搜索

1 user has reported that he/she has successfully carried out the experiment using this protocol.
Single-step Precision Genome Editing in Yeast Using CRISPR-Cas9
使用CRISPR-Cas9在酵母中进行单步精确基因组编辑   

下载 PDF 引用 收藏 提问与回复 分享您的反馈 Cited by

本文章节

参见作者原研究论文

本实验方案简略版
eLIFE
Jun 2017

Abstract

Genome modification in budding yeast has been extremely successful largely due to its highly efficient homology-directed DNA repair machinery. Several methods for modifying the yeast genome have previously been described, many of them involving at least two-steps: insertion of a selectable marker and substitution of that marker for the intended modification. Here, we describe a CRISPR-Cas9 mediated genome editing protocol for modifying any yeast gene of interest (either essential or nonessential) in a single-step transformation without any selectable marker. In this system, the Cas9 nuclease creates a double-stranded break at the locus of choice, which is typically lethal in yeast cells regardless of the essentiality of the targeted locus due to inefficient non-homologous end-joining repair. This lethality results in efficient repair via homologous recombination using a repair template derived from PCR. In cases involving essential genes, the necessity of editing the genomic lesion with a functional allele serves as an additional layer of selection. As a motivating example, we describe the use of this strategy in the replacement of HEM2, an essential yeast gene, with its corresponding human ortholog ALAD.

Keywords: CRISPR (CRISPR), Homologous recombination (同源重组), Humanization (人源化), Ortholog complementation (直系同源互补), Genome editing (基因组编辑), Yeast engineering (酵母工程)

Background

Saccharomyces cerevisiae (Baker’s yeast) has a long history as a genetically tractable organism, and there are an array of methodologies to manipulate the yeast genome. However, until recently it has been necessary to apply selection to isolate clones possessing the desired genetic alteration (Kearse et al., 2012; DiCarlo et al., 2013; Lee et al., 2015; Kachroo et al., 2017). In cases where arbitrary, scar-less editing of the genome is desired, the solution is typically a two-step process: First a selectable cassette (containing the URA3 marker, for example), flanked by homology arms targeting the region of interest, and sometimes containing nuclease targeting sites (i.e., I-SceI sites) to aid in the removal of the cassette at the later stage, is knocked in via homologous recombination (HR). The small subpopulation of successful integrants is isolated by selecting for the cassette. Second, the marker is eliminated through highly efficient sequence specific methods such as site-specific recombination or endonuclease cleavage (I-SceI) to generate the desired form of the edited genomic locus. Two steps are necessary because no method was available which is both scar-less and efficient enough such that no selection is required.

The development of CRISPR/Cas9 technology in yeast has eliminated the need for this two-step process. Cas9 efficiently creates double-stranded breaks (DSBs) in yeast DNA at virtually any arbitrary locus–provided a PAM sequence is proximal to the desired cut site. When an appropriate repair template is provided, these DSBs are repaired through the endogenous HR system of yeast. Cas9 directed to the desired genomic locus via the guide RNA sequence creates double-stranded break (DSB) in the genome. The CRISPR target site is retained in cells which fail to repair the target site as expected, which allows Cas9 to repeatedly cleave the same region until HR-mediated editing takes place. Rarely, non-homologous end-joining (NHEJ) can generate mutations which block Cas9 cleavage despite failing to incorporate the expected genomic alterations. More commonly, cells simply succumb to the stress of repeated Cas9-induced genomic cleavages. In an appropriately conducted experiment, the majority of the surviving population tends to be cells which have lost their CRISPR target site by incorporating the desired genomic alteration via HR. Cas9 thus acts as a counter-selection acting directly on genomic sequence, rather than its phenotypic manifestations.

Here, we use an approach developed by Dueber and colleagues (Lee et al., 2015) to rapidly generate single, self-contained plasmids that express both the Cas9 nuclease and guide RNA required for targeting a desired locus. These plasmids, when co-transformed with an appropriate repair template provided as a linear PCR product, allow efficient, precise, single-step replacement of any arbitrary yeast gene with an introduced sequence of interest. Only selection for the Cas9 and gRNA-expressing plasmid is required, which tends to select for correct genomic modification by proxy due to efficiency of targeting and repair. This strategy was used extensively in our ortholog complementation research (Kachroo et al., 2017) to rapidly humanize, bacterialize and plantize many essential yeast genes. A CRISPR based approach is uniquely suited to this case, because it strongly encourages HR with functional alleles. False positives, arising from CRISPR sites being mutated by NHEJ without incorporation of a new allele, are minimal because they are often not viable. Additionally, disruption of the target gene’s function is brief, eliminating the need for constructing and maintaining a complementing plasmid to sustain yeast through an otherwise lengthy engineering process. Further, given that CRISPR selects against sequence regardless of function, it is still possible and practical to alter non-essential genes (or even non-genic regions) with this technique; indeed, we have reported successful humanization of the non-essential yeast gene HEM14 with this method (Kachroo et al., 2017) and we have used this system to incorporate site-directed changes in proteins with high efficiency.

Materials and Reagents

  1. Pipette tips (Mettler Toledo, catalog numbers: 17005872 , 17005874 , 17007089 )
  2. 96-well plate (VWR, catalog number: 82006-636 )
  3. 0.2 µm filter (Fisher Scientific, catalog number: 09-719C )
  4. Petri plates (VWR, catalog number: 25384-342 )
  5. Yeast (BY4741)
  6. MoClo Yeast Toolkit (YTK, Addgene kit, Addgene, catalog number: 1000000061 ). Toolkit includes plasmids pYTK050, pYTK003, pYTK072, pYTK083, pYTK036, pYTK008, pYTK047, pYTK073, pYTK074, pYTK081 and pYTK084
  7. PCR template for the sequence which will replace the target gene (e.g., cDNA, plasmid-based clone, etc.)
    Note: For demonstration purposes, this protocol will assume replacement of S. cerevisiae HEM2 with its human ortholog ALAD.
  8. NEB 5-alpha Competent E. coli (New England Biolabs, catalog number: C2987 )
  9. DNA stain (Thermo Fisher Scientific, InvitrogenTM, catalog number: S33102 )
  10. T7 ligase (New England Biolabs, catalog number: M0318S )
  11. T4 ligase buffer (New England Biolabs, catalog number: B0202S )
  12. Restriction enzymes BsaI (New England Biolabs, catalog number: R0535S ) and BsmBI (New England Biolabs, catalog number: R0580S )
  13. LB plates with antibiotic selection
    a. Ampicillin (Sigma-Aldrich, Roche Diagnostics, catalog number: 10835242001 )
    b. Spectinomycin (Sigma-Aldrich, catalog number: PHR1426 )
  14. Chloramphenicol (Sigma-Aldrich, catalog number: C0378 )
  15. High-fidelity DNA polymerase for repair template PCR, such as KAPA HiFi (Kapa Biosystems, catalog number: KK2601 )
  16. Zymo DNA Clean&Concentrator-25 kit (Zymo Research, catalog number: D4005 )
  17. Zymo EZ yeast transformation II kit (Zymo Research, catalog number: T2001 )
  18. Optional: 100 mM lithium acetate can be used in place of EZ 1 solution from the EZ competent yeast cell kit. (Lithium acetate can be obtained from Sigma-Aldrich, catalog number: L6883 )
  19. Accuprime Pfx (Thermo Fisher Scientific, InvitrogenTM, catalog number: 12344024 )
  20. Optional: 5-fluoroorotic acid (Sigma-Aldrich, catalog number: F5013 ), if counter-selection will be used (see Procedure E)
  21. D-Sorbitol (Sigma-Aldrich, catalog number: S3889 )
  22. Zymolyase (MP Biomedicals, catalog number: 320921 )
  23. LB Broth, Lennox (BD, catalog number: 240210 )
  24. YPD powder (BD, catalog number: 242820 )
  25. Agarose (Thermo Fisher Scientific, InvitrogenTM, catalog number: 16500500 )
  26. Agar (SERVA Electrophoresis, catalog number: 11396 )
  27. Yeast nitrogen base without amino acids (BD, catalog number: 291940 )
  28. Ammonium sulfate (Sigma-Aldrich, catalog number: A4418 )
  29. Dextrose (Avantor Performance Materials, catalog number: 1919 )
  30. SC-Ura dropout powder (Sigma-Aldrich, catalog number: Y1501 )
  31. Zymolyase solution (see Recipes)
  32. Lithium acetate (see Recipes)
  33. LB medium (see Recipes)
  34. YPD agar plates (see Recipes)
  35. SD-Ura agar plates (see Recipes)

Equipment

  1. Thermocycler (Bio-Rad Laboratories, catalog number: 1861096 )
  2. Light source for visualization of DNA stain (Thermo Fisher Scientific, InvitrogenTM, catalog number: G6600 )
  3. 12-channel pipette (Mettler Toledo, catalog number: 17013810 )
  4. Standard gel electrophoresis tank and accessories (Bio-Rad Laboratories, catalog number: 1640302 )
  5. Autoclave

Software

  1. Geneious v8.0 (Kearse et al., 2012) or higher, to design gRNA and repair template (replacement gene). Other gRNA design software can be used as well, such as E-CRISP (Heigwer et al., 2014)
  2. BLAT (Kent, 2002)

Procedure

  1. Preparation of CRISPR plasmid (for a diagrammatic overview of the cloning process, see Figure 1)


    Figure 1. Overview of the CRISPR plasmid construction process. In the first step Xs and Ys represent the gRNA sequence selected, and BsmBI recognition site is indicated in bold.

