Engineering point mutant and epitope-tagged alleles in mice using Cas9 RNA-guided nuclease

Abstract

Mice with patient-associated point mutations are powerful tools to define causality of single nucleotide variants to disease states. Epitope tags enable immuno-based studies of genes for which no antibodies are available. These alleles are powerful tools for developmental, mechanistic, and translational research. The first step to generating these alleles is to identify within the target sequence – the orthologous sequence for PMs or the N- or C-terminus for epitope tags – appropriate Cas9 protospacer sequences. Subsequent steps include design and acquistition of a single-stranded oligonucleotide repair template, single-guide RNA (sgRNA) synthesis, collection of zygotes, microinjection or electroporation of zygotes with Cas9 mRNA or protein, sgRNA, and repair template followed by screening born mice for the presence of the desired sequence change. Quality control of mouse lines includes screening for random or multi-copy insertions of the repair template, and depending on sgRNA sequence, off-target mutations introduced by Cas9.

INTRODUCTION

The application of Cas9 RNA-guided nucleases to mammalian genome editing (Shen et al., 2013; H. Wang et al., 2013) (and reviewed in Hsu, Lander, & Zhang, 2014; Zhang, Wen, & Guo, 2014) has significantly simplified the generation of precision mouse models of human disease. Timelines to engineer point mutant and epitope-tagged alleles are now in the order of six months compared to the one to two-year timeframe using ES cell-based technologies. These timelines are further reduced when strain backgrounds not available as ES cells are used as embryo donors for manipulation. The mouse can now efficiently be used to evaluate variants of unknown significance and for mechanistic and pre-clinical studies of disease-associated alleles.

This chapter describes protocols for design and synthesis of single guide RNAs (sgRNAs) to direct Cas9 endonuclease to the target site (BASIC PROTOCOL 1 and ALTERNATE PROTOCOL 1) as well as the design of the oligonucleotide repair template needed to effect precise genome sequence changes (BASIC PROTOCOL 2). The synthesis of Cas9 mRNA has been described elsewhere (H. Wang et al., 2013) and with the plethora of commercial sources of Cas9 mRNA suitable for microinjection, is not described here. The composition of RNA and template mixes for mouse zygote microinjection (BASIC PROTOCOL 3) and electroporation (ALTERNATE PROTOCOL 2) are described.

Superovulation and mating mice for embryo production depends on the strain background and animal facility infrastructure. Microinjection is a specialist technique with practical details largely dependent on the specialized infrastructure available. Expert surgical techniques are required for embryo transfer into pseudopregnant recipients for gestation and parturition of mice from manipulated embryos. These techniques are usually performed by academic or commercial service providers rather than at individual laboratories. Furthermore, these techniques are well described elsewhere (Behringer, Gertsenstein, Nagy, & Nagy, 2014; Wefers et al., 2012) and so are not detailed here. We do describe the collection and preparation of 0.5-dpc zygotes for microinjection (BASIC PROTOCOL 4) and electroporation (ALTERNATE PROTCOL 3 and ALTERNATE PROTOCOL 4) delivery of Cas9 reagents and repair templates. Finally, we describe how to screen founders and quality control mice produced by founder breeding to establish high-quality lines (BASIC PROTOCOL 5).

STRATEGIC PLANNING (optional)

Typically, embryo donors must be acclimated to the light cycle of the animal facility before superovulation and mating. Hormone injections begin 3 days prior to the day of embryo harvest. In order to ensure mice are not ordered and/or superovulated prematurely, we synthesize and quality control the sgRNAs prior to ordering embryo donors. Depending on synthesis protocol used, this can take up to 2 days (BASIC PROTOCOL 1) or several hours (ALTERNATE PROTCOL 1). On the day of microinjection or electroporation, we prepare the Cas9 RNA or RNP mixes while the embryos are being collected and prepared for microinjection or electroporation. It is possible to prepare the Cas9 RNA mix ahead of time and store at -80°C for up to one week or the Cas9 RNP mix and store at 4°C overnight, but anecdotal evidence suggests that freshly prepared reagents perform better.

BASIC PROTOCOL 1

SINGLE GUIDE RNA DESIGN AND SYNTHESIS

There are several software tools available to assist with sgRNA design and selection. The choice of tool is largely dependent on user preference. Table 1 provides the names, URLs and published references for some commonly used tools. It is important to choose a tool that provides an accurate assessment of potential off-target sites, with some tools (e.g. Cas-OFFinder) providing more flexibility for off-target searching than others. The selection of the sgRNA will be constrained by the location of the desired sequence change or tag insertion. As a result, it may be necessary to select sgRNAs that are somewhat less specific. Ideally, sgRNAs should have ≥3 mismatches to all putative off-target sites. We generally consider off-target sites adjacent to both −NGG and −NAG PAMs. When an apparently less specific sgRNA sequence must be used due to constraints surrounding the target site, it is important to determine whether any of the off-target sites are linked to the target locus. Off-target sites that are unlinked can be screened and bred away during quality control and production of the mouse line. For both point mutations and epitope tags, an sgRNA with a predicted endonuclease cleavage site within ~20 bp is best, though we have had success introducing PMs up to 27 bp from the cleavage site.

Table 1.

Software tools for sgRNA design

Tool	URL	Ref.
Cas-Database	http://www.rgenome.net/cas-database/	(Park, Kim, & Bae, 2016)
Cas-Designer & CasOFF Finder	http://www.rgenome.net/cas-designer/ http://www.rgenome.net/cas-offinder/	(Bae, Park, & Kim, 2014)
CHOPCHOP	http://chopchop.cbu.uib.no/	(Labun, Montague, Gagnon, Thyme, & Valen, 2016; Montague, Cruz, Gagnon, Church, & Valen, 2014)
CRISPOR	http://crispor.tefor.net/	(Haeussler et al., 2016)
CRISPR Design	http://crispr.mit.edu/	(Hsu et al., 2013)
Guide Picker	https://www.deskgen.com/guide-picker/	(Hough et al., 2017a, 2017b)
WTSI Genome Editing	http://www.sanger.ac.uk/htgt/wge/	(Hodgkins et al., 2015)

Two protocols are presented for sgRNA synthesis – one uses a PCR-derived template (Bassett, Tibbit, Ponting, & Liu, 2013) followed by in vitro transcription using a commercial kit and the other uses a commercial kit for primer extension and in vitro transcription in a single reaction. Both approaches provide sufficient sgRNAs for multiple rounds of zygote microinjection or electroporation, though in our experience injection or electroporation of 80-160 embryos in one or two sessions is usually sufficient to obtain the desired point mutant or epitope-tagged allele. After synthesis and quantitation, synthesized RNA is subjected to quality control by gel electrophoresis (Masek, Vopalensky, Suchomelova, & Pospisek, 2005).

Please refer to Commentary for further details.

Materials

• gRNA FWD PCR primer, 10 μM

• gRNA REV PCR primer, 10 μM, or plasmid pX330, RNAse-free, 1 ng/μL [Addgene 42230]

• Nuclease-free PCR tubes

• 2X HiFi HotStart ReadyMix [KAPA KK2602]

• Nuclease-free water [e.g. Life Technologies 10977015]

• Monarch PCR & DNA Cleanup Kit [New England Biolabs T1030] or equivalent

• MEGAshortscript™ T7 Kit [Life Technologies AM1354M]

• RNA Clean & Concentrator-25 [Zymo Research R1017]

• Qubit RNA BR Assay Kit [Life Technologies Q10210]

• Fluorometer assay tubes [Life Technologies Q32856] or Axygen PCR-05-C tubes [VWR 10011-830]

• Agarose, RNAse-free [e.g. Life Technologies 16500500]

• 10X TAE, RNAse-free [e.g. Life Technologies AM9869]

• 2X RNA loading dye [e.g. New England Biolabs E2040S]

• Qubit fluorometer for RNA quantitation

• Gel electrophoresis apparatus, RNAse-free, for RNA integrity check

sgRNA design and primers

1. Select a guide RNA sequence (18-20 nt) using any of the many publicly available tools or by hand.

   General rules for sgRNA design for Cas9 are

   1. The sgRNA sequence should match a 20 nt target sequence in the DNA followed by a protospacer adjacent motif (PAM) of NGG. The NGG is absolutely required for cleavage but should NOT be included in the gRNA construct.
   2. For T7 promoter transcription, the 1^st 2 bases should be a GG dinucleotide, however appending GG to the 5’ end of the sequence seems to be well tolerated for most sgRNAs and may improve specificity (Cho et al., 2014) when a sequence starting with GG is not available.

