Skip to main content

Thank you for visiting You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

Creation of CRISPR-based germline-genome-engineered mice without ex vivo handling of zygotes by i-GONAD


Methods to create genetically engineered mice involve three major steps: harvesting embryos from one set of females, microinjection of reagents into embryos ex vivo and their surgical transfer to another set of females. Although tedious, these methods have been used for more than three decades to create mouse models. We recently developed a method named GONAD (genome editing via oviductal nucleic acids delivery), which bypasses these steps. GONAD involves injection of CRISPR components (Cas9 mRNA and guide RNA (gRNA)) into the oviducts of pregnant females 1.5 d post conception, followed by in vivo electroporation to deliver the components into the zygotes in situ. Using GONAD, we demonstrated that target genes can be disrupted and analyzed at different stages of mouse embryonic development. Subsequently, we developed improved GONAD (i-GONAD) by delivering CRISPR ribonucleoproteins (RNPs; Cas9 protein or Cpf1 protein and gRNA) into day-0.7 pregnant mice, which made it suitable for routine generation of knockout and large-deletion mouse models. i-GONAD can also generate knock-in models containing up to 1-kb inserts when single-stranded DNA (ssDNA) repair templates are supplied. i-GONAD offers other advantages: it does not require vasectomized males and pseudo-pregnant females, the females used for i-GONAD are not sacrificed and can be used for other experiments, it can be easily adopted in laboratories lacking sophisticated microinjection equipment, and can be implemented by researchers skilled in small-animal surgery but lacking embryo-handling skills. Here, we provide a step-by-step protocol for establishing the i-GONAD method. The protocol takes 6 weeks to generate the founder mice.


Genetically engineered animal models can be broadly categorized into two types: transgenic and gene-targeted animals. Transgenic animals are typically a result of insertion of foreign DNA cassettes into random locations in the genome, whereas gene-targeted animals are created by modification (either disruption or insertion of new sequences) at a specific locus in the genome. The traditional methods used to generate transgenic and gene-targeted animals also differ slightly. Transgenic animals are generated by injecting a recombinant DNA cassette into early-stage embryos called zygotes/pronuclei (the stage before the fusion of the male and female nuclei in the fertilized egg), whereas gene-targeted mice are produced by injecting embryonic stem (ES) cells that have been previously modified to contain the genetically engineered mutation into advanced-stage embryos called blastocysts. In regard to animal experimentation protocols, both transgenic and gene-targeted animal generation protocols involve three common steps: isolation of embryos (zygotes/blastocysts) from one set of pregnant females, called donors; microinjection of gene-modification reagents (DNA or ES cells) into embryos, followed by brief culturing of the injected embryos ex vivo; and surgical transfer of the manipulated embryos to another set of females (pseudo-pregnant/recipient females). These standard methods were developed more than three decades ago1,2,3,4 and they have served to generate several thousands of genetically engineered mouse models5. Because of the importance of genetically engineered mouse models for biomedical research, major funding organizations worldwide joined their efforts to develop knockout and conditional knockout ES cell clones for nearly 90% of protein-coding mouse genes, operating through the International Knockout Mouse Consortium (IKMC) and the International Mouse Phenotyping Consortium (IMPC)6.

The advent of programmable nucleases such as zinc-finger nucleases (ZFNs), transcription activator–like effector nucleases (TALENs) and CRISPR has dramatically affected how the decades-old genome-engineering technologies are practiced. ZFNs7 and TALENs8 are nuclease enzymes consisting of two domains: a custom-designed DNA binding domain to bind a specific nucleotide sequence in the genome and a nuclease domain from FokI endonuclease that cleaves the target DNA sequence. By contrast, the CRISPR system constitutes an RNA component, termed gRNA, that recognizes a specific sequence in the genome and recruits a Cas nuclease protein to the site to generate a double-strand break (DSB) at this target site. The cellular repair processes for fixing the DSB are utilized for genome editing, to create simple knockouts via frameshift mutations or to create a point mutation via a homology-directed repair (HDR) using a repair DNA template9. The CRIPSR system offers many advantages over ZFNs and TALENs; it is versatile, simple to design, rapid and less expensive.