    1. Design two guide RNA (gRNA) sequences targeting the open reading frame (ORF) for the yeast gene to be replaced using Geneious, or a similar tool such as E-CRISP (Heigwer et al., 2014).
      1. gRNA sequences can often have low activity in practice, despite being predicted to be highly efficient by software tools. In order to minimize setbacks due to a gRNA which turns out to function poorly, we advise designing multiple gRNAs from the outset, and taking them through the cloning steps in parallel, up to and including the construction of the CRISPR plasmids. Both plasmids should then be tested for their ability to target the yeast genome and kill cells (described in later steps) to empirically determine and confirm their activity.
      2. We have not noticed a strong effect of the location of the gRNA within the ORF. During homologous repair, DNA can be resected up to several kilobases from the break site (Mimitou and Symington, 2009; Chen et al., 2011), so the gRNA need not be very close to either terminus of the ORF. It is however important to select a gRNA such that the target site is not present after replacement (i.e., the gRNA should target the yeast ORF, but not the replacement gene).
      3. Example: For targeting HEM2, the sequences GGATTATCGGAGATGAATAG (‘sg1’, on the non-coding strand) and CCTGGTACCAAGGATCCAGT (‘sg2’, on the coding strand) were predicted to have high activity (see Figure 2).


        Figure 2. Diagram of the native yeast HEM2 locus, showing positions of the example guide RNAs sg1 and sg2

    2. Order forward and reverse oligonucleotides with the gRNA sequence and Golden Gate compatible overlaps:
      1. Forward oligo consists of the 5’ insert GACTTT followed by the 20 bp guide sequence specific to the target gene. Example forward oligo for HEM2 sg1 (underline indicates 5’ Golden Gate overhang): GACTTTGGATTATCGGAGATGAATAG.
      2. Reverse oligo consists of the 3’ insert AAAC, followed by the reverse complement of the 20 bp guide sequence, followed by AA, which complements part of the GACTTT insert on the forward oligo. Example reverse oligo for HEM2 sg1 (underline indicates 3’ Golden Gate overhang): AAACCTATTCATCTCCGATAATCCAA.
    3. Mix forward and reverse oligos (50 µM each) for each gRNA in a total volume of 20 µl and anneal with each other using a thermocycler with the program below. It is unnecessary to phosphorylate the insert.
      95 °C for 5 min
      55 °C for 15 min
      25 °C for 15 min
    4. First Golden Gate cloning reaction to transfer into shuttle vector: Set up cloning reaction with annealed oligos and pYTK050 (Table 1).
      A 2:1 molar ratio of insert:plasmid is recommended for optimal Golden Gate cloning of linear DNA.

      Table 1. Golden Gate reaction for cloning into shuttle vector


    5. Transform the reaction into competent bacteria and plate with chloramphenicol selection (170 µg/ml). View colonies under UV light and pick the white colonies (those not showing GFP fluorescence), then grow in liquid culture and purify plasmid. The vectors used in Golden Gate reactions described in this protocol are all GFP-dropout vectors: They contain a GFP gene which will be silenced upon successful cloning. Therefore, GFP fluorescence indicates an invalid construct, while successful constructs will lose the GFP gene and the resulting colonies will be white.
      Optionally, the plasmid can be sequenced to check for errors or mutations in the gRNA sequence, such as may occur during synthesis.
    6. Second Golden Gate cloning reaction to create gRNA cassette plasmid: Set up cloning reaction which includes connector plasmids ConL1 and ConRE (Table 2).
      For best efficiency, all plasmids should be present at the same molarity in plasmid-based Golden Gate assemblies.

      Table 2. Golden Gate reaction for cloning gRNA cassette plasmid


    7. Transform the reaction into competent bacteria and plate with ampicillin selection (60 µg/ml). View colonies under UV light and pick the white colonies (those not showing GFP fluorescence), then grow in liquid culture and purify plasmid.
    8. Third and final Golden Gate cloning reaction to construct the yeast-compatible, complete CRISPR plasmid: Set up Golden Gate cloning reaction with connector plasmid from the previous step, and yeast –Ura backbone plasmid, and Cas9 plasmid (Table 3).

      Table 3. Golden Gate reaction for cloning CRISPR plasmid

      *Cen6-Ura is constructed by assembling YTK plasmids (008, 047, 073, 074, 081, and 084).

    9. Transform the reaction into competent bacteria and plate with kanamycin selection (50 µg/ml). View colonies under UV light and pick the white colonies (those not showing GFP fluorescence), then grow in liquid culture and purify plasmid.
      The resulting construct is a self-contained CRISPR plasmid, which when transformed into yeast will cause double-stranded breaks (DSBs) at the locus determined by the gRNA sequence cloned into it. 500 ng of this will be used for each yeast transformation, so if multiple replacements are planned, it is helpful to dilute the CRISPR plasmid to a standardized concentration for easier transformation set up later on.

  2. Preparation of repair template DNA
    1. Design the template DNA using Geneious or any other cloning software. Obtain the genomic sequence of the target yeast gene (‘old gene’), and the coding sequence (CDS) of the replacing gene (‘new gene’). The CDS should not contain introns. Create a gene model for the replaced locus by editing the sequence of the old gene so that it contains the new gene in the correct position (i.e., the desired outcome of replacement).
      We find that replacement works best if the original yeast stop codon is left intact. Otherwise, modifying the new gene, for instance to codon optimize for yeast, has proven unnecessary.
    2. Design template PCR primers which anneal to about 25 bp of the 5’ and 3’ ends of the new gene’s CDS, and also the 5’ and 3’ UTR immediately adjacent to the ORF (the homology arms). Figure 3 shows an example of primer design for replacing the yeast HEM2 gene with its human ortholog ALAD. This process is much easier using the gene model constructed in the previous step: The sequence covering the junction points between yeast genome and the new gene CDS can be used directly as primer sequence.
      1. The length of the region complementary to the new gene CDS is determined only by standard PCR efficiency concerns, such as melting temperature. This area will serve as a toehold for the first few cycles of the PCR.
      2. The length of the homology arms is critical for efficient replacement. We find that homologies of at least 70 bp are necessary (in which case the entire primer oligo will be about 90 bp long), and for some genes, 170 bp homologies may be necessary. For even more difficult replacements, longer homology arms can be cloned separately, but we have found that homologies longer than 500 bp are unlikely to increase efficiency further.


        Figure 3. Diagrams of example template primer designs for the replacement of HEM2 with hsALAD

    3. Use template PCR primers to amplify a large amount of repair template DNA using a high-fidelity polymerase.
      1. We find that it is helpful to first conduct several test PCRs with different polymerases. Due to the particular design of the template primers, this PCR can sometimes run inefficiently or generate unwanted non-specific products. Different polymerases have different characteristics, and often a reaction which fails with one polymerase will run efficiently with another, rendering laborious PCR optimization unnecessary.
      2. At least 5 µg of template DNA is needed per yeast transformation, which can usually be obtained from a single 50 µl PCR. Difficult replacements can often be facilitated by using more (10 µg) template DNA, and if multiple transformations are to be performed the amount will also need to be scaled up accordingly. Often several PCRs are necessary to produce enough DNA.
      3. If very large amounts of template DNA are needed, or an efficient PCR is difficult to set up, an alternative method is to clone the template sequence onto a plasmid, which can be amplified in bacteria with the template DNA excised using restriction enzymes.
    4. Check the template PCR with agarose gel electrophoresis.
      As long as a sufficient amount of the correct template is produced, non-specific products do not necessarily constitute a problem for the replacement. Because the non-specific products usually lack appropriate homologies, they will not be efficiently integrated into the yeast genome. However, if significant amounts of them are present, they will cause over-estimation of template DNA during spectrophotometry-based quantification; thus the amount of template DNA used in the transformation would need to be adjusted accordingly. Alternatively, the PCR can be optimized to reduce non-specific products, or only the correct product can be quantified from the gel using a DNA ladder calibrated for quantity estimation.
    5. Purify template PCR using the Zymo DNA Clean&Concentrator-25 kit. Elute in double distilled water.
      Ideally, the volume of DNA included in yeast transformation should be small, so as to not interfere with the transformation reagents. The elution volume should be adjusted accordingly so that the resulting concentration of DNA is not too low. In our experiments, we have found that eluting with 25 µl double distilled water will usually yield 400-800 ng/µl DNA, which is suitable for transformations.