2. Order the forward PCR oligonucleotide for synthesis of a T7 promoter driven sgRNA construct in the form of spacer-T7 promoter-gRNA sequence (N₂₀)-template anchor. The N’s in the sequence below represent the sgRNA spacer sequence. One or both of the two G’s in bold can be removed if the gRNA sequence starts with one or two G’s, respectively. The sequence of the common reverse primer is also shown.

There are two options for preparation of gRNA templates for in vitro synthesis. One uses pX330 as the PCR template with FWD and pX330-REV primers; one uses a reverse primer that encodes the complete 3’ sequence of the sgRNA, sgRNA-REV and FWD primers. Both methods work equally well, though the second obviates the need for purified plasmid. The sgRNA-REV primer should be PAGE purified.

FWD: TACGTGTAATACGACTCACTATAGG(N)_18-20gttttagagctagaaatagc

pX330-REV: AGCACCGACTCGGTGCCACT

sgRNA-REV: AAAAGCACCGACTCGGTGCCACTTTTTCAAGTTGATAACGGACTAGCCTTATTTTAACTTGCTATTTCTAGCTCTAAAAC

Preparation of sgRNA template

• 3Prepare the template synthesis PCR.

  For each sgRNA

  Component	Volume
  2X HiFi ReadyMix	20 μL	20 μL
  10 μM FWD	1.2 μL	2.0 μL
  10 μM pX330-REV	1.2 μL	-
  10 μM sgRNA-REV	-	2.0 μL
  DNAse/RNAse-free H2O	13.6 μL	16 μL
  1 ng/μL pX330	4 μL	-
  TOTAL	4 μL	4 μL

• 4
  Cycle on a “fast” PCR machine as follows:

  • With pX330-REV:
  • 95°C for 3 min
  • [98°C for 20 sec; 55°C for 15 sec + 1°C/cycle; 72°C for 10 sec] × 5 cycles
  • [98°C for 20 sec; 60°C for 15 sec; 72°C for 10 sec] × 20 cycles
  • 72 °C for 3 min
  • 8°C hold

  • With sgRNA-REV:
  • 95°C for 3 min
  • [98°C for 20 sec; 60°C for 15 sec; 72°C for 10 sec] × 35 cycles
  • 72 °C for 3 min
  • 8°C hold
• 5
  Purify the PCR fragment from the reaction using an appropriate fragment purification kit. When using the Monarch PCR and DNA Cleanup Kit add 60 μL of nuclease-free water and 500 μL of DNA binding buffer from the Monarch kit. Mix well.

  The PCR fragment can be assayed by gel electrophoresis before purification per step 11, if desired.

• 6With the column in a collection tube, load sample onto the column. Centrifuge at 16,000 ×g for 30 seconds to 1 minute. Discard flow through.
• 7
  Add 200 μL of wash buffer from the Monarch kit and centrifuge at 16,000 ×g for 30 seconds to 1 minute. Repeat with another 200 μL of wash buffer.

  Ensure wash buffer is diluted with ethanol per kit instructions.

• 8Discard flow through and centrifuge column at 16,000 ×g for 30 seconds to 1 minute.
• 9Transfer column to an RNAse-free microcentrifuge tube and add 20 μL of DNA elution buffer from the Monarch kit. Incubate at room temperature for 1 minute. Centrifuge at 50 ×g for 1 minute and then at 16,000 ×g for 1 minute. Discard column and cap the microcentrifuge tube.
• 10Quantitate template using a spectrophotometer (Gallagher, 2011).
• 11
  Electrophorese ~200 ng on a 1.5% agarose/1X TAE gel (Voytas, 2000) to ensure a single band of 125 bp.

  If multiple bands are seen, the 125-bp product should be gel purified using a gel purification kit (e.g. Machery-Nagel NucleoSpin Gel and PCR Purification Kit 740609). Quantitate the purified band per step 10 and continue with protocol.

• 12
  Store the DNA at -20°C until needed.

  This is a good pause point. The PCR product can be stored at -20°C indefinitely.

sgRNA synthesis by in vitro transcription

• 13Set up a 20-μL in vitro transcription reaction at room temperature using the MEGAshortscript™ T7 Kit with

  • Nuclease-free water to 20 μL final volume

  • 2 μL 10X T7 reaction buffer

  • 8 μL NTP premix (2 μL each of T7 ATP, CTP, GTP & UTP solutions)

  • 200 ng (~120 nM final concentration) of the template DNA prepared above

  • 2 μL T7 enzyme mix

• 14Mix contents well by gently flicking tube and collect contents by pulse centrifugation. Incubate the transcription reaction at 37°C for at least 2 hours and up to 4 hours.
• 15Add 1 μL of Turbo DNAse and incubate at 37°C for 15 minutes to remove DNA template.
• 16Bring volume to 50 μL by adding 30 μL of nuclease-free water.
• 17Purify sgRNA with the RNA Clean & Concentrator-25 by adding 100 μL Zymo RNA binding buffer. Mix well by pipetting.
• 18Add 150 μL of 95-100% ethanol. Mix well by pipetting.
• 19Apply mixture to a column in a collection tube. Centrifuge at 16,000 ×g for 30 seconds.
• 20Apply 400 μL of Zymo RNA Prep Buffer to the column and centrifuge at 16,000 ×g for 30 seconds. Discard flow through.
• 21Apply 700 μL of Zymo RNA Wash Buffer to the column and centrifuge at 16,00 ×g for 30 seconds. Discard flow through. Repeat wash with 400 μL of wash buffer. Discard flow through.
• 22Centrifuge column for 2 minutes at 16,000 ×g to dry the column.
• 23Transfer the column to a nuclease-free microcentrifuge tube. Apply 30 μL of nuclease-free water (provided in the kit) to the column. Centrifuge at 50 ×g for 1 minute then 16,000 ×g for 1 minute at room temperature. Discard column, cap tube, and place on ice until dilutions for quantitation and quality control are made. Store the sgRNA at -80°C.

sgRNA quantitation and quality control

• 24Prepare a 1:5 dilution of the sgRNA with 3 μL of sgRNA and 12 μL of nuclease-free water in a nuclease-free microcentrifuge tube. Vortex briefly to mix and pulse centrifuge to collect tube contents.
• 25Prepare Qubit RNA BR assay solution by adding 200 μL of Qubit RNA BR buffer and 1 μL of Qubit RNA BR reagent for each standard (2) and sgRNA to be assayed in a nuclease-free plastic tube. Vortex to mix.
• 26Into individual assay tubes aliquot 190 μL of Qubit RNA BR assay solution into each of two tubes for standards and 197 μL of Qubit RNA BR assay solution for each sgRNA to be assayed.
• 27
  Add 10 μL of each standard (provided with the Qubit buffers) and 3 μL of each sgRNA to individual tubes.

  We generally recover ~25 μL of 500-1000 ng/μL of sgRNA. Yields seem to vary depending on gRNA sequence but even in the low range, sufficient RNA is recovered from a single in vitro transcription reaction for several microinjections.

• 28
  Check RNA purity by UV absorbance at 230 nm, 260 nm and 280 nm (Gallagher, 2011). A260/A280 and A260/A230 ratios should be ≥2.0.

  While UV absorbance can be used to determine concentration of RNA, we find that using the concentrations determined with the Qubit give more consistent and reproducible results.

• 29
  Check RNA integrity by electrophoresis (Masek et al., 2005) on a RNAse-free 1.2% (w/v) agarose, 1X TAE gel after denaturing RNA in 2X RNA loading dye by heating at 65°C-70°C for 5-10 minutes and snap cooling in an ice water bath immediately prior to loading; or analyse on a Bioanalyzer, ScreenTape or similar instrument using a service provider (e.g. The Centre for Applied Genomics) or following manufacturer’s instructions. A single band of ~100 nt should be observed.

  Smearing of the band below the expected size indicates RNAse contamination. Clean workspace with RNAse Away and repeat template and RNA synthesis. Discrete bands smaller than the expected size suggest premature transcription termination during synthesis and/or incomplete extension of template during PCR. A ladder of RNA products can indicate incomplete denaturation prior to electrophoresis. Note that the expected 100-nt product will run slower than a 100-bp DNA band.