The CRISPR tool has brought about several paradigm shifts in mouse genome-manipulation technologies10, enabling the long-practiced methods to be bypassed. The two most notable advances are first, that there is no need to use ES cells for gene targeting, and second, that there is no need for microinjection to deliver reagents into the embryos. With the use of CRISPR tool, gene targeting can be performed directly in zygotes via injection of CRISPR reagents, without requiring modification of the gene in ES cells and injection of the ES cells into embryos11. The CRISPR genome-editing reagents can be introduced into several dozens of zygotes simultaneously by in vitro electroporation rather than microinjection of zygotes one at a time12,13,14,15,16. Thus, the electroporation approach has obviated the need for microinjection—one of the most tedious steps of genetic engineering.

Even though the CRISPR-based approaches have dramatically simplified genetic engineering methods, they still require proficiency in performing complex procedures involving ex vivo handling of embryos. Those procedures include isolation of zygotes from dissected oviducts, dispersing them from clumps of cumulus cells, transferring them by mouth pipetting into different sets of culture dishes and then onto microinjection/electroporation slides, and their subsequent transfer to recipient females (following microinjection/electroporation) to allow them to develop further. Mastery of such ex vivo embryo-handling skills and embryo-transfer surgeries typically takes several months of practice. Furthermore, regularly performing such procedures is critical to maintaining the skill sets. Because ex vivo handling of zygotes is an inevitable step for either microinjection- or ex vivo electroporation–based approaches, devising a technical improvement that could completely avoid ex vivo embryo handling could benefit many researchers wanting to create mouse models of their own. To our knowledge, no technical advances to CRISPR-based approaches have been able to bypass the ex vivo embryo-handling steps before now.

Until very recently, the ES cell reagents were inevitably required for creating gene-targeted mouse models, and microinjection was the predominant way to introduce transgenic constructs or gene-modified ES cells into mouse embryos. We recently developed GONAD, a method that not only bypasses both the use of ES cells and microinjection, but can also dispense with ex vivo handling of embryos completely. The GONAD method enables creation of gene-targeted models without the need to perform all three major steps of transgenesis, namely, isolation, microinjection (or ex vivo electroporation) and surgical transfer of embryos (Fig. 1)10,17. We also made several improvements to GONAD (i-GONAD), making it suitable for creating knockout, knock-in and large-deletion models18. The efficiency of i-GONAD was equivalent to that of microinjection-based CRISPR methods (e.g., 49% and 52% knock-in efficiency in the i-GONAD and microinjection-based methods, respectively)18. Furthermore, i-GONAD does not require sophisticated micromanipulation equipment and specialized skills among the technicians to perform these steps. Therefore, it promises to relieve the bottlenecks of the long-used transgenesis protocols. i-GONAD uses substantially fewer animals because the females used for the i-GONAD procedure will not be sacrificed and thus the method does not require recipient (pseudo-pregnant) females and vasectomized males. Sacrificing of donor females is inevitable in traditional methods of isolating embryos. In addition, vasectomized males are necessary in traditional methods to mate with (another set of) females to generate embryo-recipient females.

Fig. 1: Overview of the i-GONAD workflow.

(Left) The procedure consists of three main stages: (i) preparation of CRISPR components such as gRNA and repair template (Steps 1–3); (ii) the main i-GONAD procedure (Steps 4–32); and (iii) genotyping analyses of offspring (Steps 33–41). Steps 1–3: the CRISPR reagents used in i-GONAD have three components: gRNA, Cas9 protein and a repair DNA template. The gRNA consists of two components: crRNA and tracrRNA. (Top, center) The crRNA consists of two regions: a 20-nt region complementary to the genomic target sequence (orange rectangle) and a region complementary (orange line) to the tracrRNA (light blue wavy line). The genomic region complementary to the crRNA is shown as two separated strands (black). Cas9 protein is denoted by dotted orange ovals. (Top, right) Two types of repair template DNA are used in i-GONAD: long ssDNA prepared from plasmid DNA through an in vitro transcription and reverse transcription (ivTRT) method, and single-stranded oligodeoxynucleotides (ssODNs). The central red regions in the repair DNA templates (ssDNA or ssODN) correspond to the genomic modification introduced by the repair template, whereas the blue lines correspond to the homology arms complementary to the genomic cleavage sites. Steps 4–32: the main i-GONAD procedure (bottom right). This is further divided into four sub-stages: preparation of pregnant females, preparation of the genome-editing mixture, surgical procedure and injection, and in vivo electroporation. Steps 33–41: genotyping analyses of offspring. The offspring of the i-GONAD method can be analyzed for the intended genomic modification at the two-cell stage (2-d protocol; see Box 2), fetal stage or newborn stage.