  3. Yeast transformation
    1. Prepare competent yeast cells using the Zymo EZ competent yeast kit according to the kit instructions.
      The EZ 1 solution in this kit can be substituted with 100 mM lithium acetate without significant change in transformation efficiency.
      The amounts given in the kit manual can be slightly modified: 2 ml yeast culture can be used to produce 100 µl of competent yeast, which is sufficient for two transformations, 50 µl each.
    2. Set up a transformation reaction: Mix 50 µl competent yeast, 500 µl EZ 3 solution, 500 ng of CRISPR plasmid and 5 µg repair template DNA (up to 50 µl total volume). Incubate at 30 °C as directed by kit manual and plate on –Ura medium.
      When using a new gRNA for the first time, gRNA efficiency can be estimated with a control transformation, which is performed as stated but without repair DNA. When the CRISPR plasmid is introduced without a repair template, it will repeatedly cleave the target locus, causing toxicity. Very few or no colonies are the ideal outcome, since this indicates highly efficient CRISPR cleavage and low background rate. Cells can survive the CRISPR plasmid uptake without repair DNA if the CRISPR activity is stochastically low (such as due to poor gRNA efficiency) or mutations at the CRISPR target locus can be tolerated (which produces false transformants even in presence of the repair template).
    3. When colonies appear on the –Ura plates, collect up to 12 of them with a pipette tip and suspend in 50 µl water. These suspensions will be screened for confirmed replacements. Yeast suspensions can be stored at 4 °C and used to start new cultures for up to 2 weeks.
      1. Typically, colonies will appear on –Ura plates (Figure 4) after 1-3 days. In some cases, the replacement will impose a significant fitness defect such that up to 6 days may be required for colonies to appear, but we have not encountered cases where colonies from a successful transformation take longer than 6 days to grow.


        Figure 4. Representative assay results. Yeast cells are rescued from DSB lethality (center plate) when an appropriate repair template is provided (right plate). The left plate is a negative control of cells carrying a control plasmid with the same selectable marker (URA3) done to estimate the transformation efficiency of the yeast strains being used.

      2. The uracil dropout medium will select against cells which failed to take up the CRISPR plasmid (which confers uracil prototrophy), but because the CRISPR plasmid is toxic to cells unless a successful replacement occurs (eliminating the CRISPR target locus) only cells which have a replaced locus are expected to survive. However, due to spontaneous hypoactivity of the CRISPR system, mutations in the CRISPR target locus (DiCarlo et al., 2013), and cells which manage to survive CRISPR-associated DSBs, there will be a background rate in the form of false transformant colonies which do not carry the correct genomic replacements. To save time, we recommend collecting several transformant colonies and screening them in parallel.
      3. To streamline this process (especially when several replacements are performed in parallel), pick colonies with pipette tips and manually attach them to a multichannel pipette (Figure 5). The multichannel pipette can then be used to suspend all 12 samples in one row of small PCR tubes or a 96-well plate.


        Figure 5. Demonstration of colony picking technique with 12-channel pipette

  4. Colony screening via PCR
    1. Design confirmation PCR primers: Primer pairs should be selected such that the forward primer anneals to the yeast UTR while the reverse primer anneals only to the new gene CDS but not the old gene’s ORF. Thus, the product should span the junction point between foreign sequence and native yeast genome. The yeast UTR primer should preferably not overlap the homology region.
      1. Ideally, the product size should be small, about 300 bp, for a faster and more robust PCR.
      2. It is sufficient to check only the 5’ junction point, since it is rare for integration to proceed as expected at one end of the gene but introduce artifacts at the other.
      3. If desired, the absence of the yeast ORF can also be tested by using a reverse primer which anneals to yeast ORF only. However, lack of product from such a primer pair is not sufficient to confirm a clone, since the reaction is liable to fail for unrelated reasons (such as poor lysis of cells).
    2. Prepare lysates of harvested transformants: Mix 5 µl of each yeast suspension with 15 µl zymolyase solution.
    3. Incubate lysates for 30 min at room temperature, then 15 min at 37 °C and 5 min at 95 °C.
    4. Set up 20 µl colony PCRs with confirmation primers and using Accuprime Pfx as the polymerase. Use 1 µl of the lysate as template DNA.
      1. We find that other polymerases do not perform well due to impurities from the yeast lysates.
      2. Due to the impurities introduced by the lysate, the colony PCR may spontaneously fail, leading to false negatives. To ameliorate this problem, a positive control PCR can be performed for each lysate, which is identical to the confirmation PCR but uses primers complementary to an unrelated, unmodified locus in the genome. We use two primers targeting a 500 bp segment of the yeast ERG13 promoter for this purpose (forward CGAACTGGATGAGATGGCCG and reverse CATGCTGCACCTTTTATAGTAATTTGGC).
    5. Check the colony PCRs for product by agarose electrophoresis. Lysates from clones with the correct modifications should generate a product with the confirmation primers. Background false transformants (e.g., mutants) will not produce a band.
      1. A PCR product from the confirmation primers is sufficient evidence of successful integration of the repair template. For further verification, the locus can be sequenced, but we have found that dramatic sequence artifacts rarely occur in clones confirmed by PCR, the most common mutations are single-basepair substitutions or indels, which typically constitute a minority of confirmed clones.
      2. Lack of product from the confirmation primers is inconclusive per se. In such cases, it is worthwhile to consider additional evidence, such as whether the positive control PCR worked (if not, the lysis may have failed).
    6. Confirmed clones can be propagated by starting a new culture from the original suspensions of yeast in water.

  5. Curing of the CRISPR plasmid
    1. Streak original water suspensions of confirmed clones on YPD.
      The CRISPR plasmid is low copy and can be spontaneously lost in absence of selection.
    2. Pick 10 colonies from the YPD plate and patch each one on YPD and SD-Ura plates.
    3. Incubate both plates, and collect cells from patches which grew only on YPD but not on SD-Ura.
      Isolates which still carry the CRISPR plasmid will grow on uracil dropout medium, but those which have lost the plasmid will not. Typically, 3 days is sufficient to confirm lack of Ura prototrophy, but if slow growth on uracil dropout is suspected, incubation can be extended to up to 6 days to definitively confirm no growth on uracil dropout.
      The plasmid can also be cured by counterselecting on 5-fluoroorotic acid (FOA) plates (Boeke et al., 1987). However, there is a possibility that this FOA method will generate some colonies that are not cured of the plasmid but rather have acquired a mutation in the Ura marker (thus continuing to express the gRNA). Thus, FOA counterselection should not be used (as opposed to replicate patches on YPD and –Ura) if it is important to ensure curing of the plasmid, rather than simply abrogating Ura prototrophy. On the other hand, the FOA method can save time if only loss of –Ura heterotrophy is desired, for instance to enable a subsequent transformation with a different Ura-selectable plasmid.

Data analysis

The data analysis needs for this procedure are minimal. Most importantly, when using Geneious to design gRNA sequences, it is desirable to select gRNA sequences that have high predicted on-target activity (automatically calculated by Geneious). gRNA sequences with high predicted activity may have low actual activity, but they will be less likely to exhibit low activity than sequences with low predicted activity. The distance of the gRNA target site can be up to 1 kb away from either homology region without perceptible negative consequence, thus gRNAs should be selected primarily based on high activity rather than location (provided that they lie between the two homology arms).