ALTERNATE PROTOCOL 1

Synthesis of sgRNA by primer extension and in vitro transcription

There are several commercially available kits for sgRNA synthesis. Commercial kits can offer a variety of advantages, including the ability to use shorter oligonucleotides which can reduce both costs and probability of errors during synthesis. They may also be more accessible to the novice user. We have used the EnGen^® sgRNA synthesis kit (New England Biosciences) with very good results and find its one-tube primer extension and in vitro transcription saves significant time for sgRNA synthesis.

Materials

• gRNA FWD primer, 1 μM in TE buffer or water

• Nuclease- and pyrogen-free PCR and microfuge tubes

• Nuclease-free water

• EnGen^® sgRNA synthesis kit [NEB E3322]

• RNA Clean & Concentrator-25 [Zymo Research R1017]

• Qubit RNA BR Assay Kit [Life Technologies Q10210]

• Fluorometer assay tubes [Life Technologies Q32856] or Axygen PCR-05-C tubes [VWR 10011-830]

• Agarose, RNAse-free [e.g. Life Technologies 16500500]

• 10X TAE, RNAse-free [e.g. Life Technologies AM9869]

• 2X RNA loading dye [e.g. New England Biolabs E2040S]

• Qubit fluorometer for RNA quantitation

• Gel electrophoresis apparatus, RNAse-free, for RNA integrity check

1. Following the design guidelines outlined in Step 1 of BASIC PROTOCOL 1, order the forward PCR oligonucleotide per kit instructions in the form of spacer-T7 promoter-gRNA spacer sequence-template anchor. The N’s in the sequence below represent the sgRNA spacer sequence. One or both of the two G’s in bold can be removed if the gRNA sequence starts with one or two G’s, respectively. The reverse primer is provided in the kit reaction mix.

   • FWD: TTCTAATACGACTCACTATAGG(N)_18-20GTTTTAGAGCTAGA

2. Thaw the FWD oligonucleotide and 2X EnGen^® sgRNA reaction mix on ice. Vortex well to mix and pulse centrifuge to collect contents. Leave enzyme mix at -20°C until use.

3. Assemble the reaction at room temperature by adding components in order to a nuclease-free PCR tube.

   • 3 μL nuclease-free water
   • 10 μL 2X EnGen^® sgRNA reaction mix
   • 5 μL FWD primer
   • 2 μL EnGen^® enzyme mix

4. Mix well by gently flicking tube and pulse centrifuge to collect contents.

5. Incubate at 37°C for 60 minutes.

   We find it convenient to perform these incubations in a PCR machine set for a 60-minute 37°C incubation, a 10°C hold for addition of the water and DNAse I, followed by a 25-minute 37°C incubation that is started by manually advancing the cycling program from the 10°C hold (step 6), followed by a final hold at 10°C.

6. Transfer tube to ice and add 30 μL of nuclease-free water and 2 μL of DNAse I provided in the kit. Incubate at 37°C for 25 minutes. Transfer tube to ice.

7. Clean up the RNA using the Zymo Clean & Concentrator-25 as described in the BASIC PROTOCOL 1, steps 17-23.

8. Resume protocol at step 24 of the BASIC PROTOCOL 1 for sgRNA quantitation and quality control.

BASIC PROTOCOL 2

DESIGN OF OLIGONUCLEOTIDE REPAIR TEMPLATE

To introduce point mutations or epitope tags, synthetic single-strand oligonucleotides are provided as a repair template for the endogenous cellular DNA repair machinery. The oligonucleotide repair template should be in the form of 50 to 70-nt homology arm-[desired change-endonuclease cut site]-50 to 70-nt homology arm. The desired change can be either 5’ or 3’ of the endonuclease cut site. An important consideration for repair template design is whether the sequence change eliminates the ability of the Cas9 RNP to re-cut the repaired allele and consequential NHEJ repair with the introduction of cis-located insertions or deletions (indels). To obviate the risk of re-cutting, we typically introduce one or more silent changes to mutate the PAM sequence from −NGG to any sequence except −NAG. If destruction of the PAM by silent mutation is not possible, 2 or more silent changes in the protospacer sequence “seed” region (11-bp adjacent to the PAM Hsu et al., 2013) could be used to reduce the risk of Cas9 re-cutting the repaired allele. Either the target or silent mutations may introduce or destroy a restriction enzyme site that could then be used for genotyping founders by restriction length polymorphism after PCR (see BASIC PROTCOL 6). In general, silent mutations should mimic the codon usage of the original codon as close as possible. It should be noted that there is evidence in the literature that silent mutations may have unintended consequences on transcription, splicing and/or translation (e.g. Nielsen et al., 2007; Biro, 2008; Doktor et al., 2014). This must be balanced against the possibility of Cas9 re-cutting and introducing cis-located indels. It may be possible to obtain the correct allele in the absence of silent mutations depending on the location and number of target changes relative to the protospacer sequence and/or PAM. In the absence of a restriction site polymorphism, derived alleles can be genotyped by real-time or digital-droplet PCR (Mazaika & Homsy, 2014) for the presence of the nucleotide variant.

Please refer to Commentary for further details.

Materials

Software for protein alignments

1. Align amino acid sequences of source orthologous protein (usually human) with the mouse orthologue.

2. Identify the target amino acid(s) and necessary nucleotide changes to effect the desired amino acid change(s).

   Epitope tags are usually inserted at the N- or C-terminus of proteins – either immediately following the START or immediately before the STOP codons, respectively. The location of the tag depends on what is known about protein domains, structure and function. Often a flexible linker (e.g. [GGSG]_1-3) is used to join the epitope tag to the target protein to reduce the risk of structural interference and maximize the availability of the epitope tag to antibodies.

3. Calculate the distance between the target nucleotide changes or site of insertion and the endonuclease cut site. This is considered the “middle” of the repair template.

4. Extend the repair template 50-70 nt on each side of the middle.

5. Order the synthetic oligonucleotide (e.g. IDT Ultramers) normalized at 100 μM in TE buffer, pH 8.

6. Before use, dilute the repair oligonucleotide to an appropriate stock concentration in TE or RNAse-free microinjection buffer.

   For microinjection, a working stock solution of 10 μM is suitable. For electroporation (ALTERNATE PROTCOL 2), working stock solutions of 2.5 μM or 5 μM are more appropriate.

BASIC PROTOCOL 3

PREPARATION OF RNA MIXTURE FOR ZYGOTE MICROINJECTION

Materials

• Nuclease- and pyrogen-free microcentrifuge tubes

• RNAse-free microinjection buffer (see recipe)

• Cas9 mRNA, 1 mg/mL [e.g. Life Technologies A29378, TriLink L-6112]

• sgRNA, ≥300 ng/μL

• Oligonucleotide repair template, 10 μM in TE or nuclease-free microinjection buffer (see recipe)

1. Calculate the amount of Cas9 mRNA, sgRNA and oligonucleotide repair template needed for 75 μL of the desired final concentration.

   Component	Final Concentration - pronuclear	Final Concentration - cytoplasmic
   Cas9 mRNA	30 ng/μL	100 ng/μL
   gRNA	10 ng/μL	10 ng/μL
   Repair template	10 ng/μL (~0.25 μM)	100 ng/μL (~2.5 μM)
   Microinjection buffer	to 75 μL	to 75 μL

2. Dilute the mRNA, gRNA and repair template in RNAse- and pyrogen-free microinjection buffer to a final volume of 75 μL in microinjection buffer.

3. Mix well by vortexing. Centrifuge at 20,000 ×g for 5 minutes at 4°C to pellet any particulates that may clog the injection needle.

4. Carefully remove the top 50 μL to a clean RNAse- and pyrogen-free microcentrifuge tube. Store the remaining 25 μL at -80°C as the pre-injection fraction for later quality control.

5. Store the microinjection mix on ice until injection.

   The microinjection mix may be stored at -80°C for up to 1 week prior to injection. If the microinjection mix clogs the needle and/or is too difficult to inject, it can be filtered to improve performance (see SUPPORT PROTOCOL 1).

SUPPORT PROTOCOL 1

Filtration of Cas9 RNA microinjection mixes

Microinjection mixes prepared as above can occasionally perform poorly as indicated by clogged microinjection needles or high embryo lysis rates. Filtration of the mixes through an RNAse-free filter column can improve injection performance.

Materials

Cas9 microinjection mix

Spin-X^® RNAse-free centrifuge tube filters, 0.22μm CA membrane [Corning 8160]

1. Apply 50 μL of RNAse-free microinjection buffer to the Spin-X filter in a collection tube. Centrifuge for 15 sec at 10,000 ×g at room temperature.

