in yeast suggest that it might be possible to use a single-stranded oligonucleotide instead of a double-stranded DNA fragment as a repair template for gene deletion (Dicarlo et al., 2013;Storici et al., 2003), even though we have not tested it in our hands.  Optional: Load 5 mL PCR product on agarose gel and verify PCR via DNA electrophoresis (expected band size: 120 bp).

Proceed with yeast transformation (step 32).
Design and production of the repair fragment for N-terminal tagging Timing: 2-5 h The repair fragment for N-terminal tagging of a gene is made up of the 57 nucleotides upstream the ATG start codon, followed by the tag, followed by the 60 nucleotides downstream the ATG start codon. The repair fragment is obtained by standard PCR amplification ( Figure 5). If the sgRNA does not span through the ATG start codon, it is also necessary to introduce one or more (synonymous) mutations in the sgRNA sequence, to prevent re-cutting by Cas9 after repair, as described in detail below.
CRITICAL: As the wide majority of tags start with an ATG codon, this design yields a repair fragment with 60 nucleotides of homology at both ends ( Figure 5). However, if the tag does not start with an ATG codon, it is necessary to add the ATG codon to the upstream homology region (step 20c).
20. In case the sgRNA sequence spans through the ATG start codon, design the primers for the repair fragment as follows. a. Design the annealing sequence for the forward primer on the top strand of the template for tag amplification, according to the following rules: i. Start at the beginning of the tag sequence (from the ATG) and extend the sequence in 5 0 to 3 0 direction. ii. The annealing sequence should be at least 20 nucleotides long. iii. If possible, the CG content should be between 35% and 65%. iv. The last nucleotide should be G or C, and the last 5 nucleotides should enclose at least 2 G/C. v. The Tm of the annealing sequence should be compatible with the Tm of the annealing sequence for the reverse primer (no more than 5 C difference). b. Design the annealing sequence for the reverse primer on the top strand of the template for tag amplification, according to the following rules: i. Start at the end of the tag sequence (excluding the STOP codon, in case it is present) and extend the sequence in 3 0 to 5 0 direction. ii. The annealing sequence should be at least 20 nucleotides long. iii. If possible, the CG content should be between 35% and 65%. iv. The last nucleotide should be G or C, and the last 5 nucleotides should enclose at least 2 G/C. v. The Tm of the annealing sequence should be compatible with the Tm of the annealing sequence for the forward primer (no more than 5 C difference). c. Design the homology regions for genomic integration.
i. Upstream homology region: 57 nucleotides upstream the ATG starting codon.
ii. Downstream homology region: 60 nucleotides downstream the ATG starting codon. d. Combine the annealing regions and the homology regions to create the primers.
i. Forward primer: upstream homology region + annealing sequence for the forward primer.
ii. Reverse primer: reverse complement downstream homology region + reverse complement annealing sequence for the reverse primer. e. Amplify the repair fragment through standard PCR.
Note: If the sgRNA sequence spans through the ATG start codon, introduction of an N-terminal sequence will automatically disrupt the sgRNA sequence. 21. In case the sgRNA sequence does not span through the ATG start codon, design the primers for the repair fragment as follows. a. Design the primers as indicated in step 20. b. Introduce point mutations in the upstream homology region or in the downstream homology region of the primers, in order to disrupt the sgRNA sequence, but still preserving the amino acid sequence (if the sgRNA encompasses the coding region). This can be done in two ways: i. Mutate one of the ''Gs'' of the PAM site (NGG) so that the target sequence is no longer recognized (but avoid NAG, because it can still be recognized as a PAM). ii. Mutate at least 2 bases among the 10 PAM-proximal bases (this is enough to prevent binding of the sgRNA). c. Amplify the repair fragment through standard PCR. 22. Proceed with yeast transformation (step 32).