Notes

  1. We have found that even among gRNAs with high predicted activity, some will fail to induce double-strand breaks with sufficient efficiency for editing. It is highly recommended that for each target locus, several gRNA are designed and tested in parallel, to ensure that at least one will be a sufficiently good DSB inducer for purposes of genome editing.
  2. If a given gRNA exhibits significant off-target activity, the likely outcome is that off-target cleavage will kill most of the transformed yeast cells. Successful, efficient genome editing in yeast relies on lethality associated with DSBs at the target locus being rescued by HR (allowing efficient repair of the DSB) and abrogation of the gRNA target site (preventing further cleavage). In the event off-target activity, HR may likely not take place because no repair template with homology to the off-target site has been supplied, moreover the gRNA site will not be eliminated for the same reason. Further, the confirmation strategy we suggest is such that only repair at the correct locus will produce a positive result. However, it is nevertheless worthwhile to ensure that selected gRNA target sites do not occur at other locations in the genome, where cleavage is not intended. Although it is very unlikely for the combined 23 bp target sequence to appear multiple times in the yeast genome, we recommend confirming that candidate gRNA sites appear only in the target locus using a tool such as BLAT.
  3. gRNA targets consist of a 20 bp sequence (which will also be included in sgRNA sequence and become part of the Cas9 complex) followed by a 3 bp PAM sequence (which takes the form of NGG for Cas9 described in this protocol). The PAM sequence does not become part of the gRNA, but it must be present in the target genome for Cas9 cleavage to occur. This can be verified by attempting to align the gRNA sequence to the sequence of the repair template–typically, CRISPR activity will be very low with more than 5 mismatching basepairs, although mismatches in the PAM and proximal to the PAM appear to have more significance (Kuscu et al., 2014). When replacing with very similar sequences, such that it is difficult to find good gRNA sites unique to the target locus, one strategy that can be adopted is to introduce synonymous mutations in the repair template sequence which alter the PAM site or PAM-proximal nucleotides. Alternatively, recent research suggests that using shorter gRNA may increase specificity, since the 8-17 PAM-proximal nucleotides contribute disproportionately to CRISPR target recognition (Xu et al., 2017).
  4. There is some variability in the yeast transformation step, and depending on how the competent cells were prepared, and how the transformation was performed. Most commonly, the number of resulting colonies will vary somewhat between transformations of identical strains with identical reagents, but usually this variation will be less than tenfold. When a transformation produces a fair number of colonies (at least 10) yet none of them are found to be correct clones upon screening, simply repeating the transformation is unlikely to improve results. The most straightforward avenues of increasing the number of correct clones are to increase the amount of repair template DNA, and to produce repair template DNA with longer homologies.
  5. If no colonies appear after transformation, the reason may be low transformation efficiency. In this case, several troubleshooting steps can be taken (described in detail in the documentation of the Zymo EZ competent yeast kit). We have found the following to be effective:
    1. Thoroughly vortexing the mixture of competent cells and DNA.
    2. Longer incubation time for the transformation (1.5 h instead of the 45 min).
    3. Including more cells in the transformation.
    4. Competent cells seem to perform slightly better when frozen once (slowly in -80 °C) than freshly prepared cells.
  6. When the CRISPR reagents and repair template are transformed into yeast cells, the resulting transforming colonies will be of three kinds with respect to the targeted locus:
    1. Correct transformants which bear the sequence of the repair template.
    2. False transformants which bear the original, unedited sequence.
    3. Mutants.
  7. In our experiments, we have found that the first two classes predominate unless mutants are specifically selected for. Even in the absence of a repair template, the majority of false transformants will not be mutants. Due to the efficient HR system of S. cerevisiae, if the conditions of the experiment are adequate then editing will take place at a very high rate. Thus, typically, the proportion between the first two of the three classes listed above will be such that the transformants are either mostly correct or all false. The third class, or mutants, we have found to be very rare in either case unless specifically selected for. As a consequence, it is rarely necessary to screen a very large number of colonies to determine whether an editing experiment has succeeded. However, it is desirable to collect several confirmed clones to minimize issues caused by artifacts, such as mutant edited sequence caused by errors during PCR (with the reagents and protocols described in this text, we have found clones with mutant edited sequence also be very rare).
  8. Selecting yeast transformants with a single amino-acid dropout medium is normally a straightforward process, and colonies can be seen within 1-2 days of plating. However, occasionally the genome editing process itself, or the resulting edited sequence, can result in a growth defect in the resulting cells. Thus, if no colonies appear, incubating the plate for a longer period can produce colonies. In the most extreme case we observed, it took 6 days for colonies to appear on a uracil dropout medium, but several clones were later confirmed by PCR and sequencing; these clones consistently exhibited slow growth in subsequent culture on rich medium (YPD) as well.
  9. Some combinations of target locus and repair template may lead to a mixture of large and small yeast colonies after transformation. If this occurs, generally it is best to screen an adequate number of colonies for each size class. It may be that the correct edits create much slower growing strains, thus the large colonies are false while the small ones have the desired edit. Conversely, if the desired sequence does not interfere with normal growth, but mutations arising from NHEJ do, then larger colonies will tend to be the correct clones. We have observed examples of either case when humanizing and bacterializing various loci. It is difficult to predict a priori which case will be evident for a given transformation, therefore it is often more practical to screen colonies and recording their size, and also ensuring that each size is adequately represented in the screen.
  10. When picking colonies for the colony PCR screen, only a small quantity of cells is needed. Most likely as little as 1,000 cells will be sufficient to obtain a PCR product. We have often chosen to collect slightly larger numbers of cells to visually confirm their suspension in water by turbidity. However, too many cells lead to incomplete lysis and inhibition of the colony PCR. With cell clumps larger than 1-2 mm the colony PCR will often fail. So ideally, the cells collected from the colony should form only a tiny speck, 0.5 mm or smaller in diameter. It is helpful to include the positive control PCR when screening, to identify samples which failed to produce a PCR product due to poor lysis. Lysis and PCR can be repeated for these samples if needed.
  11. It is possible to adapt the protocol described here for the simultaneous replacement of multiple genes. The Mo Clo toolkit allows for cloning up to 4 different gRNA cassettes on the same CRISPR plasmid; for this, the gRNAs would be captured on pYTK050 as described here, but in the second Golden Gate reaction, instead of the ConL1 and ConRE plasmids, the first gRNA would be cloned with ConL1 and ConR2, the second with ConL2 and ConR3, the third with ConL3 and ConR4 and the fourth with ConL4 and ConRE (this process is explained in detail in Lee et al., 2015). All of these cassette plasmids would then be included in the final Golden Gate reaction to assemble the CRISPR plasmid. Then, during transformation of yeast, templates for each of the included gRNAs will need to be co-transformed. However, multiple replacements are even more dependent on efficient transformation, cleavage and repair than single replacements, and some additional work may be necessary to optimize these parameters in practice.

Recipes

  1. Zymolyase solution (50 ml)
    1. Weigh 9.11 g D-sorbitol
    2. Dissolve in 50 ml distilled, deionized water to make 1 M sorbitol and autoclave
    3. Weigh 0.25 g zymolyase and dissolve in sorbitol solution
    4. Aliquot and store at -20 °C
  2. Lithium acetate, 100 mM (40 ml)
    1. Weigh 0.408 g lithium acetate dehydrate
    2. Dissolve in 40 ml distilled, deionized water
    3. Filter sterilize (0.2 µm filter) and store at room temperature
  3. LB medium (1 L)
    1. Weigh 25 g LB powder
    2. For solid medium, add 15 g agar
    3. Dissolve in distilled, deionized water for 1 L total volume
    4. Autoclave and let it cool to 60-70 °C
    5. Pour in Petri plates so that the medium covers the visible area of the plate
    6. Let plates cool and solidify at room temperature, store at 4 °C
  4. YPD (1 L)
    1. Weigh 50 g YPD powder
    2. For solid medium, add 20 g agar
    3. Dissolve in distilled, deionized water for 1 L total volume
    4. Autoclave and let it cool to 60-70 °C
    5. Pour in Petri plates so that the medium covers the visible area of the plate
    6. Let plates cool and solidify at room temperature, store at 4 °C
  5. SD-Ura (1 L)
    1. Weigh 1.5 g yeast nitrogen base w/o amino acids, 5 g ammonium sulfate, 20 g dextrose, 2 g SC-Ura dropout powder
    2. For solid medium, add 20 g agar
    3. Dissolve in distilled, deionized water for 1 L total volume
    4. Autoclave and let it cool to 60-70 °C
    5. Pour in Petri plates so that the medium covers the visible area of the plate
    6. Let plates cool and solidify at room temperature, store at 4 °C

Acknowledgments

This work was supported by grants from the NIH (R21 GM119021, R01 HD085901, DP1 GM106408, R01 DK110520, R35 GM122480), Army Research Office (ARO) grant W911NF-12–1–0390, and the Welch Foundation (F-1515) to E.M.M. We would like to thank John Dueber & colleagues for producing the excellent plasmid toolkit which greatly facilitated our work. The authors declare no conflicts of interest.

References

  1. Boeke, J. D., Trueheart J., Natsoulis G. and Fink G. R. (1987). 5-Fluoroorotic acid as a selective agent in yeast molecular genetics. Methods Enzymol 154: 164-75.
  2. Chen, X., Niu, H., Chung, W. H., Zhu, Z., Papusha, A., Shim, E. Y., Lee, S. E., Sung, P. and Ira, G. (2011). Cell cycle regulation of DNA double-strand break end resection by Cdk1-dependent Dna2 phosphorylation. Nat Struct Mol Biol 18(9): 1015-1019.
  3. DiCarlo, J. E., Norville, J. E., Mali, P., Rios, X., Aach, J. and Church, G. M. (2013). Genome engineering in Saccharomyces cerevisiae using CRISPR-Cas systems. Nucleic Acids Res 41(7): 4336-4343.
  4. Heigwer F., Kerr G. and Boutros M. (2014). E-CRISP: fast CRISPR target site identification. Nat Methods 11: 122-123.
  5. Kachroo, A. H., Laurent, J. M., Akhmetov, A., Szilagyi-Jones, M., McWhite, C. D., Zhao, A. and Marcotte, E. M. (2017). Systematic bacterialization of yeast genes identifies a near-universally swappable pathway. Elife 6: e25093.
  6. Kearse, M., Moir, R., Wilson, A., Stones-Havas, S., Cheung, M., Sturrock, S., Buxton, S., Cooper, A., Markowitz, S., Duran, C., Thierer, T., Ashton, B., Meintjes, P. and Drummond, A. (2012). Geneious Basic: an integrated and extendable desktop software platform for the organization and analysis of sequence data. Bioinformatics 28(12): 1647-1649.
  7. Kent, W. J. (2002). BLAT – The BLAST-Like Alignment Tool. Genome Res 12(4): 656-64.
  8. Kuscu, C., Arslan, S., Singh, R., Thorpe, J. and Adli, M. (2014). Genome-wide analysis reveals characteristics of off-target sites bound by the Cas9 endonuclease. Nat Biotechnol 32(7): 677-83.
  9. Lee, M. E., DeLoache, W. C., Cervantes, B. and Dueber, J. E. (2015). A highly characterized yeast toolkit for modular, multipart assembly. ACS Synth Biol 4(9): 975-986.
  10. Mimitou, E. P. and Symington, L. S. (2009). DNA end resection: many nucleases make light work. DNA Repair (Amst) 8(9): 983-995.
  11. Xu, X., Duan, D. and Chen, S. (2017). CRISPR-Cas9 cleavage efficiency correlates strongly with target-sgRNA folding stability: from physical mechanism to off-target assessment. Scientific Reports 7(143).