2. Discard flow through and collection tube. Place filter in a clean collection tube.

3. Apply diluted Cas9 mRNA mixture to the rinsed Spin-X filter. Centrifuge for 15 sec at 10,000 ×g at room temperature.

4. Remove filter. Centrifuge eluent at 20,000 ×g for 5 minutes at 4°C.

5. Carefully remove the top 50 μL of filtered RNA mix to a new RNAse-free 1.5 mL tube. This can be used for immediate microinjection or stored at -80°C for up to 1 week until needed. Store the remaining 25 μL of filtered RNA at -80°C (pre-injection fraction).

   This step is necessary to remove fibres from the Spin-X filter from the microinjection solution. Without this step, needles may get clogged with particulate matter, likely from the filter.

BASIC PROTOCOL 4

MOUSE ZYGOTE COLLECTION AND PREPARATION FOR MICROINJECTION

One advantage of Cas9 mutagenesis in zygotes is the ability to produce mutants in specific inbred backgrounds rather than relying on a few specific background strains available as embryonic stem cells. Zygotes from many inbred strains can be collected for microinjection with some variation for age and hormone doses for superovulation (Byers, Payson, & Taft, 2006; Luo et al., 2011). Below we describe the preparation of embryos for microinjection. As mentioned in the introduction, zygote microinjection requires specialized equipment and skill sets and are typically performed by service cores located in accredited animal facilities rather than at individual labs so the specifics of superovulation and mating, microinjection, and embryo transfer are not detailed here. These protocols are fully described elsewhere (Behringer et al., 2014; Wefers et al., 2012) and should be routine for any transgenic or mouse model service core facility.

Materials

• Superovulated and mated 0.5-dpc female mice as embryo donors

• Absorbent bench pads [e.g. VWR 82020-845]

• 26G needles

• Plastic Petri dishes, 35-mm, 100-mm [e.g. Falcon 351008, 353003]

• 10X Hyaluronidase (see recipe)

• M2 [Zenith Biotech ZFM2-100] with 4 mg/mL embryo-tested BSA [Sigma A3311]

• KSOM^AA [Zenith Biotech ZEKS-50] with 2 mg/mL embryo tested BSA [Sigma A3311]

• Paraffin oil [Zenith Biotech ZPOL-50]

• Incubator 37°C, 6% CO₂

• Stereomicroscope with transmitted light, 16X, 25X and 40X magnification

• Instruments (scissors, forceps) for oviduct dissection

• Embryo handling pipette with freshly pulled glass capillaries

1. Set up a 35-mm Petri dish with KSOM^AA microdrops under paraffin oil for embryo culture (Behringer et al., 2014). Pre-equilibrate at 37°C/6% CO₂ overnight.

2. On the morning of injection, euthanize the superovulated and mated females by cervical dislocation. Lay them on absorbent paper and wet their abdomen with 70% ethanol. Pinch the skin and make an incision at the midline with scissors. Holding the skin firmly above and below the incision, pull the skin apart and toward the head and tail until the abdomen is completely exposed and the fur is out of the way.

3. Using the watchmaker’s forceps and fine scissors, cut the body wall (peritoneum). Push the coils of the gut out of the way and locate the two horns of the uterus, the oviducts, and the ovaries.

4. Grasp the upper end of one of the uterine horns with fine forceps and gently pull the uterus, oviduct, ovary, and fat pad away from the body cavity. Use the closed tips of fine forceps or scissors to separate the membrane (the mesometrium) that connects the reproductive tract to the body wall and carries a prominent blood vessel near the oviduct.

5. Pull the oviduct, ovary, and fat pad with fine forceps and cut between the oviduct and ovary with fine scissors. Reposition the forceps and cut the uterus near the oviduct.

6. Transfer the oviduct and attached segment of uterus to a Petri dish containing M2 medium at room temperature. Continue with the second side and the remaining animals until all oviducts are dissected out.

7. Transfer one oviduct at a time into a new Petri dish with M2 medium and view through the stereomicroscope at 16X or 25X magnification to locate the ampulla where the zygotes are located.

8. Use watchmaker’s forceps to grasp the oviduct next to the infundibulum and hold it firmly on the bottom of the dish. Use another watchmaker’s forceps or a 26-gauge needle to tear the oviduct near ampulla, releasing the clutch of cumulus cells.

9. Using a P20 pipette with tip transfer all cumulus masses into a separate drop with ~100 μL of freshly diluted 0.3 mg/ml of hyaluronidase in M2. Incubate at room temperature until the cumulus cells separate from the zygotes.

   Pipette the cumulus mass up and down with an embryo handling pipette, if necessary. Do not leave the embryos in the hyaluronidase for more than a few minutes as it is harmful for the embryos.

10. Transfer the embryos through at least two fresh M2 drops to rinse off the hyaluronidase, cumulus cells, and debris.

11. Using stereomicroscope at 40X magnification (or higher if available), screen the embryos to select only fertilized zygotes by checking for the presence of the polar body (rotating the embryos with the pipette to visualize it) and pronuclei.

12. Transfer fertilized embryos in groups of 30-40 embryos into KSOM^AA microdrops and keep at 37°C/6% CO₂ until injection.

13. Proceed with either pronuclear or cytoplasmic injection of embryos using the microinjection mix prepared above.

14. Transfer injected embryos into 0.5-dpc pseudopregnant females on the same day as injection.

SUPPORT PROTOCOL 2

POST-INJECTION RNA INTEGRITY CHECK

Quality control of the microinjection mixtures can assist with trouble shooting. Pre- and post-injection fractions of the microinjection mixture are checked for integrity by agarose gel electrophoresis.

Materials

• Agarose, RNAse-free [e.g. Life Technologies 16500500]

• 10X TAE, RNAse-free [e.g. Life Technologies AM9869]

• 2X RNA loading dye [e.g. New England Biolabs E2040S]

• Gel electrophoresis apparatus, RNAse-free, for RNA integrity check

1. After microinjection, store remaining microinjection mix at -80°C as the post-injection fraction.

2. Test integrity by agarose gel electrophoresis (see step 28 of BASIC PROTOCOL 1) of the post-injection mixture in a lane adjacent to the pre-injection fraction from step 4 of BASIC PROTOCOL 3 or step 3 of SUPPORT PROTOCOL 1. Bioanalyzer, ScreenTape or equivalent analysis is an acceptable alternative to agarose gel electrophoresis.

3. Two bands, one at ~4.1 kb (Cas9 RNA) and one at ~100 nt (gRNA) and 140-200 nt (oligonucleotide) should be seen. Intensities should be equivalent between the two samples with no obvious degradation.

   The gRNA and oligonucleotide often overlap each other in agarose gels. In these cases, integrity of the Cas9 mRNA can be used as a reporter of gRNA integrity.

ALTERNATE PROTOCOL 2

PREPARATION OF CAS9 RIBONUCLEOPROTEIN COMPLEXES FOR ELECTROPORATION

Zygote microinjection is a specialist technique with practical limitations on throughput. Recently several groups have reported successful mutagenesis of mouse zygotes by delivering Cas9 ribonucleoprotein (RNP) by electroporation (S. Chen, Lee, Lee, Modzelewski, & He, 2016; Hashimoto, Yamashita, & Takemoto, 2016; W. Wang et al., 2016). Cas9 protein is pre-complexed with sgRNAs before delivery with an oligonucleotide repair template into zygotes using electroporation. Electroporation is higher throughput, is more efficient, and requires less technical expertise than microinjection, however it is still critical to have expertise in embryo handling and embryo transfer surgery.

Materials

• Nuclease- and pyrogen-free 0.5 mL microcentrifuge tubes

• 5X Cas9 RNP buffer (see recipe)

• Cas9 protein [e.g. PNABio CP01, IDT 1074182]

• sgRNA, ≥400 ng/μL

• Oligonucleotide repair template, 2.5 μM in TE buffer

1. Thaw gRNA and oligonucleotide on ice.

2. For each target, prepare 11 μL of 2X RNP in 1X Cas9 RNP buffer, when performing electroporation in a cuvette; 5 μL when using a slide for electroporation.