Design and production of the repair fragment for C-terminal tagging
Timing: 2-5 h The repair fragment for C-terminal tagging of a gene is made up of the 60 nucleotides upstream the STOP codon, followed by the tag, followed by the STOP codon, followed by the 57 nucleotides downstream the STOP codon. The repair fragment is obtained by standard PCR amplification (Figure 6). If the sgRNA does not span through the STOP codon, it is also necessary to introduce one or Figure 5. Design of repair fragment for N-terminal tagging (A) Schematic of the repair fragment design. The repair fragment is obtained through PCR amplification of the tag (in green) with primers annealing at the beginning (in dark red) and at the end (in orange) of the tag, and carrying tails for the homology regions upstream (in black) and downstream (in blue) the ATG start codon. After Cas9-dependent cleavage near the ATG start codon, repair fragment integration results in introduction of the tag at the N-terminal of the gene. (B) Example of primers design for N-terminal tagging (step 20). The ATG start codon is highlighted in yellow. (C) Introduction of synonymous mutations in the target sequence to prevent Cas9 cleavage after repair (step 21). The sgRNA sequence is highlighted in gray. The possible base changes for synonymous mutations are marked in red. The amino acid sequence is displayed using the one-letter code under the DNA sequence. See main text for details. more (synonymous) mutations in the sgRNA sequence, to prevent re-cutting by Cas9 after repair, as described in detail below.
23. In case the sgRNA sequence spans through the STOP codon, design the primers for the repair fragment as follows. a. Design the annealing sequence for the forward primer on the top strand of the template for tag amplification, according to the following rules: i. Start at the beginning of the tag sequence and extend the sequence in 5 0 to 3 0 direction.
ii. The annealing sequence should be at least 20 nucleotides long.
iii. If possible, the CG content should be between 35% and 65%. iv. The last nucleotide should be G or C, and the last 5 nucleotides should enclose at least 2 G/C. v. The Tm of the annealing sequence should be compatible with the Tm of the annealing sequence for the reverse primer (no more than 5 C difference). b. Design the annealing sequence for the reverse primer on the top strand of the template for tag amplification, according to the following rules: i. Start at the end of the tag sequence and extend the sequence in 3 0 to 5 0 direction.
ii. The annealing sequence should be at least 20 nucleotides long.
iii. If possible, the CG content should be between 35% and 65%. iv. The last nucleotide should be G or C, and the last 5 nucleotides should enclose at least 2 G/C. v. The Tm of the annealing sequence should be compatible with the Tm of the annealing sequence for the forward primer (no more than 5 C difference). c. Design the homology regions for genomic integration.
i. Upstream homology region: 60 nucleotides upstream the STOP codon.
ii. Downstream homology region: STOP codon + 57 nucleotides downstream the STOP codon. d. Combine the annealing regions and the homology regions to create the primers.
i. Forward primer: upstream homology region + annealing sequence for the forward primer.
ii. Reverse primer: reverse complement downstream homology region + reverse complement annealing sequence for the reverse primer. e. Amplify the repair fragment through standard PCR.
Note: If the sgRNA sequence spans through the STOP codon, introduction of a C-terminal sequence will automatically disrupt the sgRNA sequence.
24. In case the sgRNA sequence does not span through the STOP codon, design the primers for the repair fragment as follows. a. Design the primers as indicated in step 23. b. Introduce point mutations in the upstream homology region or in the downstream homology region of the primers, in order to disrupt the sgRNA sequence, but still preserving the amino acid sequence (if the sgRNA encompasses the coding region). This can be done in two ways: i. Mutate one of the ''Gs'' of the PAM site (NGG) so that the target sequence is no longer recognized (but avoid NAG, because it can still be recognized as a PAM). ii. Mutate at least 2 bases among the 10 PAM-proximal bases (this is enough to prevent binding of the sgRNA). c. Amplify the repair fragment through standard PCR. 25. Proceed with yeast transformation (step 32).
Design and production of the repair fragment for sequence integration or sequence replacement Timing: 2-5 h It is possible to insert a donor sequence at any genomic location, with or without concomitant deletion of an endogenous sequence, based on the design of the repair fragment. The repair fragment for integration of the desired donor DNA sequence at a specific genomic position without deleting any endogenous sequence (assuming the sgRNA sequence contains the integration site) is made up of the 60 nucleotides upstream the integration site, followed by the donor sequence, followed by the 60 nucleotides downstream the integration site (Figure 7). The repair fragment for replacement of a specific genomic sequence with the desired donor sequence (assuming the sgRNA sequence is internal to the sequence to be replaced) is made up of the 60 nucleotides upstream the genomic sequence, followed by the donor sequence, followed by the 60 nucleotides downstream the genomic sequence (Figure 8). In both cases, the repair fragment is obtained by standard PCR amplification.
26. For both genome editing applications, design the primers for the repair fragment as follows.
a. Design the annealing sequence for the forward primer on the top strand of the template for amplification of the donor sequence, according to the following rules: i. Start at the beginning of the donor sequence and extend the sequence in 5 0 to 3 0 direction. ii. The annealing sequence should be at least 20 nucleotides long. iii. If possible, the CG content should be between 35% and 65%. iv. The last nucleotide should be G or C, and the last 5 nucleotides should enclose at least 2 G/C. v. The Tm of the annealing sequence should be compatible with the Tm of the annealing sequence for the reverse primer (no more than 5 C difference). b. Design the annealing sequence for the reverse primer on the top strand of the template for amplification of the donor sequence, according to the following rules: i. Start at the end of the donor sequence and extend the sequence in 3 0 to 5 0 direction.
ii. The annealing sequence should be at least 20 nucleotides long.
iii. If possible, the CG content should be between 35% and 65%. iv. The last nucleotide should be G or C, and the last 5 nucleotides should enclose at least 2 G/C. v. The Tm of the annealing sequence should be compatible with the Tm of the annealing sequence for the forward primer (no more than 5 C difference). c. Design the homology regions for genomic integration.
i. Upstream homology region for sequence integration: 60 nucleotides upstream the integration site. ii. Downstream homology region for sequence integration: 60 nucleotides downstream the integration site. iii. Upstream homology region for sequence replacement: 60 nucleotides upstream the sequence to be replaced. iv. Downstream homology region for sequence replacement: 60 nucleotides downstream the sequence to be replaced. d. Combine the annealing regions and the homology regions to create the primers.
i. Forward primer: upstream homology region + annealing sequence for the forward primer.