简介

芽殖酵母中的基因组修饰已经非常成功,主要归功于其高度同源性的DNA修复机制。之前已经描述了几种用于修饰酵母基因组的方法,其中许多方法涉及至少两个步骤:插入选择标记并用该标记取代预期的修饰。在这里,我们描述了CRISPR-Cas9介导的基因组编辑方案,用于在没有任何选择标记的情况下在单步转化中修饰任何感兴趣的酵母基因(基本或非必需)。在该系统中,Cas9核酸酶在选择的基因座处产生双链断裂,这在酵母细胞中通常是致死的,而不管由于无效的非同源末端连接修复导致的靶基因座的重要性。该致死性通过使用源自PCR的修复模板的同源重组导致有效的修复。在涉及必需基因的情况下,用功能性等位基因编辑基因组病变的必要性作为额外的选择层。作为一个激励性的例子,我们描述了使用这种策略替代HEM2,一种必需的酵母基因,以及相应的人类直向同源物ALAD。

【背景】酿酒酵母(Baccharomyces cerevisiae,Baker's酵母)作为一种遗传易处理的生物体具有悠久的历史,并且有许多操作酵母基因组的方法。然而,直到最近,有必要应用选择以分离具有所需遗传改变的克隆(Kearse等人,2012; DiCarlo等人,2013; Lee等人,等,2015; Kachroo等,,2017)。在需要对基因组进行任意的无瘢痕编辑的情况下,该解决方案通常是两步过程:首先是可选择的盒(例如含有URA3标记),两侧是针对感兴趣区域的同源臂;以及有时含有核酸酶靶向位点( ie ,I-SceI位点)以辅助在后期阶段除去盒,通过同源重组(HR)敲入。通过选择盒子来分离成功整合体的小亚群。其次,通过诸如位点特异性重组或核酸内切酶切割(I-SceI)的高效序列特异性方法消除标记以产生编辑的基因组基因座的期望形式。有两个步骤是必要的,因为没有方法可用,既没有疤痕又没有足够的效率,因此不需要选择。

CRISPR / Cas9技术在酵母中的开发已经消除了这个两步过程的需要。 Cas9在酵母DNA中几乎在任意位置有效地产生双链断裂(DSBs) - 只要PAM序列靠近期望的切割位点即可。当提供适当的修复模板时,这些DSB通过内源性酵母系统修复。通过引导RNA序列指向期望的基因组座位的Cas9在基因组中产生双链断裂(DSB)。 CRISPR靶位点被保留在未按预期修复靶位点的细胞中,这允许Cas9反复切割相同的区域,直到进行HR介导的编辑。很少,非同源末端连接(NHEJ)可以产生阻断Cas9切割的突变,尽管未能整合预期的基因组改变。更常见的是,细胞只是屈服于重复的Cas9诱导的基因组切割的压力。在一个适当的实验中,大部分存活的人群倾向于通过通过HR引入期望的基因组改变而失去其CRISPR靶位点的细胞。因此,Cas9充当直接作用于基因组序列的反选择,而不是其表型表现。

在此,我们使用Dueber及其同事开发的方法(Lee等人,2015)快速产生单个自包含的质粒,其表达靶向期望位点所需的Cas9核酸酶和指导RNA 。这些质粒当与作为线性PCR产物提供的适当修复模板一起共转化时,允许用引入的目的序列有效,精确,单步替换任意任意酵母基因。只需要选择Cas9和gRNA表达质粒,其倾向于选择由于靶向和修复的效率而通过代理进行正确的基因组修饰。该策略广泛用于我们的直向同源互补研究(Kachroo et al。,2017),以快速人源化,细菌化和平板化许多重要的酵母基因。基于CRISPR的方法非常适合这种情况,因为它强烈鼓励HR提供功能性等位基因。由于NHIS没有引入新的等位基因而引起的CRISPR位点突变引起的误报很少,因为它们往往不可行。另外,目标基因功能的破坏是短暂的,消除了构建和维持补充质粒以维持酵母通过另外冗长的工程过程的需要。此外,鉴于CRISPR针对序列选择而不考虑功能,使用该技术改变非必需基因(甚至非基因型区域)仍然是可能的和实际的;实际上,我们已经报道了用这种方法成功地将非必需酵母基因HEM14人源化(Kachroo等人,2017),并且我们已经使用该系统以高效率掺入蛋白质中的定点改变。

关键字:CRISPR, 同源重组, 人源化, 直系同源互补, 基因组编辑, 酵母工程

材料和试剂

  1. 移液器吸头(Mettler Toledo,产品目录号:17005872,17005874,17007089)
  2. 96孔板(VWR,目录号:82006-636)
  3. 0.2微米过滤器(Fisher Scientific,目录号:09-719C)
  4. 培养皿(VWR,目录号:25384-342)
  5. 酵母(BY4741)
  6. MoClo酵母工具包(YTK,Addgene试剂盒,Addgene,目录号:1000000061)。工具包包括质粒pYTK050,pYTK003,pYTK072,pYTK083,pYTK036,pYTK008,pYTK047,pYTK073,pYTK074,pYTK081和pYTK084。
  7. 用于替代靶基因(例如,cDNA,基于质粒的克隆,等等)的序列的PCR模板。
    注:为了演示目的,本协议将假定用其人类直向同源物ALAD取代酿酒酵母HEM2。
  8. NEB 5-alpha Competent大肠杆菌(New England Biolabs,目录号:C2987)
  9. DNA染色(Thermo Fisher Scientific,Invitrogen TM,目录号:S33102)
  10. T7连接酶(New England Biolabs,目录号:M0318S)
  11. T4连接酶缓冲液(New England Biolabs,目录号:B0202S)
  12. 限制性酶BsaI(New England Biolabs,目录号:R0535S)和BsmBI(New England Biolabs,目录号:R0580S)
  13. 选择抗生素的LB平板
    一个。氨苄青霉素(Sigma-Aldrich,Roche Diagnostics,目录号:10835242001)
    湾壮观霉素(Sigma-Aldrich,目录号:PHR1426)
  14. 氯霉素(Sigma-Aldrich,目录号:C0378)
  15. 用于修复模板PCR的高保真DNA聚合酶,如KAPA HiFi(Kapa Biosystems,目录号:KK2601)
  16. Zymo DNA Clean& Concentrator-25试剂盒(Zymo Research,目录号:D4005)
  17. Zymo EZ酵母转化II试剂盒(Zymo Research,目录号:T2001)
  18. 可选:可使用100 mM醋酸锂代替EZ主管酵母细胞试剂盒中的EZ 1溶液。 (乙酸锂可以从Sigma-Aldrich获得,目录号:L6883)
  19. Accuprime Pfx(Thermo Fisher Scientific,Invitrogen TM,目录号:12344024)
  20. 任选:如果将使用反选择(参见程序E),则使用5-氟乳清酸(Sigma-Aldrich,目录号:F5013)。
  21. D-山梨糖醇(Sigma-Aldrich,目录号:S3889)
  22. Zymolyase(MP Biomedicals,目录号:320921)
  23. LB肉汤,Lennox(BD,目录号:240210)
  24. YPD粉末(BD,目录号:242820)
  25. 琼脂糖(Thermo Fisher Scientific,Invitrogen TM,目录号:16500500)
  26. 琼脂(SERVA电泳,目录号:11396)
  27. 不含氨基酸的酵母氮碱(BD,目录号:291940)
  28. 硫酸铵(Sigma-Aldrich,目录号:A4418)
  29. 葡萄糖(Avantor Performance Materials,目录号:1919)
  30. SC-Ura辍学粉末(Sigma-Aldrich,目录号:Y1501)
  31. Zymolyase溶液(见食谱)
  32. 醋酸锂(见食谱)
  33. LB培养基(见食谱)
  34. YPD琼脂平板(见食谱)
  35. SD-Ura琼脂平板(见食谱)

设备

  1. 热循环仪(Bio-Rad Laboratories,目录号:1861096)
  2. 用于DNA染色可视化的光源(Thermo Fisher Scientific,Invitrogen TM,目录号:G6600)
  3. 12通道移液器(Mettler Toledo,产品目录号:17013810)
  4. 标准凝胶电泳槽和附件(Bio-Rad Laboratories,目录号:1640302)
  5. 高压灭菌器

软件

  1. Geneious v8.0(Kearse等人,2012)或更高,设计gRNA和修复模板(替换基因)。其他gRNA设计软件也可以使用,例如E-CRISP(Heigwer等人,2014)
  2. BLAT(Kent,2002)

程序

  1. CRISPR质粒的制备(用于克隆过程的图解概述,参见图1)


    图1. CRISPR质粒构建过程概述在第一步中,Xs和Ys代表所选的gRNA序列,BsmBI识别位点以粗体显示。

    1. 设计两个靶向开放阅读框(ORF)的指导RNA(gRNA)序列,用Geneious或诸如E-CRISP(Heigwer等人,2014)等类似工具替换酵母基因。 。
      1. gRNA序列在实践中通常可能具有较低的活性,尽管预测其通过软件工具高效。为了最大限度地减少由于gRNA导致功能不佳而引起的挫折,我们建议从一开始就设计多个gRNA,并通过克隆步骤平行进行,直至并包括构建CRISPR质粒。然后应该测试两种质粒靶向酵母基因组并杀死细胞的能力(在后面的步骤中描述)以经验确定并确认它们的活性。
      2. 我们还没有注意到ORF中gRNA位置的强烈影响。在同源修复过程中,DNA可从断裂位点切除数千个碱基(Mimitou和Symington,2009; Chen等人,2011),所以gRNA不必非常靠近ORF。然而选择一个gRNA使得替换后目标位点不存在是非常重要的(即,gRNA应该靶向酵母ORF,而不是替换基因)。
      3. 实施例:为了靶向HEM2,预测序列GGATTATCGGAGATGAATAG(在非编码链上的'sg1')和CCTGGTACCAAGGATCCAGT(编码链上的'sg2')具有高活性(参见图2)。