3. Add the first four components from the following table to a 0.5 mL nuclease- and pyrogen-free microfuge tube.

   Component	Final concentration	Notes
   Nuclease-free water	to 11 μL for cuvette electroporation; to 5 μL for slide electroporation	a
   5X Cas9 buffer	1X
   gRNA(s)	8.8 μM	b
   61 μM Cas9 protein	8 μM
   2.5 or 5 μM repair oligonucleotide	0.25 μM	c

   Notes:

   1. 10 μL of RNP mix is used in a cuvette; 2.5 μL for slide electroporation.
   2. gRNAs are provided in 10% molar excess to Cas9 protein. 6.6 μM of gRNA is ~217 ng/μL.
   3. We have tested a range of 10 ng/μL (~0.25 μM) to 20 ng/μL (~0.5 μM) of repair template concentrations without any significant difference in founder rates observed (unpublished data). Higher concentrations of oligonucleotide may provide better founder rates, but this is balanced against a predicted increased chance for random integration of the repair template.

4. Mix well by flicking the tube and pulse centrifuge to collect any droplets.

5. Incubate at 37°C for 10 minutes. Pulse centrifuge to collect droplets and transfer to ice.

6. Add the appropriate amount of oligonucleotide. Mix well by flicking the tube and pulse centrifuge to collect droplets. Place on ice until use.

ALTERNATE PROTOCOL 3

ZYGOTE ELECTROPORATION OF CAS9 RNP IN CUVETTES

Electroporation of zygotes can be done in 0.1-mm gap cuvettes or on an electrode slide. The former enables the use of standard equipment and consumables available in most laboratories. The drawback of using cuvettes is the relatively large volumes of buffer needed to recover the embryos necessitates additional time and effort to recover the embryos after electroporation for transfer. The zona pellucida of the embryo is weakened with Acid Tyrode’s prior to electroporation. After electroporation, embryos are transferred to pseudopregnant females for gestation and birth.

Materials

• Plastic Petri dishes, 35-mm and 100-mm [e.g. Falcon 351008, 353003]

• KSOM^AA [Zenith Biotech ZEKS-50] with 2 mg/mL embryo tested BSA [Sigma A3311]

• Paraffin oil [Zenith Biotech ZPOL-50]

• Incubator 37°C, 6% CO2

• 0.5-dpc zygotes (from BASIC PROTOCOL 4)

• Acid Tyrode’s [Sigma T1788]

• M2 [Zenith Biotech ZFM2-100] with 4 mg/mL embryo-tested BSA [Sigma A3311]

• Embryo handling pipette with freshly pulled glass capillaries

• Stereomicroscope with transmitted light, 16X, 25X and 40X magnifications

• Opti-MEM [Thermo Fisher Scientific 31985062]

• Cas9 RNP mix (from ALTERNATE PROTOCOL 2)

• 1-mm gap cuvettes [e.g. Cell Projects EP-201, BioRad 165-2089]

• BioRad Gene Pulser XCell™ Electroporator with ShockPod™ or similar

• Rainin Pipet-Lite XLS L20 [Mettler Toledo 17014392] and tips [Mettler Toldeo 17002930] for embryo retrieval

Weakening the zona pellucida

1. Set up a 35-mm Petri dish with KSOM^AA microdrops under paraffin oil for embryo culture (Behringer et al., 2014). Set up a second 35-mm Petri dish with KSOM^AA drops under paraffin oil for rinsing embryos after acid Tyrode’s treatment. Pre-equilibrate at 37°C/6% CO₂ overnight.

2. Place ~100 μl drop of acid Tyrode’s solution and several drops of M2 in a 100-mm Petri dish.

3. Place groups of embryos into separate M2 drops.

   If working with embryos located in KSOM^AA microdrops it is also possible to place the group of embryos directly into acid and then into an M2 drop.

4. Using 16X or 25X magnification, transfer one group of 30-40 embryos within a minimal amount of medium from M2 or KSOM^AA into the acid Tyrode’s drop and quickly spread them around.

   While it is possible to transfer larger groups of embryos into acid Tyrode’s, this extends the time that embryos are exposed to the acid Tyrode’s because it takes longer to collect all the embryos back into the pipette. The longer incubation can lead to loss of the zona pellucida and compromise embryo survival.

5. Immediately retrieve the embryos with the embryo transfer pipette in a minimal volume of acid and place them back into original M2 drop.

   Acid exposure should be no more than 7-10 seconds.

6. Proceed with zona pellucida weakening for the remaining sets of embryos.

7. Remove groups of embryos from the M2 drop and rinse through three to four drops of CO₂-equilibrated KSOM^AA.

8. Place the embryos back into KSOM^AA microdrops culture in the incubator until needed or proceed with electroporation as described below.

Zygote electroporation

• 9
  Set up the electroporator to deliver twelve 30V square wave pulses with 100-msec intervals between pulses.

  For the BioRad XCell this requires setting up a program that delivers 6 × 30V pulses with 100 msec intervals and executing that program twice in succession on each set of embryos.

• 10Set up a 100-mm Petri dish with ~50 μL of Opti-MEM in the centre and a 10-μL drop of Opti-MEM more peripherally.
• 11
  Collect one or two groups of embryos from KSOM^AA using an embryo handling pipet and rinse in the 50 μL of Opti-MEM.

  Electroporation of up to 80 embryos can be performed in the same cuvette.

• 12Transfer Opti-MEM rinsed embryos into the 10-μl drop of Opti-MEM medium minimizing media carry-over from the previous rinse.
• 13Immediately add 10 μl of Cas9 RNP mix prepared above into the 10-μl drop of Opti-MEM containing embryos.
• 14Using a Rainin L20 pipette with 20-μl tip, transfer the embryos in the 20-μl mix into the bottom of a 1-mm gap cuvette.
• 15Place the cuvette into the cuvette chamber and apply 12 × 30V square wave pulses with 100-msec intervals between pulses.
• 16Remove the cuvette from the chamber and carefully add 50 μL to 100 μl, depending on cuvette capacity, of CO₂-equilibrated KSOM^AA to the cuvette.
• 17With the Rainin L20 pipette set to 20 μL, gently but vigorously pipette up and down and remove ~18 μl of embryo suspension at a time to a Petri dish, avoiding air bubbles in the resulting drops.
• 18
  Retrieve the embryos from the KSOM^AA drops in the Petri dish using an embryo handling pipet and collect them into one KSOM^AA microdrop. Count the embryos to ensure the majority were recovered. Rinse cuvette with more KSOM^AA, if necessary.

  Typically, at least two rinses are required.

• 19Transfer the microdrop dish to the 37°C/6% CO₂ incubator until ready for embryo transfer surgery.

ALTERNATE PROTOCOL 4

ZYGOTE ELECTROPORATION OF CAS9 RNP ON AN ELECTRODE SLIDE

Using a slide for electroporation simplifies embryo manipulations and decreases the amount of Cas9 reagents used. However, slides are not disposable and must be thoroughly rinsed with nuclease-free water and dried between uses to minimize the chance of reagent carryover between embryo groups.

Materials

• Plastic Petri dishes, 35-mm and 100-mm [e.g. Falcon 351008, 353003]

• KSOM^AA [Zenith Biotech ZEKS-50] with 2 mg/mL embryo tested BSA [Sigma A3311]

• Paraffin oil [Zenith Biotech ZPOL-50]

• Incubator 37°C, 6% CO2

• 0.5-dpc zygotes (from BASIC PROTOCOL 4)

• Acid Tyrode’s [Sigma T1788]

• M2 [Zenith Biotech ZFM2-100] with 4 mg/mL embryo-tested BSA [Sigma A3311]

• Embryo handling pipette with freshly pulled glass capillaries

• Stereomicroscope with transmitted light, 16X, 25X and 40X magnification

• Opti-MEM [Thermo Fisher Scientific 31985062]

• 1-mm gap electrode slide [e.g. BEX via Protech International LF501PT1-10] and wires to attach to the electroporator

• BioRad Gene Pulser XCell™ Electroporator or similar

1. Prepare embryos by weakening the zona pellucida as in ALTERNATE PROTOCOL 3, steps 1-8.

2. Set up the electroporator to deliver twelve 30V square wave pulses with 100 msec intervals between pulses.

   For the BioRad XCell this requires setting up a program that delivers 6 × 30V pulses with 100 msec intervals and executing that program twice in succession on each set of embryos.

3. Set up two stereomicroscopes next to each other: one with the electrode slide connected to the electroporator inside a 100-mm Petri dish; one with a 100-mm Petri dish for preparation of embryos and RNP mixtures for electroporation. This latter Petri dish is the Embryo Prep dish.

   We find it convenient to perform this procedure with two adjacent microscopes, however a single microscope can be used.