OPEN ACCESS
ii. Reverse primer: reverse complement downstream homology region + reverse complement annealing sequence for the reverse primer. e. Amplify the repair fragment through standard PCR. Figure 7. Design of repair fragment for sequence integration at a desired genomic location (A) Schematic of the repair fragment design. The integration site is located within the target sequence (marked in red). The repair fragment is obtained through PCR amplification of the donor sequence (in green) with primers annealing at the beginning (in dark red) and at the end (in orange) of the sequence, and carrying tails for the homology regions upstream (in black) and downstream (in blue) the integration site. After Cas9-dependent cleavage near the integration site, repair fragment integration results in the insertion of the donor sequence at the integration site with concomitant disruption of the target sequence. Note: Integration of the donor sequence at the desired genomic location or replacement of the desired genomic sequence with the donor sequence will disrupt the sgRNA sequence.

Proceed with yeast transformation (step 32).
Design and production of the repair fragment for introduction of point mutations Timing: 2-3 h Figure 8. Design of repair fragment for sequence replacement (A) Schematic of the repair fragment design. The repair fragment is obtained through PCR amplification of the donor sequence (''new sequence'', in green) with primers annealing at the beginning (in dark red) and at the end (in orange) of the sequence, and carrying tails for the homology regions upstream (in black) and downstream (in blue) the sequence to be replaced (''old sequence'', in pink). After Cas9-dependent cleavage within the old sequence, repair fragment integration results in the insertion of the new sequence with concomitant deletion of the old sequence. (B) Example of primers design for sequence replacement (step 26). The old sequence is highlighted in pink. See main text for details.