        图2.天然酵母HEM2基因座的图,显示了实例指南RNA sg1和sg2的位置

    2. 使用gRNA序列和Golden Gate兼容重叠序列转发和反转寡核苷酸:
      1. 正向寡核苷酸由5'插入GACTTT组成,然后是对目标基因特异的20bp指导序列。 HEM2 sg1的示例性正向寡核苷酸(下划线表示5'金色突出端):GAC TTGGATTATCGGAGATGAATAG。
      2. 反向寡核苷酸由3'插入序列AAAC组成,其后是20bp指导序列的反向互补序列,接着是AA,其补充正向寡核苷酸上的部分GACTTT插入片段。 HEM2 sg1的反向寡核苷酸示例(下划线表示3'Golden Gate overhang):AAAC CTATTCATCTCCGATAATCCAA。
    3. 混合正向和反向寡核苷酸(每个50μM),每个总RNA量为20μl,使用下面的程序使用热循环仪互相退火。
      磷酸化插入是没有必要的 95°C 5分钟
      55°C 15分钟
      25°C 15分钟
    4. 第一个金门克隆反应转移到穿梭载体中:用退火的寡核苷酸和pYTK050建立克隆反应(表1)。

      插入片段:质粒的摩尔比为2:1,推荐用于线性DNA的最佳金门克隆
      表1.克隆到穿梭载体中的金门反应


    5. 用氯霉素选择(170μg/ ml)将反应转化为感受态细菌和平板。在紫外光下观察菌落并挑选白色菌落(不显示GFP荧光的菌落),然后在液体培养物中生长并纯化质粒。在本议定书中描述的金门反应中使用的载体都是GFP-脱落载体:它们含有GFP基因,其在克隆成功时将被沉默。因此,GFP荧光指示无效构建体,而成功的构建体会失去GFP基因,并且所得到的菌落将是白色的。
      任选地,可以对质粒进行测序以检查gRNA序列中的错误或突变,例如可能在合成过程中发生。
    6. 第二次金门克隆反应以产生gRNA盒质粒:建立包含连接质粒ConL1和ConRE的克隆反应(表2)。
      为获得最佳效率,所有质粒都应以基于质粒的Golden Gate组件中的相同摩尔浓度存在。

      表2.用于克隆gRNA盒质粒的Golden Gate反应


    7. 将反应转化成具有氨苄青霉素选择(60μg/ ml)的感受态细菌和平板。在紫外光下观察菌落,挑选白色菌落(不显示GFP荧光的菌落),然后在液体培养基中生长并纯化质粒。
    8. 第三次也是最后一次金门克隆反应,构建酵母兼容的完整CRISPR质粒:用前一步骤的连接物质粒,酵母-Ura主链质粒和Cas9质粒(表3)建立金门克隆反应。 >
      表3.用于克隆CRISPR质粒的金门反应

      * Cen6 Ura由YTK质粒(008,047,073,074,081和084)组装而成。

    9. 用卡那霉素筛选(50μg/ ml)将反应转化为感受态细菌和平板。在紫外光下观察菌落,挑选白色菌落(不显示GFP荧光的菌落),然后在液体培养基中生长并纯化质粒。
      所得到的构建体是一种自含式CRISPR质粒,当转化到酵母中时,会在由克隆到其中的gRNA序列确定的基因座处引起双链断裂(DSB)。 500 ng将用于每次酵母转化,因此如果计划进行多次替换,将CRISPR质粒稀释到标准浓度以便稍后设置更容易的转化是有帮助的。

  2. 修复模板DNA的制备
    1. 使用Geneious或任何其他克隆软件设计模板DNA。获得目标酵母基因('旧基因')的基因组序列和替换基因('新基因')的编码序列(CDS)。 CDS不应含有内含子。通过编辑旧基因的序列,为被替换的基因座创建一个基因模型,以便它将新基因包含在正确的位置(即, ,即替换所需的结果)。
      我们发现,如果原始酵母终止密码子保持完整,替换效果最好。否则,修改新基因,例如密码子优化酵母,已被证明是不必要的。
    2. 设计模板PCR引物退火到新基因CDS的5'和3'末端的约25bp,以及紧邻ORF(同源臂)的5'和3'UTR。图3显示了用其人直系同源物ALAD取代酵母HEM2基因的引物设计实例。使用上一步构建的基因模型,这个过程更容易:覆盖酵母基因组与新基因CDS之间接合点的序列可直接用作引物序列。
      1. 与新基因CDS互补的区域的长度仅由标准PCR效率问题(例如解链温度)确定。
        这个区域将作为PCR前几个周期的起始点
      2. 同源臂的长度对于有效替换是至关重要的。我们发现至少70bp的同源性是必需的(在这种情况下,整个引物寡核苷酸长约90bp),对于一些基因,可能需要170bp的同源性。对于更加困难的替换,可以单独克隆更长的同源臂,但我们发现超过500bp的同源性不可能进一步提高效率。


        图3.用hsALAD替代HEM2的示例模板引物设计图

    3. 使用模板PCR引物使用高保真聚合酶扩增大量修复模板DNA。
      1. 我们发现首先用不同的聚合酶进行多个测试PCR是有帮助的。由于模板引物的特殊设计,该PCR有时可能效率低下或产生不需要的非特异性产物。不同的聚合酶具有不同的特性,并且通常一个聚合酶失败的反应将与另一个聚合酶高效运行,从而导致不必要的繁琐的PCR优化。
      2. 每个酵母转化需要至少5μg的模板DNA,通常可以从一个50μlPCR获得。通过使用更多(10μg)模板DNA,通常可以促进难以取代,并且如果要进行多重转化,则量也将相应地扩大。常常需要几个PCR来产生足够的DNA。
      3. 如果需要非常大量的模板DNA,或者难以建立有效的PCR,另一种方法是将模板序列克隆到质粒上,该质粒可以用限制酶切割的模板DNA在细菌中扩增。 />
    4. 用琼脂糖凝胶电泳检查模板PCR。
      只要生产出足够数量的正确模板,非特定产品就不一定构成替换问题。由于非特异性产物通常缺乏适当的同源性,因此不能有效整合到酵母基因组中。但是,如果存在显着量的它们,则在基于分光光度法的量化过程中它们会导致模板DNA的高估;因此转化中使用的模板DNA的量将需要相应地调整。或者,可以对PCR进行优化以减少非特异性产物,或者使用经过校准的DNA梯形图进行数量估算,只能从凝胶中定量正确的产物。
    5. 使用Zymo DNA Clean& Concentrator-25试剂盒纯化模板PCR。在双蒸水中洗脱。
      理想情况下,包含在酵母转化中的DNA体积应该很小,以免干扰转化试剂。洗脱体积应相应调整,以使得到的DNA浓度不会太低。在我们的实验中,我们发现用25μl双蒸水洗脱通常会产生400-800 ng /μl的DNA,适合转化。

  3. 酵母转化
    1. 根据试剂盒使用说明书,使用Zymo EZ主管酵母试剂盒准备感受态酵母细胞。
      该试剂盒中的EZ 1溶液可用100 mM乙酸锂代替,而不会显着改变转化效率。
      试剂盒手册中给出的数量可以稍作修改:2 ml酵母培养物可用于生产100μl感受态酵母,这足以进行两次转化,每次转化50μl。
    2. 设置转化反应:将50μl感受态酵母,500μlEZ 3溶液,500ng CRISPR质粒和5μg修复模板DNA(总体积达50μl)混合。
      按照试剂盒手册的指示在30°C孵育,并在-Ura培养基上培养 当第一次使用新的gRNA时,可以用对照转化来估计gRNA效率,如所述的那样进行,但没有修复DNA。当CRISPR质粒在没有修复模板的情况下被引入时,它将反复切割目标基因座,导致毒性。很少或无殖民地是理想的结果,因为这表明高效CRISPR卵裂和低背景率。如果CRISPR活性随机低(例如由于差的gRNA效率)或CRISPR靶基因座处的突变可以被容忍(即使存在修复模板时产生假转化体),则细胞可以在CRISPR质粒摄取而没有修复DNA的情况下存活。
    3. 当菌落出现在-Ura平板上时,用移液器尖端收集其中的12个,并悬浮在50μl水中。这些悬浮液将被筛选以确认替换。酵母悬浮液可以储存在4°C,用于启动新的培养物长达2周。
      1. 通常情况下,1-3天后菌落会出现在ura平板上(图4)。在某些情况下,替代品会造成显着的适应缺陷,使得殖民地可能需要长达6天的时间才能出现,但我们还没有遇到成功转化的菌落需要6天以上才能生长的情况。


        图4.代表性测定结果。 当提供合适的修复模板(右侧板)时,酵母细胞从DSB致死性(中心板)中被拯救出来。左侧板是携带对照质粒的细胞的阴性对照,所述对照质粒具有用于估计所用酵母菌株的转化效率的相同选择标记(URA3)。