Development of i-GONAD

We hypothesized that if genome-editing components could be injected into the oviductal lumen, followed by subsequent electroporation of the tissue to deliver them into the developing zygotes within the oviduct, the whole process of ex vivo handling of zygotes, which is practiced in traditional transgenesis methods, can be avoided. Such a method would also save lives of embryo-donor females because the mice are inevitably euthanized in order to isolate the embryos from them in the traditional methods.

In the first stage of testing the above hypothesis, we targeted the EGFP locus in a mouse line containing a single-copy EGFP transgene in the ROSA26 locus19. We injected Cas9 mRNA and a gRNA targeting EGFP into the oviduct of embryonic day (E)1.5 pregnant females, and observed successful disruption of the EGFP in the mid-gestation embryos17. This preliminary result prompted us to systematically test several parameters in order to develop the GONAD method for routine generation of knockout, knock-in and large-deletion animal models. The most important parameters we evaluated during the development of GONAD, which resulted in i-GONAD, were stage of pregnancy (day 1.5, day 0.7 and day 0.4), different electroporators and electroporation conditions, and different CRISPR reagent formats (Cas9 mRNA, Cas9 protein or Cpf1 protein)18.

Stage of pregnancy

After mating of females at the estrus stage, ovulated eggs are transported to the ampulla, an intermediate dilated oviduct region proximal to the infundibulum, where fertilization occurs. We evaluated several time points post mating (from 0.4 d up to 1.5 d) to determine the optimal time point for generating gene-modified animals. We also assessed the efficiencies of reagent delivery by injecting mRNA coding for EGFP or 3-kDa tetramethylrhodamine–labeled dextran (as a marker for efficient electroporation of the zygotes and the oviductal tissues). These series of experiments indicated that the day 0.7 stage of pregnancy produces optimal results; the scientific rationale for this observation is described as follows.

On the basis of standard protocols20, day 0.5 of gestation is considered to correspond to 12:00 pm on the day vaginal plugs are observed in mated females. Embryos at day 0.4 of gestation are at the early zygote (1-cell) stage, and they are tightly surrounded by cumulus cells and exist within a largely expanded ampulla. At day 0.7 of gestation (4 pm), embryos will be at the late zygote stage, and they will be almost free of cumulus cells. Day 1.5 is less optimal, compared to day 0.7, because (i) the zygotes will be at the two-cell stage, which can increase mosaicism, and the overall efficiency of obtaining the gene-edited allele is lower than when the procedure is performed at the 1-cell stage; and (ii) the ampulla region is no longer dilated and the zygotes will have traversed further toward the uterus. Taken together, upon systematic evaluation of the parameters discussed above, we observed that the day 0.7 stage would be most favorable because (i) it provides the optimal stage of embryonic development (1-cell zygote stage); (ii) the zygotes are no longer tightly covered with cumulus cells, making them accessible for achieving higher electroporation efficiencies; and (iii) it is relatively easy to identify the dilated ampulla to expose it for the i-GONAD procedure.

Electroporators and electroporation conditions

We tested electroporators from three different vendors that can deliver square-wave pulses. In our initial experiments, we used a BTX 820 electroporator, which produced successful results; however, the model is no longer available17. Subsequently we tested electroporators from two different manufacturers, BEX and NepaGene, which produced comparable results18. We tested various electroporation conditions for these electroporators to define the optimal conditions for obtaining higher genome-editing efficiencies (described in detail in Step 25).

The CRISPR reagent formats

CRISPR reagents constitute gRNA, Cas9 or Cpf1 (also known as Cas12a) (mRNA or protein) and an optional repair ssDNA template. We used in vitro–transcribed single guide RNA (sgRNA)11 and Cas9 mRNA in our preliminary experiments with good results. Later, we compared these reagent formats with RNP delivery consisting of fully synthetic gRNAs (as CRISPR RNA (crRNA) + trans-activating crRNA (tracrRNA)) and Cas9 protein21, and observed that the RNP formats were much more efficient in genome editing18. We recently developed a microinjection-based CRISPR method, called Easi-CRISPR, to create knock-in and conditional knockout mouse models. Easi-CRISPR uses long ssDNAs as repair templates along with CRISPR RNPs21,22,23. We then demonstrated that long single-stranded repair template donors (one of the Easi-CRISPR components) can be used in combination with i-GONAD to create knock-in models containing insertions of up to 1 kb.