OPEN ACCESS
If the sgRNA sequence includes the desired mutation site, the repair fragment for the introduction of point mutations is made up of the 50 nucleotides upstream the target sequence, followed by the 20-nucleotides target sequence, followed by the 50 nucleotides downstream the target sequence. The fragment carries the desired mutation either inside the target sequence or in the 50 nucleotides upstream or downstream. If needed, extra (synonymous) mutations are introduced to disrupt the sgRNA sequence. In case the sgRNA sequence does not include the desired mutation site, the repair fragment is designed similarly, but it is shifted a few nucleotides upstream or downstream, so that the mutation site and the target sequence are more or less equally distant from the edges of the repair fragment. In both instances, repair fragment is obtained by PCR amplification of partially overlapping primers (''no template PCR'', Figure 9).
28. In case the sgRNA sequence spans through the desired mutation site, design partially overlapping primers as follows. a. Forward primer: 50 nucleotides upstream target sequence + 20 nucleotides mutated target sequence. b. Reverse primer: reverse complement 50 nucleotides downstream target sequence + reverse complement 20 nucleotides mutated target sequence. c. The mutated sequence contains: i. The desired point mutation(s).
ii. If necessary, one or multiple additional synonymous mutations in order to disrupt the sgRNA sequence, but still preserving the amino acid sequence. d. Introduction of additional synonymous mutations that disrupt the sgRNA sequence can be done in two ways: i. Mutate one of the ''Gs'' of the PAM site (NGG) so that the target sequence is no longer recognized (but avoid NAG, because it can still be recognized as a PAM). ii. Mutate at least 2 bases among the 10 PAM-proximal bases (this is enough to prevent binding of the sgRNA).
Note: With such a design, the 20 3 0 -proximal nucleotides of the two primers anneal with each other, and each can provide the template for the extension of the other primer, yielding a complete 120 bp repair fragment.
29. In case the sgRNA sequence does not include the desired mutation site, design partially overlapping primers as follows. Figure 9. Design of repair fragment for the introduction of point mutations (A) Schematic of the repair fragment design when the desired mutation is located within the sgRNA sequence. The repair fragment is obtained through ''no template PCR'' amplification of partially overlapping primers containing the mutated target sequence and the homology regions upstream (in black) and downstream (in blue) the target sequence. The mutation is indicated by a yellow star. After Cas9-dependent cleavage near the mutation site, repair fragment integration results in the introduction of the desired mutation and the concomitant disruption of the target sequence.
(B) Schematic of the repair fragment design when the desired mutation is located outside the sgRNA sequence. The repair fragment is obtained through ''no template PCR'' amplification of partially overlapping primers containing (part of) the target sequence and the homology regions upstream (in black) and downstream (in blue) the target sequence. The desired mutation (indicated by a yellow star) is introduced in the upstream or downstream homology region. Moreover, one or more additional synonymous mutations (indicated by a white star) are introduced in the target sequence. After Cas9-dependent cleavage near the mutation site, repair fragment integration results in the introduction of the desired mutation and the disruption of the target sequence.
(C) Example of primers design for the introduction of point mutations, when the desired mutation is located within the target sequence (step 28). The sgRNA sequence is highlighted in gray. The mutated bases (ATG>GGT) and the corresponding amino acid change (M>G) are marked in red. The amino acid sequence is displayed using the oneletter code under the DNA sequence. Since the mutated bases are distal to the PAM, it is possible, out of precaution, to introduce an additional synonymous mutation (in magenta) in the target sequence near the PAM to ensure abrogation of Cas9 cleavage after repair. See main text for details.

OPEN ACCESS
a. Select a 120-nucleotide sequence centered on the target sequence and the desired mutation site. b. Obtain the sequence of the repair fragment by adding the desired point mutation(s) and one or multiple additional synonymous mutations in order to disrupt the sgRNA sequence: i. Mutate one of the ''Gs'' of the PAM site (NGG) so that the target sequence is no longer recognized (but avoid NAG, because it can still be recognized as a PAM). ii. Alternatively, mutate at least 2 bases among the 10 PAM-proximal bases (this is enough to prevent binding of the sgRNA). c. Forward primer: the first 70 nucleotides of the repair fragment. d. Reverse primer: reverse complement of the last 70 nucleotides of the repair fragment.
Note: With such a design, the 20 3 0 -proximal nucleotides of the two primers anneal with each other, and each can provide the template for the extension of the other primer, yielding a complete 120 bp repair fragment.
a. Prepare the PCR mix: b. Divide the PCR mix in 4 PCR tubes (50 mL per tube). c. Perform the PCR reaction in the thermocycler with the following program: Note: The PCR program performs 20 cycles with an annealing temperature corresponding to the overlapping region (20 nucleotides), and then another 10 cycles with an annealing temperature = extension temperature (=72 C), corresponding to annealing of the whole primers (70 nucleotides).
d. Pool the 4 PCR reactions together.
Optional: Load 5 mL PCR product on agarose gel and verify PCR via DNA electrophoresis (expected band size: 120 bp).

Yeast transformation
Timing: 6-7 days The desired yeast strain is co-transformed with the Cas9+sgRNA(s) multi-gene plasmid and the appropriate repair fragment. As a negative control, the same transformation without the repair fragment is performed ( Figure 10A).