4. Place ~50 μl of Opti-MEM medium in the middle of the Embryo Prep dish.

5. Near the periphery of the Embryo Prep dish, place one drop of M2 medium (~50 μl) for each group of embryos to be electroporated and three drops of KSOM^AA (~50 μL) for embryo rinses.

   We typically divide the dish into two halves to process two groups of embryos in quick succession.

6. Transfer one group of embryos into an M2 drop in the Embryo Prep dish.

   Up to 80 embryos may be processed for electroporation in a single slide.

7. Prepare the RNP mix for the first group of embryos by placing 2.5 μl of Opti-MEM medium next to the M2 drop with the embryos.

8. Add 2.5 μl of the Cas9 RNP mix to the 2.5 μL Opti-MEM drop. Transfer the 5 μl of Opti-MEM/RNP mix into the gap between the electrodes on the slide.

9. Rinse the group of embryos in the 50-μl Opti-MEM drop taking care to transfer a minimal volume of M2.

10. Transfer the rinsed embryos with minimal volume of Opti-MEM to the Opti-MEM/Cas9 RNP drop between the slide electrodes and line them up parallel to the electrodes. Work quickly to minimize evaporation of the mix in the slide.

11. Apply the 12 × 30V square wave pulses with 100-msec intervals between pulses.

12. Remove the embryos from the slide and return them to their original M2 drop in the Embryo Prep Petri dish.

13. To process additional groups of embryos, rinse the slide with copious amounts of nuclease-free water and wipe dry with a clean laboratory wipe (e.g. Kimwipe).

14. Proceed with preparing RNP mix and electroporating the next group of the embryos in the Embryo Prep dish as described above.

15. After all embryos have been treated and are back their M2 drops, rinse each group of embryos through at least three KSOM^AA drops and return them to the microdrops culture dish.

16. Incubate embryos at 37°C/6% CO₂ until embryo transfer into pseudopregnant recipients the same day.

BASIC PROTOCOL 5

SCREENING AND QUALITY CONTROL OF DERIVED MICE

Mice born from injected or electroporated embryos must be screened to identify founders. Founders are then backcrossed to the original strain (when using inbred lines) or to the desired background strain (when using F1 or other embryo donors). The derived N1 mice are subjected to quality control to ensure that the allele is the correct sequence and that no random or extra integrations of the repair template has occurred. N1 mice may also be screened for the presence of Cas9-induced off-target mutations. Once N1 mice have passed quality control, they can be used to establish the line.

Please refer to the Commentary for additional details.

Materials

• PCR primers ≥150-bp distal to and flanking the target change

• Tail DNA lysates from founder mice (Behringer et al., 2014)

• Reagents and equipment for PCR (Kramer & Coen, 2001), gel electrophoresis of PCR amplicons (Voytas, 2000) and sequencing of PCR amplicons (Dorit, Ohara, Hwang, Kim, & Blackshaw, 2001)

• Tail DNA purified from N1 mice (Liu & Harada, 2013)

• Reagents and equipment for subcloning PCR amplicons (e.g. TOPO TA cloning kit, ThermoFisher K457501 or (Finney, Nisson, & Rashtchian, 2001) and sequencing plasmids (Slatko, Albright, Tabor, & Ju, 1999; Slatko, Heinrich, Nixon, & Eckert, 2000)

• Amplification primers and hydrolysis probe for real-time PCR detection of the oligonucleotide repair template sequence that does not differentiate between wild-type and mutant sequences

• Reagents and equipment for quantitative real-time PCR (L. Ma & Chung, 2014) or digital droplet PCR (Mazaika & Homsy, 2014)

Founder screening and backcross

1. Identify (e.g. by ear punch) and obtain tissue biopsies from mice born from microinjected or electroporated embryos (Behringer et al., 2014).

2. Prepare genomic DNA (Behringer et al., 2014) and perform PCR (Kramer & Coen, 2001) to amplify the target sequence with ≥150 bp on each side flanking the target site.

   Crude DNA preparations are sufficient for most end-point PCR applications. However, if the target region is difficult to amplify, DNA can be purified using commercial kits or organic extraction prior to PCR.

3. If a restriction site polymorphism is predicted by the sequence change, digest the PCR products with the appropriate restriction enzyme and analyze by gel electrophoresis.

   Founders are likely to be mosaic and thus both wild-type and mutant alleles may be present – either from heterozygous or mosaic animals. If destruction of a restriction site is being used for genotyping, be aware that indels introduced by non-homologous end-joining repair of the DNA double-strand break rather than homology-directed repair could also destroy the restriction site.

4. Analyze amplicons by gel electrophoresis and sequence the PCR amplicons (Dorit et al., 2001) to confirm the presence of the desired point mutation or epitope tag.

   Founders may contain alleles with the templated change at the target site and indels. If using Sanger sequencing capillary-based fluorescent sequencing, chromatograms will show two peaks at the target site(s) – one with the base(s) for the wild-type allele and one with the base(s) for the mutant allele. For insertions of epitope tags, two overlapping chromatograms will be observed beginning at the target site. Overlapping chromatograms can be deconvoluted in silico using PolyPeak Parser (Hill et al., 2014). If using Sanger sequencing with autoradiography, two bands will appear on the gel at the target sites, similar to the appearance of two peaks on a sequence chromatogram. Manual deconvolution of sequences may be required for epitope tag insertions.

5. Set up founders with the desired sequence change to breed with the desired background strain.

N1 screening and quality control

• 6Identify (e.g. by ear punch) and obtain tissue biopsies from born N1 mice (Behringer et al., 2014).
• 7
  Purify genomic DNA (Liu & Harada, 2013) and perform PCR to amplify the target sequence.

  For quantitative PCR, inhibitors and variations in DNA amounts found in crude tissue lysates can compromise results. Using purified DNA, from commercial kits or organic extraction, is recommended for these applications. Using such DNA for the initial screen obviates the need to resample mice.

• 8Verify the integrity of the flanking genomic sequence and the presence of the targeted changes by sequencing the PCR amplicon and by restriction site polymorphism, if applicable.
• 9
  For epitope tag insertions, subclone the PCR products (Finney et al., 2001) and sequence the resultant plasmid inserts (Slatko et al., 1999; Slatko et al., 2000) to confirm the epitope tag has the correct sequence.

  Subcloning and sequencing is usually not necessary for single or a few nucleotide changes as they can be readily confirmed by direct PCR amplicon sequencing. However, it can be done, if desired.

• 10
  Design and acquire PCR primers and hydrolysis probes for real-time or digital droplet PCR (qrtPCR and ddPCR, respectively).

  Most companies that supply qrtPCR or ddPCR reagents also offer free or low-cost design services. Design principles for primers and probes for ddPCR are outlined elsewhere (Mazaika & Homsy, 2014). We typically use Tfrc or Tert, depending on primer and probe compatibility with the target gene, as our reference gene for both qrtPCR and ddPCR of genomic DNA.

• 11
  Determine the copy number of the target sequence using quantitative real-time PCR or ddPCR.

  By designing the amplification primers and hydrolysis probe to recognize both wild-type and mutant sequences within the repair template sequence, wild-type DNA can be used as the 2-copy (for autosomal genes) or 1-copy (for sex-linked genes in males) controls.

• 12
  Using the appropriate software for the qrtPCR or ddPCR platform, determine the copy number of the target sequence.

  Typically, we consider target copy numbers ranging from 1.8-2.2 for autosomal genes and 0.8-1.2 for sex-linked genes in males as wild type. Copy numbers outside of these ranges can be considered as provisional pass (1.6-2.4 for autosomal and 0.6-1.4 for sex-linked) or as failing (>2.4 for autosomal or >1.4 for sex-linked) QC.

REAGENTS AND SOLUTIONS

Microinjection buffer

• 1 M Tris, pH 7.6 [Sigma T2444]

• 0.5 M EDTA [Sigma E7889]

• Embryo-tested water [Sigma W1503]

• 0.22-μm surfactant-free cellulose acetate vacuum filter units [e.g. Nalgene^® 157-0020]

• Whatman Anotop Plus filters [VWR 28138-011]

• 20-mL luer lok sryinges [BD 309611]

• Sterile PETG serum vials [e.g. Nalgene^® 342032-0010] and sterile closures [e.g. Nalgene^® 342158-0021]

• Agarose, RNAse-free [e.g. Life Technologies 16500500]

• 10X TAE, RNAse-free [e.g. Life Technologies AM9869]

• 2X RNA loading dye [e.g. New England Biolabs E2040S]

• Gel electrophoresis apparatus, RNAse-free, for RNA integrity check

Prepare this solution in a biosafety cabinet to maintain sterility of stock solutions. For 200 mL of microinjection buffer add 2 mL of 1 M Tris, pH 7.6 (10 mM final) and 100 μL of 0.5 M EDTA (0.25 mM final) to 198 mL of embryo-tested water in a sterile PETG bottle. Mix well by swirling.