Taken together, the i-GONAD protocol performed at day 0.7 of pregnancy, by using streamlined electroporation conditions along with CRISPR RNPs (with the optional inclusion of up to 200-base-long single-stranded oligodeoxynucleotides (ssODNs) or longer ssDNA repair templates), provides high-efficiency germline-genome-engineering protocols for creating genome-modified animals. Further, these procedures completely bypass ex vivo zygote-handling steps, which are inevitable in the traditional genome-modification protocols practiced for more than three decades.

Applications of i-GONAD

The method can be used for routine generation of various kinds of genome-edited animals, including gene-disruption, point-mutation, deletion (from a few bases to tens of kilobases) and larger-fragment-insertion (up to 1 kb) models. We have generated mouse and rat models with genomic deletions ranging from a few bases up to 16.2 kb, as well as insertions up to 1 kb18,24,25. i-GONAD can potentially be used for creating genome-edited animal models in other species, such as hamster, rabbit and pig. Our experiments involving the delivery of EGFP mRNA demonstrate that the approach can also be used for delivering nucleic acids and CRISPR reagents to the oviductal tissue itself to achieve gene editing of the oviductal (somatic) cells. As an extension of this somatic cell gene-editing (SCGE) approach, the method can be adapted for delivering reagents to other somatic tissues, by altering the surgical procedures, and by using suitable electrodes and/or electroporation conditions (M.O. and H.M., unpublished data).

Comparison of i-GONAD to traditional animal transgenesis methods

The three major steps of traditional transgenesis methods—i.e., isolation of embryos, introduction of transgenic DNA/modified ES cells into them, and transfer to another set of females—have largely remained unchanged since they were first developed in the 1980s5. Because these steps require sophisticated equipment as well as highly skilled personnel, animal transgenesis technologies are typically performed at centralized core facilities in most organizations. In direct comparison, i-GONAD does not rely on any of the steps involving ex vivo handling of embryos. Furthermore, it also offers several advantages over the traditional methods18. First, i-GONAD offers gene-editing efficiencies comparable to those of the most commonly used traditional method, embryo microinjection. Second, i-GONAD does not require the second set of pregnant females (embryo recipients) and the vasectomized males to generate recipient females. Third, because the females used for i-GONAD are not sacrificed (to collect their eggs), they can be used for other experiments. This feature would be advantageous when i-GONAD-treated mice are precious. For example, some genetically modified mouse line with just one or two mice are available (that the researchers received from another laboratory or were newly created) and the researchers do not wish to sacrifice the animals for reasons such as pathogen testing requirements of the animal facility. Fourth, i-GONAD can be easily adopted at laboratories lacking sophisticated microinjection equipment. Last, the method can be performed by researchers without embryo-handling skills.

Experimental design

i-GONAD involves three major stages: (i) preparation of CRISPR reagents, (ii) the main i-GONAD procedure and (iii) analysis of offspring (Fig. 1). Note that the main i-GONAD procedure (Steps 4–32) is the major focus of this paper and is covered in detail, whereas stages (i; Steps 1–3) and (iii; Steps 33–41) are quite similar to those of the microinjection-based genome editing procedure described in our previous paper23. The minor differences or considerations in stages (i) and (iii) that are relevant to the i-GONAD method are described below.