Note:
The transformation protocol is adapted from the lithium acetate (LiAc)/SS carrier DNA/ PEG method (Gietz and Schiestl, 2007).
32. Purify the repair fragment(s) with a PCR purification kit and measure DNA concentration with the NanoDrop (expected concentration: 50-250 ng/mL).
Pause point: Purified repair fragments can be stored at À20 C for several months.
33. Transform the desired yeast strain(s) using the LiAc method. CRITICAL: Pipette PEG slowly and carefully, since it is a very viscous solution.
Note: If you want to perform simultaneous genome editing at two genomic loci with two sgRNAs and two repair fragments, 3 mg of each repair fragment should be added to the CRISPR mix. Furthermore, it is recommended to simultaneously perform also single locus genome editing (using each sgRNA and the corresponding repair fragment separately) as a control, to make sure that both sgRNAs work. Troubleshooting 6.
k. Resuspend by vortexing and, if needed, pipetting up and down. l. Incubate at 42 C for 40 min (in a water bath). m. Pellet cells at 5,000 g for 2 min at RT (20 C-25 C) and carefully remove the TMIX (by pipetting). n. Resuspend cells in 1 mL YPD liquid medium and incubate 2-3 h at 30 C to allow expression of the resistance marker. o. Pellet cells at 5,000 g for 2 min at RT (20 C-25 C), resuspend in 150 mL sterile ddH 2 O and plate on the appropriate selective plates: i. YPD + G418 plates if the plasmid contains the KanR marker.
ii. YPD + clonNAT plates if the plasmid contains the NatR marker.
iii. YPD + HygB plates if the plasmid contains the HygR marker. p. Incubate plates at 30 C for 2-3 days.
34. Check the transformation plates to verify the proper functioning of the sgRNA(s): the plate with the repair fragment should have at least 10-fold more colonies than the plate without repair fragment ( Figure 10A, see also expected outcomes and troubleshooting sections). 35. Choose 8-10 colonies from the transformation plate (with repair fragment!) and streak for single colony on selective plates (using the same selection drug used for the transformation).
CRITICAL: Streaking for single colonies is essential because the CRISPR genome editing event happens during the first cell divisions on plate, therefore colonies on the transformation plate might be heterogeneous.
CRITICAL: it is recommended to still keep the plasmid selection while streaking for single colony, to avoid false positives that lost the Cas9-gRNA plasmid but somehow are still alive on the transformation plate.
36. Make patches for temporary storage on YPD plates. a. For every re-streaked independent transformant, chose one individual colony and make a small patch on a YPD plate. b. If no single colony can be isolated for a particular transformant, that transformant should be left aside.
Note: From this moment on, the selection for the plasmid marker is not needed anymore, and it is recommended to remove the selection to facilitate the loss of the plasmid (see removal of the Cas9+sgRNA(s) multi-gene plasmid). If your strain contains other antibiotic-resistance markers, you can decide to add one antibiotic to the YPD plates to minimize contamination.
37. Incubate the plate(s) at 30 C for 1 day.
Optional: To save one day, you can use a small portion of the colony directly for genomic DNA extraction for colony PCR (see verification of genome editing), and patch the rest of the colony for temporary storage.
Pause point: plates can be stored at 4 C for a several days.

Verification of genome editing
Timing: 2-8 h Genomic DNA (gDNA) is prepared from the single colonies, and the desired genome editing event is verified by PCR and/or sequencing ( Figure 10B).  Note: It is possible to perform another round of CRISPR genome editing before plasmid removal, provided the second Cas9+sgRNA(s) multi-gene plasmid has a different selectable marker. The first plasmid will be likely lost in the process. At the end of the second round, loss of both plasmids has to be verified.