Set up a disposable 0.22-μm SFCA vacuum filter unit in the biosafety cabinet. Rinse the filter with 100 ml of the microinjection buffer by vacuum filtration. Discard buffer. Transfer the rinsed filter to a new collection bottle. Filter the remaining 100 mL of buffer and store in the sterile collection bottle for the next step.

Aspirate 20 mL of microinjection buffer into a gamma-radiation sterilized 20-mL luer lok syringe. Attach a 0.02 μM Whatman Anotop Plus filter to the filled syringe. Filter and discard 5 mL of microinjection buffer to rinse the filter. Filter the remaining buffer into two aliquots in sterile PETG serum vials. Close with a sterile closure. Store at room temperature.

Test buffer to confirm it is RNAse-free by incubating 2 μg of RNA (e.g. ThermoFisher Century™-Plus RNA Markers, AM7145) with 18 μg of 0.02 μm-filtered microinjection buffer at 37°C for ≥30 minutes. Assess integrity of RNA by agarose gel electrophoresis (Masek et al., 2005) or bioanalyzer adjacent to freshly prepared RNA. No degradation of RNA should be observed.

Test buffer to confirm it is suitable for microinjection. Inject ~100 0.5-dpc embryos (see BASIC PROTCOL 4) with 0.02 μm-filtered microinjection buffer. Transfer to pseudopregnant recipients in parallel with 50-100 embryos injected with previously QC’d buffer. Dissect mid-gestation and score number of implantation sites and number of live embryos. Implantation and live embryo rates should be within 10% of each other for both injected test and control embryos.

10X Hyaluronidase

• Hyaluronidase [Sigma H4272]

• M2 medium [Zenith Biotech ZFM2-100] containing 4 mg/mL embryo-tested BSA [Sigma A3311]

Dissolve hyaluronidase at 3 mg/mL in M2 medium containing 4 mg/mL embryo-tested BSA. Stock aliquots are stored at -20 °C. One aliquot is kept at 4 °C and freshly diluted in M2 to a working concentration of 0.3 mg/ml just before use to treat cumulus masses.

5X Cas9 RNP Buffer

• Potassium chloride (KCl) solution, 1M [Sigma 60142]

• HEPES buffer solution, 1M, pH 7.2-7.4 [Life Technologies, 15630]

• Embryo-tested water [Sigma W1503]

For 1 mL of 5X buffer, add 500 μL of 1M KCl, 100 μL of 1M Hepes, pH 7.2-7.4, and 400 μL of embryo-tested water to a nuclease- and pyrogen-free 1.5 mL microcentrifuge tube. Mix well. Store at room temperature. Centrifuge briefly to clear lid of any droplets before opening.

COMMENTARY

Background Information

The mouse has been the pre-eminent mammalian model system due to the ease with which its genome can be manipulated. Beginning with germline competent transgenic mice (Brinster et al., 1981; Costantini & Lacy, 1981; Gordon & Ruddle, 1981) and gaining sophistication with the isolation of pluripotent embryonic stem (ES) cells (Evans & Kaufman, 1981; Martin, 1981) and subsequent ability to engineer mutations by homologous recombination (Thomas & Capecchi, 1986), genetically modified mice have been used to dissect gene function in both normal and pathological development. The application of zinc-finger nucleases and TAL effector nucleases to modify the genome of mouse zygotes (Wefers et al., 2012) was a turning point in mouse genome engineering – opening the door to direct genome modification in any strain background, not just those for which high quality ES cells were available. The failure of these technologies to completely replace ES cell-based genome engineering was primarily due to the complexity of design for zinc finger nucleases and the necessity to engineer new proteins for each target (reviewed in Gaj, Gersbach, & Barbas, 2013; H. Kim & Kim, 2014). In 2013, the re-purposing of the CRISPR-Cas9 bacterial adaptive immune system to a programmable genome engineering tool (H. Wang et al., 2013; Yang et al., 2013) transformed mouse model production.

The Cas9 proteins are a family of RNA-guided nucleases (Ran et al., 2015) whose target specificity is dictated by a short spacer sequence (18-20 nt) within a larger guide RNA (gRNA) (Cong et al., 2013; Jinek et al., 2012) in the assembled ribonucleoprotein. Genomic target sites must be adjacent to a 3’ protospacer adjacent motif (PAM) which differs between Cas9 proteins (Ran et al., 2015). The most commonly used Cas9 for mammalian genome editing is a “humanized” version of S. pyogenes Cas9 (Cong et al., 2013) that uses an −NGG PAM, but lower cleavage efficiency is observed adjacent to non-cannoncial PAMs −NAG and −NGA (Hsu et al., 2013; Y. Zhang et al., 2014). This along with some tolerance for base mismatches between the spacer sequence and the genomic target contribute to off-target mutagenesis that has been observed for Cas9 (Hsu et al., 2013; Lin et al., 2014). Approaches developed to mitigate the risk of off-target mutagenesis include the use of Cas9 nickase (Shen et al., 2014), the use of truncated (17-18 nt) gRNAs (Fu, Sander, Reyon, Cascio, & Joung, 2014), appending two G residues (5’-GG) to the 5’ end of gRNAs (Cho et al., 2014), and using Cas9 ribonucleoprotein (RNP) instead of plasmid expression for mutagenesis (S. Kim, Kim, Cho, Kim, & Kim, 2014). Cas9 nickase is consistently less efficient at introducing mutations than Cas9 endonuclease, while the changes in efficiency using truncated or 5’-GG gRNAs varies by gRNA. To date, next generation sequencing of mouse mutants made with Cas9 have demonstrated a very low off-target mutagenesis rate, even when wild-type Cas9 endonuclease is used (Iyer et al., 2015; Mianne et al., 2016; Nakajima et al., 2016; Shen et al., 2014).

In the absence of a homologous repair template, double-strand breaks are most often repaired by non-homologous end-joining, which can result in the introduction of indels (Guirouilh-Barbat et al., 2004). However, homology-directed repair does occur at double-strand DNA breaks, such as those introduced by Cas9, when an appropriate homologous template is present, though at lower rates than NHEJ-mediated repair (Inui et al., 2014; Yang et al., 2013). The efficiency of homology-directed repair in mouse zygotes is sufficient for the recovery of point mutations and/or insertion of short sequence tags using synthetic oligonucleotides as repair templates (Inui et al., 2014; Yang et al., 2013). As a result, direct genome editing in zygotes using Cas9 and single-strand oligonucleotide repair templates has become the primary method for producing alleles with point mutations and short sequence tags.

Critical Parameters

sgRNA design

A critical aspect of sgRNA design is to select a spacer sequence that is specific, but specificity must sometimes be compromised in order to introduce a double-strand break near to the target site. In general, templated repair is more efficient closer to the double-strand break than further away (unpublished data and (F. Chen et al., 2011; Liang, Potter, Kumar, Ravinder, & Chesnut, 2017; Paquet et al., 2016). As noted above, there are methods to improve Cas9 specificity. We routinely use 5’-GG sgRNAs (Cho et al., 2014) and efficiencies to date (75-85%, depending on allele type, unpublished data) are largely using these types of sgRNAs. The 5’-GG also facilitate synthesis by in vitro transcription thereby removing the requirement for the selected sgRNA spacer sequence to begin with one or more G’s for T7-based in vitro transcription. In addition, using Cas9 mRNA for microinjection or Cas9 RNP for electroporation limits the exposure of the genome to Cas9 and is likely to reduce the risk of off-target mutagenesis (S. Kim et al., 2014).

A second consideration is that the efficiency with which double-strand breaks are introduced by Cas9 is likely correlated with stimulation of homology-directed repair (Liang et al., 2017). Several groups have developed algorithms to predict Cas9 RNP activity (e.g. Doench et al., 2014; Moreno-Mateos et al., 2015) with recent evidence presented that the prediction algorithm developed in zebrafish predicts activity in mouse embryos better than algorithms developed in cell lines (Haeussler et al., 2016). The choice of sgRNAs is often limited by the target site, but when a choice of sgRNAs is available, predicted activity can assist in selecting between two sgRNAs with equivalent specificity.