      2. 尿嘧啶缺失培养基将针对未能吸收CRISPR质粒(其赋予尿嘧啶原养型)的细胞进行选择,但是因为CRISPR质粒对细胞有毒,除非发生成功的替换(消除CRISPR靶基因座),只有具有替换的细胞预计基因座可以存活。然而,由于CRISPR系统的自发活动减少,CRISPR靶基因座中的突变(DiCarlo等人,2013)和设法存活CRISPR相关DSB的细胞,将会有背景速率以不携带正确基因组替换的假转化菌落形式存在。为了节省时间,我们建议收集几个转化体菌落并平行筛选。
      3. 为了简化这个过程(尤其是当多个替代品并行进行时),用移液枪尖端挑选菌落并手动将它们连接到多通道移液器上(图5)。多道移液器然后可用于将所有12个样品悬浮在一排小PCR管或96孔板中。


        图5.使用12通道移液器的集群采摘技术示范

  4. 通过PCR进行菌落筛选
    1. 设计确认PCR引物:应该选择引物对,使得正向引物与酵母UTR退火,而反向引物只与新基因CDS退火,而不与旧基因的ORF退火。因此,产物应该跨越外源序列和天然酵母基因组之间的连接点。酵母UTR引物应该优选不与同源区重叠。
      1. 理想情况下,产品大小应该很小,大约300 bp,以便更快速和更强大的PCR。
      2. 仅检查5'连接点就足够了,因为整合在基因的一端按预期进行的情况很少,但在另一端引入了人为因素。
      3. 如果需要,也可以通过使用仅与酵母ORF退火的反向引物来测试酵母ORF的缺失。然而,从这样的引物对缺乏产物不足以确认克隆,因为由于不相关的原因(例如细胞溶解差),反应很可能失败。
    2. 准备收获转化体的溶胞产物:将5μl每种酵母悬浮液与15μlzymolyase溶液混合。
    3. 在室温孵育裂解物30分钟,然后在37°C 15分钟和95°C 5分钟。
    4. 用确认引物设置20μl菌落PCR,并使用Accuprime Pfx作为聚合酶。使用1微升裂解物作为模板DNA。
      1. 我们发现其他聚合酶由于酵母裂解物中的杂质而表现不佳。
      2. 由于裂解物引入的杂质,菌落PCR可能会自发失败,从而导致假阴性。为了改善这个问题,可以对每个裂解物进行阳性对照PCR,其与确认PCR相同,但使用与基因组中不相关,未修饰基因座互补的引物。我们使用两个靶向酵母ERG13启动子500 bp片段的引物(正向CGAACTGGATGAGATGGCCG和反向CATGCTGCACCTTTTATAGTAATTTGGC)。
    5. 通过琼脂糖电泳检查菌落PCR产物。来自具有正确修饰的克隆的裂解产物应使用确认引物生成产物。背景假转化体(,例如,突变体)不会产生带。
      1. 来自确认引物的PCR产物是修复模板成功整合的充分证据。为了进一步验证,可以对基因座进行测序,但是我们发现在通过PCR确认的克隆中很少出现戏剧性的序列伪影,最常见的突变是单碱基对替换或插入缺失,这通常构成少数确认的克隆。 >
      2. 确认引物缺乏产品本身并不确定。在这种情况下,值得考虑更多的证据,例如阳性对照PCR是否起作用(如果不是,裂解可能失败)。

    6. 通过从水中原始酵母悬浮液开始新的培养物,可以繁殖已确认的克隆。

  5. CRISPR质粒的固化
    1. 确定YPD上已确认克隆的原始水悬浮液。
      CRISPR质粒是低拷贝的,可以在缺乏选择的情况下自发丢失。

    2. 从YPD平板挑选10个菌落,并在YPD和SD-Ura平板上贴上每个菌落。
    3. 孵育这两个板块,并收集只生长在YPD但不在SD-Ura上的贴片细胞。
      仍然携带CRISPR质粒的分离物将在尿嘧啶缺失培养基上生长,但那些丢失质粒的分离物不会。通常情况下,3天足以确认Ura原营养缺乏,但如果怀疑尿嘧啶辍学的生长缓慢,则可将孵化延长至6天以明确证实尿嘧啶辍学时无生长。
      也可以通过在5-氟乳清酸(FOA)平板上进行反选来使质粒固化(Boeke等人,1987)。然而,这种FOA方法可能会产生一些菌落,这些菌落没有治愈质粒,而是在Ura标记中获得了突变(从而继续表达gRNA)。因此,如果重要的是确保质粒的固化,而不是简单地废除Ura原养型,则不应使用FOA反选择(与在YPD和-Ura上复制贴片相反)。另一方面,如果只需要失去-Ura异质性,例如为了使用不同的Ura-可选择质粒进行后续转化,则FOA方法可以节省时间。

数据分析

该程序的数据分析需求很小。最重要的是,当使用Geneious设计gRNA序列时,需要选择具有高预测靶向活性的gRNA序列(由Geneious自动计算)。具有高预测活性的gRNA序列可能具有低实际活性,但与具有低预测活性的序列相比,它们不太可能表现出低活性。 gRNA靶位点的距离可以远离任一同源区域1kb而没有可察觉的负面结果,因此gRNA应主要基于高活性而非位置(只要它们位于两个同源臂之间)来选择。

笔记

  1. 我们发现,即使在具有高预测活性的gRNA中,也有一些将不能以足够的编辑效率诱导双链断裂。强烈建议,对于每个靶基因座,设计并测试几种gRNA,以确保至少有一种gRNA能成为足够好的DSB诱导物,用于基因组编辑。
  2. 如果给定的gRNA表现出显着的脱靶活性,则可能的结果是脱靶切割会杀死大部分转化的酵母细胞。在酵母中成功高效的基因组编辑依赖于在HR被拯救的靶位点(允许DSB的有效修复)和gRNA靶位点的消除(阻止进一步的切割)与DSB相关的致死性。在脱靶活动中,HR可能不会发生,因为没有提供与脱靶位点同源的修复模板,而且由于相同的原因,gRNA位点不会被消除。此外,我们建议的确认策略是,只有在正确的位点进行修复才能产生积极的结果。然而,确保选择的gRNA靶位点不在基因组的其他位置不发生切割是值得的。尽管组合的23bp靶序列不太可能在酵母基因组中出现多次,但我们建议使用诸如BLAT的工具确认候选gRNA位点仅出现在靶基因座中。
  3. gRNA靶标由20bp序列(其也将被包括在sgRNA序列中并成为Cas9复合体的一部分)组成,接着是3bp PAM序列(其在本方案中描述为采用NGG代替Cas9)。 PAM序列不成为gRNA的一部分,但它必须存在于目标基因组中才能发生Cas9切割。这可以通过尝试将gRNA序列与修复模板的序列比对来验证 - 通常,CRISPR活性将会非常低并且具有超过5个错配碱基对,尽管PAM中和PAM近侧的错配似乎具有更多显着性( Kuscu et al。,2014)。当用非常相似的序列进行替换时,难以找到目标基因座独有的良好gRNA位点,可以采用的一种策略是在修饰模板序列中引入同义突变,其改变PAM位点或PAM近端核苷酸。或者,最近的研究表明使用较短的gRNA可以增加特异性,因为8-17个PAM近端核苷酸对CRISPR靶标识别贡献不成比例(Xu等人,2017)。
  4. 酵母转化步骤有一些变化,取决于如何制备感受态细胞以及如何进行转化。最常见的情况是,在使用相同试剂的相同菌株转化之间,所得菌落的数量会有所不同,但通常这种变化将小于十倍。当转化产生相当数量的菌落(至少10个)时,筛选时没有发现它们是正确的克隆,只是重复转化不可能改善结果。增加正确克隆数量的最直接的途径是增加修复模板DNA的量,并产生具有更长同源性的修复模板DNA。
  5. 如果转化后没有出现菌落,原因可能是转化效率低下。在这种情况下,可以采取几个故障排除步骤(详细描述在Zymo EZ主管酵母试剂盒的文档中)。我们发现以下是有效的:
    1. 彻底涡旋感受态细胞和DNA的混合物。