OPEN ACCESS
Removal of the Cas9+sgRNA(s) multigene plasmid

Timing: 5 days
Yeast cells lose the low-copy Cas9+sgRNA(s) multi-gene plasmid easily in the absence of selection. Verified strains are grown on non-selective medium and loss of the Cas9+sgRNA(s) multi-gene plasmid is verified via inability to grow on selective plates (Figure 11).
42. Inoculate a tiny amount of cells from the temporary storage patches of verified transformants in 3 mL YPD medium. 43. Grow an overnight (12-20 h) culture at 30 C, shaking at 300 rpm. 44. Plate the culture on (non-selective) YPD plates.
a. Dilute the cultures in order to achieve $100 colonies on plate. i. Prepare a first dilution in a sterile 2 mL microcentrifuge tube: 70 mL cells + 930 mL ddH 2 O (sterile). ii. From the first dilution, perform two 1:100 serial dilutions in sterile 2 mL microcentrifuge tubes: 10 mL cells + 990 mL ddH 2 O (sterile). b. Plate 100 mL of the last dilution on YPD plates.
Note: To minimize plate contamination, you can add the same antibiotic you added to the plates for temporary storage patches (step 36).
Note: If the transformed yeast strain has growth defects, it may be necessary to adjust the growth time of the liquid culture, and/or the serial dilutions factor, in order to achieve $100 colonies on plate.
c. Incubate plates at 30 C for 2 days. 45. Make new temporary storage patches and test plasmid loss.
a. Patch 4 colonies per strain on YPD plates for temporary storage (it is again possible to add the antibiotic you added to the plates for temporary storage patches in step 36). b. At the same time (using the same inoculation loop or flat toothpick you used to make the patch), streak cells onto a selective plate for the Cas9+sgRNA(s) multi-gene plasmid (Figure 11C).
Note: Include a positive and a negative control on selective plates. c. Incubate plates at 30 C for 1 day. 46. Verify plasmid loss: strains that have lost the plasmid do not grow on selective plates. 47. For each independent transformant (selected in step 41), freeze down (from the last temporary storage patches made in step 45) one colony that has lost the Cas9+sgRNA(s) multi-gene plasmid.

EXPECTED OUTCOMES
This protocol consists of two main parts: i) cloning of Cas9 and the sgRNA(s) in a yeast expression vector via three subsequence Golden Gate assembly reactions and ii) yeast co-transformation with the Cas9+sgRNA(s) multi-gene plasmid and the appropriate repair fragment for genome editing.
An example of plates with E. coli colonies resulting from the three Golden Gate assembly reactions is shown in Figure 3A. The number of colonies on plate as well as the white/green colonies ratio may vary due to the transformation efficiency of the competent cells used and the efficiency of the different Golden Gate assembly reactions. However, in our hands the cloning procedure works very consistently and we were always able to isolate a correct Cas9+sgRNA(s) multi-gene plasmid at the end of the cloning workflow.
CRISPR/Cas9 technology allows precise, marker-free genome editing, provided that a suitable sgRNA and the proper repair fragment are employed. Figure 10 displays examples of successful yeast genome editing after transformation. The control transformation without the repair fragment is essential to assess the Cas9 cleavage efficiency associated with a specific sgRNA. An unrepaired double-strand break (DSB) in yeast is lethal. DSBs in yeast are efficiently repaired via homologous recombination (HR), provided a suitable homology template is available. In the absence of a donor template, repair can occur via the less efficient non-homologous end joining (NHEJ) pathway. If perfect NHEJ occurs, the repaired sequence remains identical (and can be cleaved again by Cas9), while imperfect NHEJ results in the presence of mutations in the proximity of the cleavage site. If the sgRNA works, very few colonies are expected in the control plate (resulting from imperfect NHEJ repair with mutation of the sgRNA sequence), while many more colonies (more than 10-fold) should appear in the plate with the repair fragment (since HR is way more efficient). Conversely, if both plates display a similarly high number of colonies, this is an indication that the selected sgRNA probably does not work well in vivo (see troubleshooting section).
If the sgRNA works, usually it is enough to test $10 colonies to identify at least one transformant bearing the desired genome editing event. The efficiency of genome editing depends on several factors: 1) the size of the deleted fragment: increased length = decreased efficiency; 2) the length of the homology regions at the ends of the repair fragment: increased length (up to 500 bp) = increased efficiency; 3) the distance between the genome editing site and the target sequence: increased distance = decreased efficiency.
Please consult the troubleshooting section for suggestions on how to improve genome editing efficiency.

LIMITATIONS
The main limitation of CRISPR/Cas9 genome editing technology is the need for a suitable sgRNA sequence in the proximity of the editing site. If no suitable sgRNA is available, a different genome editing technique should be used, e.g., the delitto perfetto (Stuckey and Storici, 2013).
Secondly, the pre-assembled vectors we provide are suitable for cloning up to 2 sgRNAs together with Cas9, and contain only drug resistance yeast markers (KanR, NatR, HygR). If cloning of more than 2 sgRNAs in one vector and/or auxotrophic yeast marker are preferred, users should assemble the yeast expression vector(s) themselves (the MoClo-Yeast Toolkit design allows cloning of up to 4 sgRNAs in a single plasmid, and provides three auxotrophic markers) (Lee et al., 2015).