Oligonucleotide repair template design

The heuristics for oligonucleotide repair template design continue to evolve as experimental evidence for best practices accumulates. Asymmetric oligonucleotides designed to anneal to the strand first released from the Cas9 RNP-DNA complex – that is, oligonucleotides complementary to the PAM-containing strand – were suggested to improve point mutation recovery for sequence changes at or near the Cas9-induced double-strand break (Richardson, Ray, DeWitt, Curie, & Corn, 2016). More recent evidence suggests that 5’ modifications of single-stranded oligonucleotides, short double-stranded oligonucleotides (Liang et al., 2017), or changing the polarity of the single-strand oligonucleotide template depending on sequence change location relative to the double-strand break (Paix et al., 2017) may improve efficiencies. These systematic studies of oligonucleotide design heuristics are generally performed in human cell lines with occasional testing of identified optimal parameters in mouse zygotes, likely due to the costs associated with such systematic testing in zygotes. Using symmetric oligonucleotides with 50-70 nt homology arms complementary to the PAM-containing strand has worked well in our hands to date, but incorporating some of the newer design recommendations may enable us to improve templated sequence change efficiencies.

Cas9 mRNA and protein

The use of Cas9 mRNA versus RNP largely depends on method of delivery. We and others (e.g. (Mianne et al., 2016; Paix et al., 2017; Yang et al., 2013) have successfully used both Cas9 mRNA and Cas9 RNP for microinjection to introduce point mutations and short sequence tags. However, in our hands, Cas9 RNP is significantly more efficient when electroporation is used to deliver reagents to zygotes.

Modifications to the S. pyogenes Cas9 amino acid sequence has also been used to improve specificity. Several groups have developed high-fidelity Cas9 (e.g. Kleinstiver et al., 2016; Slaymaker et al., 2016) tested in cell lines and a proprietary version of high-fidelity Cas9 is now commercially available (Integrated DNA Technologies, 1078727). While these may improve specificity of DNA cleavage, they may not improve the integration of sequence-specific changes. Ma et al. (M. Ma et al., 2017) reported improvements in precise repair using an avidin-Cas9 fusion protein with biotinylated repair templates. Improvements in reporter allele generation using avidin-Cas9 and biotinylated repair templates have also been recently reported (Gu, Posfai, & Rossant, 2017). Thus, the incorporation of this technology may improve efficiencies for generating point mutations and epitope-tagged alleles, though this remains to be more broadly tested.

Reagent delivery

As iterated above, zygote microinjection requires specialized equipment and skill sets and are typically performed by service cores rather than at individual labs so the specifics of microinjection itself are not described here. Several excellent protocols are available elsewhere (Behringer et al., 2014; Wefers et al., 2012) for implementation of this technique. For simple alleles – exon deletions, NHEJ-mediated indels, point mutations, and epitope tags – electroporation is a straight-forward, efficient and scalable methodology for mouse line production. Our experience is that electroporation is more efficient (requires fewer embryos per founder obtained) than microinjection. This may be due to the increased survival (live born rate) of embryos manipulated by electroporation compared to microinjection.

Cas9 dose

Changes in Cas9 amounts (concentration delivered to zygotes) can affect efficiency of allele generation. Attention should be paid to the outcome of the experiment prior to either increasing or decreasing the amount of Cas9 (as either mRNA or RNP) delivered. For some mutations and/or gene targets, particularly those affecting viability and/or fertility, increasing the amount of Cas9 may actually decrease recovery rates, while for other mutations and/or targets, it may improve recovery rates. See Troubleshooting for more details.

Zona pellucida treatment

Acid Tyrode’s activity can vary between lots and it is very hard to standardize due to the variation of sizes of embryo handling pipettes and amounts of embryo handling media transferred with each group. Thus efforts are underway to remove this step from the electroporation protocol. Our early results suggest that increasing the pulse duration to 3 ms, enables production of indel and/or exon deletion alleles. Other groups have successfully eliminated this step from their protocols (Hashimoto & Takemoto, 2015; Horii et al., 2017).

Off-target mutations

Two types of off-target mutations may arise when engineering point mutations or epitope-tagged alleles in mice – random integration of the repair template or indels as a result of off-target Cas9 activity. Routine QC to detect random integration of repair templates is described (BASIC PROTOCOL 5). Detection of off-target mutagenesis by Cas9 can be done by (a) PCR and sequencing predicted target sites or (b) whole-genome sequencing. The prediction algorithms for Cas9 off-target activity are evolving (e.g. Abadi, Yan, Amar, & Mayrose, 2017) and unbiased approaches such as whole-genome sequencing are becoming more affordable. Since current evidence suggests that off-target mutagenesis rates are very low (Iyer et al., 2015; Mianne et al., 2016; Nakajima et al., 2016), unless indicated by poor sgRNA sequence specificity, backcrossing mice 3-4 generations may be sufficient to mitigate the risk of off-target mutations confounding phenotypic analyses. More studies with a broader range of gRNAs are needed to more fully assess risk and confirm the best approaches for identifying off-target mutagenesis after Cas9-mediated genome engineering in mice, however.

Mouse line maintenance

Lines should be maintained by backcrossing heterozygous mutants to the wild-type strain. Homozygous experimental cohorts should be produced by intercrossing heterozygous mice to generate wild-type controls and homozygotes from the same generation of intercrosses. Since genetic drift can result in as many as 100 new single nucleotide variants and 3-5 new indel mutations per generation (Uchimura et al., 2015), and commercial colonies provide mice from a range of generations (e.g. Charles River, 2016; Taft, Davisson, & Wiles, 2006), it is not appropriate to maintain control and experimental lines separately or to order commercial stock as controls.

Troubleshooting

Efficiencies vary from locus to locus and can range from 1-40% of live born pups carrying the desired templated sequence change and about double that number of indel alleles recovered. The most common causes of failure are (a) poor reagent quality as indicated RNA degradation upon post-injection QC, (b) a low activity Cas9-sgRNA combination, or (c) effects of the mutation(s) on viability and/or fertility. If reagents pass post-injection QC, yet the desired mutation is not recovered after one round of microinjection or electroporation, assess (a) the number of pups born and (b) the level of Cas9 activity as indicated by the recovery of indel alleles at the target site. If good birth rates (≥20% of embryos delivered after microinjection or ≥40% for electroporation) and indel rates (≥25% of alleles mutated) are observed, but no templated repair, this indicates the Cas9 RNP activity is good. Increasing the amount of oligonucleotide in the microinjection or electroporation mix for repeated attempts may be effective. If good birth rates occur, but few or no indel alleles are identified, as a first effort increase the amount of Cas9 or if available, try a different gRNA. Removing the non-templated 5’-GG may also help, if the spacer sequence begins with one of more G’s. If few or no pups are born, try reducing the amount of Cas9 used to reduce possible toxic effects of the gene modification by decreasing the likelihood of compound heterozygotes or homozygotes and/or increasing mosaicism with lower doses of Cas9.

Statistical Analyses

n/a

Understanding Results

Among mice born, 1-40% of mice should have the desired sequence change, with about 2-3 fold the number of indel alleles identified. In general, indel alleles are directly correlated with Cas9 activity whereas templated sequence changes are dependent on the activity of the endogenous cellular DNA repair machine. The variability is a result of several factors including accessibility of the locus to Cas9 activity, Cas9-sgRNA efficiency, and the activity of the cellular DNA repair machinery. The Troubleshooting section above provides possible explanations for negative results and ways to attempt to rescue failed projects.

Time Considerations

Once reagents arrive (e.g. synthetic oligonucleotides), sgRNA synthesis and QC takes 1-2 days. Timing of mouse orders for embryo donors, superovulation, mating and embryo harvest may vary slightly between animal facilities. Preparation of microinjection or electroporation mixes take less than one hour and can reasonably be completed during the time it takes to collect and prepare embryos for microinjection or electroporation. Microinjection or electroporation followed by embryo transfer surgery can take 8-10 hours, depending on the number of embryos manipulated and transferred. Once embryo transfer is complete, the remaining experimental timeline is largely is dependent on the reproductive biology of the mouse - ~3 weeks until parturition, ~2 weeks until born mice can be sampled for genotyping, ~6 weeks after than until founders can be bred for germline transmission testing. Genotyping and quality control can generally be completed in one day, depending on the number of mice being assayed, once reagents are in hand.