    2. 转化时间更长(1.5小时而不是45分钟)。

    3. 包含更多单元格

    4. 。当冷冻一次(缓慢在-80°C)时,感受态细胞似乎表现略好于新鲜制备的细胞。
  6. 当CRISPR试剂和修复模板转化为酵母细胞时,所产生的转化菌落将针对靶向基因座有三种:
    1. 正确的转化体,其中包含修复模板的序列。
    2. 含有原始未经编辑序列的假转化体。
    3. 突变体。
  7. 在我们的实验中,我们发现前两个类别占主导地位,除非特别选择突变体。即使没有修复模板,大多数假转化体也不会是突变体。由于有效的人力资源系统。如果实验条件充足,那么编辑将以非常高的速度进行。因此,典型地,上面列出的三个类别中的前两个类别之间的比例将使得转化体大部分正确或全部错误。第三类或突变体,除非特别选择,否则我们发现在任何情况下都非常罕见。因此,很少有必要筛选大量的殖民地来确定编辑实验是否成功。然而,需要收集几个确认的克隆以最小化由人工产物引起的问题,例如由PCR过程中的错误引起的突变编辑序列(使用本文中描述的试剂和方案,我们发现具有突变编辑序列的克隆也是非常罕见的)。
  8. 选择含有单个氨基酸缺失培养基的酵母转化子通常是一个简单的过程,并且在1-2天内可以看到菌落。然而,偶尔基因组编辑过程本身或编辑后的序列可能导致产生的细胞发育缺陷。因此,如果没有菌落出现,将培养皿更长时间可以产生菌落。在我们观察到的最极端的情况下,菌落出现在尿嘧啶缺失培养基上需要6天,但后来通过PCR和测序证实了几个克隆;这些克隆在富含培养基(YPD)中的后续培养中始终表现出缓慢生长。
  9. 目标基因座和修复模板的一些组合可能导致转化后大和小酵母菌落的混合物。如果发生这种情况,通常最好为每个大小类别筛选足够数量的菌落。这可能是因为正确的编辑会造成生长缓慢的菌株,因此大的菌落是错误的,而小的菌落则具有所需的编辑。相反,如果期望的序列不干扰正常的生长,但是由NHEJ引起的突变,那么较大的克隆将趋向于是正确的克隆。我们观察了任何一种情况下人类化和细菌化各种基因座的例子。事先很难预测哪种情况对于给定的转化是明显的,因此筛选菌落并记录它们的大小通常更实际,并且确保每个大小在屏幕中充分表示。
  10. 当为菌落PCR筛选挑选菌落时,只需要少量细胞。最可能少至1,000个细胞就足以获得PCR产物。我们经常选择收集稍大一些的细胞以通过浊度目测确认它们在水中的悬浮液。然而,过多的细胞导致不完全的裂解和抑制菌落PCR。随着细胞团块大于1-2毫米,菌落PCR往往会失败。理想情况下,从菌落中收集的细胞只能形成一个直径为0.5毫米或更小的细小斑点。筛选时包括阳性对照PCR是有帮助的,以鉴定由于溶解差而不能产生PCR产物的样品。
    如果需要,可对这些样品重复裂解和PCR
  11. 有可能适应这里描述的协议同时取代多个基因。 Mo Clo工具包允许在相同的CRISPR质粒上克隆多达4种不同的gRNA表达盒;为此,如本文所述,将在pYTK050上捕获gRNA,但在第二次Golden Gate反应中,代替ConL1和ConRE质粒,第一种gRNA将用ConL1和ConR2克隆,第二种用ConL2和ConR3克隆,第三种与ConL3和ConR4,第四个与ConL4和ConRE(这一过程在Lee等人的详细解释,2015年)。所有这些盒式质粒将包括在最终的Golden Gate反应中以装配CRISPR质粒。然后,在酵母转化过程中,每种包含的gRNA的模板都需要进行共转化。然而,多次替换更依赖于高效的转换,分裂和修复,而不是单个替换,并且在实践中优化这些参数可能需要一些额外的工作。

食谱

  1. 发酵酶溶液(50ml)
    1. 称重9.11克D-山梨糖醇
    2. 溶于50毫升蒸馏去离子水中,制成1 M山梨糖醇和高压灭菌器
    3. 称0.25 g酵母裂解酶并溶于山梨糖醇溶液
    4. 分装并储存在-20°C
  2. 醋酸锂,100mM(40ml)
    1. 称量0.408克醋酸锂脱水物
    2. 溶解于40毫升蒸馏去离子水中。
    3. 过滤消毒(0.2μm过滤器)并在室温下保存
  3. LB培养基(1L)
    1. 称量25克LB粉末
    2. 对于固体培养基,添加15克琼脂
    3. 溶解于蒸馏水,去离子水中,总体积为1升

    4. 高压灭菌器将其冷却至60-70°C
    5. 倒入培养皿中,使培养基覆盖培养皿的可见区域
    6. 让板材在室温下冷却并固化,存放在4°C。
  4. YPD(1 L)
    1. 称量50克YPD粉末
    2. 对于固体培养基,加入20克琼脂
    3. 溶解于蒸馏水,去离子水中,总体积为1升

    4. 高压灭菌器将其冷却至60-70°C
    5. 倒入培养皿中,使培养基覆盖培养皿的可见区域
    6. 让板材在室温下冷却并固化,存放在4°C。
  5. SD-Ura(1 L)
    1. 称重1.5克酵母氮碱w / o氨基酸,5克硫酸铵,20克葡萄糖,2克SC-Ura辍学粉末
    2. 对于固体培养基,加入20克琼脂
    3. 溶解于蒸馏水,去离子水中,总体积为1升

    4. 高压灭菌器将其冷却至60-70°C
    5. 倒入培养皿中,使培养基覆盖培养皿的可见区域
    6. 让板材在室温下冷却并固化,存放在4°C。

致谢

这项工作得到了NIH(R21 GM119021,R01 HD085901,DP1 GM106408,R01 DK110520,R35 GM122480),陆军研究办公室(ARO)资助W911NF-12-1-0390和韦尔奇基金会(F-1515) EMM我们要感谢John Dueber&我们的同事们为我们的工作提供了极好的质粒工具包。作者宣称没有利益冲突。

参考

  1. Boeke,J. D.,Trueheart J.,Natsoulis G.和Fink G.R(1987)。 5-氟乳清酸作为酵母分子遗传学的选择剂方法Enzymol 154:164-75。
  2. Chen,X.,Niu,H.,Chung,W.H.,Zhu,Z.,Papusha,A.,Shim,E.Y.,Lee,S.E.,Sung,P.and Ira,G.(2011)。 细胞周期调控通过Cdk1依赖性Dna2磷酸化进行的DNA双链断裂末端切除 Nat Struct Mol Biol 18(9):1015-1019。
  3. DiCarlo,J.E.,Norville,J.E.,Mali,P.,Rios,X.,Aach,J.and Church,G.M。(2013)。使用CRISPR-Cas系统酿酒酵母中的基因组工程 Nucleic Acids Res 41(7):4336-4343。
  4. Heigwer F.,Kerr G.和Boutros M.(2014)。 E-CRISP:快速CRISPR目标网站识别 Nat Methods 11:122-123。
  5. Kachroo,A.H.,Laurent,J.M。,Akhmetov,A.,Szilagyi-Jones,M.,McWhite,C.D。,Zhao,A。和Marcotte,E.M。(2017)。 酵母基因的系统性细菌化确定了一个近乎普遍可交换的途径。 Elife 6.
  6. Kearse,M.,Moir,R.,Wilson,A.,Stones-Havas,S.,Cheung,M.,Sturrock,S.,Buxton,S.,Cooper,A.,Markowitz,S.,Duran,C Thierer,T.,Ashton,B.,Meintjes,P.和Drummond,A。(2012)。 Geneious Basic:用于组织和分析序列数据的集成和可扩展桌面软件平台。 a> Bioinformatics 28(12):1647-1649。
  7. 肯特,W. J.(2002)。 BLAT - 类似BLAST的对齐工具。 Genome Res 12(4):656-64。
  8. Kuscu,C.,Arslan,S.,Singh,R.,Thorpe,J.和Adli,M.(2014年)。 全基因组分析揭示了由Cas9核酸内切酶结合的脱靶位点的特征。 > Nat Biotechnol 32(7):677-83。
  9. Lee,M.E.,DeLoache,W.C.,Cervantes,B。和Dueber,J.E。(2015)。 一种用于模块化,多部件组装的高度表征的酵母工具包。 ACS Synth Biol 4(9):975-986。
  10. Mimitou,E.P。和Symington,L.S.(2009)。 DNA末端切除:许多核酸酶轻松工作 DNA修复(Amst ) 8(9):983-995。
  11. Xu,X.,Duan,D。和Chen,S.(2017)。 CRISPR-Cas9切割效率与目标-sgRNA折叠稳定性强关联:从物理机制到关闭 - 目标评估。 科学报告 7(143)。
  • English
  • 中文翻译
免责声明 × 为了向广大用户提供经翻译的内容,www.bio-protocol.org 采用人工翻译与计算机翻译结合的技术翻译了本文章。基于计算机的翻译质量再高,也不及 100% 的人工翻译的质量。为此,我们始终建议用户参考原始英文版本。 Bio-protocol., LLC对翻译版本的准确性不承担任何责任。
Copyright Akhmetov et al. This article is distributed under the terms of the Creative Commons Attribution License (CC BY 4.0).
引用: Readers should cite both the Bio-protocol article and the original research article where this protocol was used:
  1. Akhmetov, A., Laurent, J. M., Gollihar, J., Gardner, E. C., Garge, R. K., Ellington, A. D., Kachroo, A. H. and Marcotte, E. M. (2018). Single-step Precision Genome Editing in Yeast Using CRISPR-Cas9. Bio-protocol 8(6): e2765. DOI: 10.21769/BioProtoc.2765.
  2. Kachroo, A. H., Laurent, J. M., Akhmetov, A., Szilagyi-Jones, M., McWhite, C. D., Zhao, A. and Marcotte, E. M. (2017). Systematic bacterialization of yeast genes identifies a near-universally swappable pathway. Elife 6.
提问与回复

(提问前,请先登录)bio-protocol作为媒介平台,会将您的问题转发给作者,并将作者的回复发送至您的邮箱(在bio-protocol注册时所用的邮箱)。为了作者与用户间沟通流畅(作者能准确理解您所遇到的问题并给与正确的建议),我们鼓励用户用图片或者视频的形式来说明遇到的问题。由于本平台用Youtube储存、播放视频,作者需要谷歌账户来上传视频。

当遇到任务问题时,强烈推荐您提交相关数据(如截屏或视频)。由于Bio-protocol使用Youtube存储、播放视频,如需上传视频,您可能需要一个谷歌账号。