OPEN ACCESS
Another limiting factor of CRISPR/Cas9 genome editing is time. The whole procedure (from design of the sgRNA till genome editing verification and plasmid removal) takes at least three weeks. More specifically, the sgRNA cloning part can take a whole week. However, since no PCR amplification steps are required in the sgRNA cloning phase, the chance of introducing undesired mutations during sgRNA cloning is virtually absent. Furthermore, simultaneous expression of the Cas9 protein and the sgRNA from the same (low copy) plasmid circumvents the need for a host yeast strain with Cas9 stably integrated in the genome. All in all, we believe that the robustness and reliability of this protocol compensates well for its length.

TROUBLESHOOTING
Problem 1 It is difficult to distinguish white and green E. coli colonies obtained during sgRNA cloning (steps 7, 11, and 15).

Potential solutions
Incubation of the plates at 4 C for 5-8 h usually improves the detection of white versus green colonies. Alternatively, the difference between white and green can be better visualized by illuminating the plates with a blue or UV lamp.
Problem 2 E. coli transformation with the Golden Gate assembly product yields only green colonies (steps 7, 11, and 15). This is an indication that probably the Golden Gate assembly reaction did not work.

Potential solutions
Double-check the design of the oligos for the sgRNA (step 5 of the before you begin section). Double-check the Golden Gate reaction table (especially the name of the plasmids and the plasmid concentrations). Repeat the Golden Gate reaction and E. coli transformation.

Problem 3
Low yield of repair fragment obtained by ''no template PCR'' (steps 18 and 30).

Potential solutions
Use the product of the ''no template PCR'' as a template for a standard PCR reaction, using two new 20-nucleotide primers annealing at the edges of the repair fragment.

Problem 4
After CRISPR transformation there is no difference in the number of colonies in the presence or absence of the repair fragment (step 34). This is likely due to an insufficient amount of the repair fragment (if the number of colonies on both plates is very low), or to poor Cas9 cleavage because the sgRNA does not work (if many colonies grow on both plates).

Potential solutions
Increase the amount of the repair fragment (5-10 mg). Try a different sgRNA.

Problem 5
The sgRNA works (clear difference À/+ repair fragment), but no clones carrying the desired editing event are detected after diagnostic PCR and/or sequencing (steps 38-40). This is an indication of low CRISPR efficiency.

Potential solutions
To increase the efficiency of CRISPR genome editing, several strategies can be tried: Verify if there is any homologous sequence in the yeast genome that might be used as a donor sequence instead of the desired repair fragment. If this is the case, increase the amount of repair fragment (5-10 mg) and test many transformants (>20). Increase the length of the homology regions at the edges of the repair fragment to improve recombination efficiency (80-100 nucleotides).
In case of gene knockout (or, more in general, deletion of any sequence), genome editing efficiency is inversely related to the length of the sequence to be deleted. To improve efficiency, increase the amount of repair fragment (5-10 mg) and test many transformants (>20). In case of large deletions (>5 Kb), to improve efficiency we suggest the simultaneous use of two sgRNAs, guiding Cas9 to cut near both ends of the sequence to be deleted, to facilitate loss of the intervening sequence. Also increasing the length of the repair fragment could contribute to improve efficiency. In case of point mutations, the distance between the target sequence and the desired mutation site affects the editing efficiency (increased distance = decreased efficiency). To improve efficiency, increase the amount of repair fragment (5-10 mg) and test many transformants (>20).

Problem 6
Simultaneous genome editing with two (or more) sgRNAs does not work (no double-edited transformants are retrieved).

Potential solutions
If the sgRNAs have been individually validated, two strategies are possible: Increase the amount of repair fragment the amount of transformants tested. Perform the two (or more) genome editing events sequentially in two independent transformations. If the Cas9+sgRNA multi-gene plasmids carry the same selection marker, removal of the plasmid is needed before the second transformation, otherwise both can be removed at the end of the editing procedure.
If the sgRNAs have not been individually validated, perform the CRISPR transformations with each sgRNA individually to verify them. The transformants of one of the transformations can be used as a starting strain to introduce the second genome editing event, as suggested above. If one of the sgRNAs does not work, try a different sgRNA.

RESOURCE AVAILABILITY
Lead contact Further information and requests for resources and reagents should be directed to and will be fulfilled by the lead contact, Andreas Milias-Argeitis (a.milias.argeitis@rug.nl).

Data and code availability
This study did not generate any datasets or code.

ll
OPEN ACCESS