Preparation of CRISPR reagents

Animal genome-engineering designs fall into one of two major categories: the first (two-component CRISPR) requires only gRNA(s) and Cas9, whereas the second (three-component CRISPR) requires gRNA(s), Cas9 and a repair DNA template. The repair DNA template consists of the intended changes in the genome, such as substitutions, insertion of new sequences encoding protein tags or expression cassettes, or loxP sequences, along with the sequences adjacent to the nuclease cleavage sites serving as homology arms26. Such donor DNAs serve as templates to repair the cleavage site through a high-fidelity repair process called HDR. The gRNA can be designed using many programs available online. Some examples are CRISPOR27,; CHOPCHOP28,; and Breaking-Cas29, The repair DNA template, used for knock-in, can be one of two types: a short oligonucleotide (an ssODN up to 200 bases long) or a longer DNA molecule (either circular double-stranded DNA (dsDNA) or linear ssDNA longer than 201 bases). Although ssODN knock-ins are quite efficient, insertion of longer repair templates (particularly the circular dsDNAs) was quite challenging until our group showed that long ssDNAs serve as very efficient repair templates21,22. Principles of the design of repair ssDNA templates are described in detail in our previous paper23.

Although gRNAs can be synthesized in-house using an in vitro transcription method11, synthetic gRNAs can be commercially procured21, which offers convenience as well as consistency in their performance. The Cas9 can be used as an mRNA or protein, with the latter emerging as the preferred form because of higher efficiency of editing and fewer mosaic offspring21,30. Mosaicism refers to the presence of more than two genetically distinct cells in the same individual, which typically occurs in CRISPR-based gene-editing experiments in which the Cas9 is active for a long time in the developing embryo (typically beyond the two-cell stage). The repair DNA templates can be procured as ssODNs (for short sequence insertions) or longer dsDNAs can be commercially synthesized or cloned in-house using standard molecular biology techniques, and long ssDNA repair templates (up to 2 kb) can be synthesized in the laboratory using the in vitro transcription and reverse transcription (ivTRT) method (described below). For i-GONAD experiments, the concentrations of the CRISPR reagents, including repair ssDNA template, are much higher compared to what is needed for microinjection-based methods.

We use the ivTRT method for preparation of long ssDNA up to 2 kb22,23. This includes three major steps: preparation of dsDNA template, RNA synthesis by in vitro transcription and cDNA synthesis by reverse transcription. Most of the steps are similar to those in our previous paper describing ssDNA preparation steps used for microinjection-based genome-engineering methods23. Two major modifications were incorporated into the cDNA synthesis steps in order to obtain a higher concentration of the repair ssDNA template needed for i-GONAD: (i) the reactions were scaled up nearly threefold and (ii) the gel purification step was replaced with column purification followed by ethanol precipitation. Note that the gel purification step causes substantial loss of samples, although it yields very high-quality DNA. Loss of DNA would not be an issue for sample preparation in the case of microinjection-based methods because small quantities of DNA (a few hundred nanograms) would be enough and thus gel purification would be a better approach. Because a high quantity of repair DNA template is more important for i-GONAD, the above modifications to the ivTRT protocol can produce the necessary amounts of repair ssDNA template for the i-GONAD protocols.

i-GONAD procedure

Preparation of pregnant females. We typically use 3–5 pregnant females for one session of i-GONAD experiments. To obtain ~4 plugged females, 6–7 females are mated with males when the Multi Cross Hybrid (Institute of Cancer Research) (MCH(ICR)) strain is used. Note that some strains require more mating pairs. For example, we begin with ~12 females when the C57BL/6 mouse strain is used, because it is often difficult to identify estrus in C57BL/6 mice by observation of vaginal tissues as compared to doing so with the MCH(ICR) strain. Smear examination, prepared from the vaginal epithelia, can be used as an indicator to identify females in the estrus stage.

Although administration of hormones to synchronize estrous cycles and stimulate ovulation (known as superovulation) can increase the embryo numbers, it is not recommended because of the potential risk of either too many embryos being implanted or the mother not successfully delivering the pups. It is known that the superovulation procedure, especially when equine chorionic gonadotropin (eCG) and human chorionic gonadotropin (hCG) are used, can impair subsequent embryo development, resulting in fewer offspring31. Therefore, we recommend using non-superovulated females and avoiding superovulation and, especially, the use of eCG/hCG.

Preparation of genome-editing mixture. As the optimal injection volume is up to 1.5 µl/oviduct in mouse, at least ~3 µl of solution is needed for one pregnant female. We usually prepare 10 µl for three mice and 14 µl for four mice (a slight excess over the required amount) to accommodate any loss during pipetting.

The genome-editing solution is slightly viscous because of the glycerol component, which comes from the Cas9 or Cpf1 protein stock. Inclusion of Fast Green FCF will help in visualizing successful injection and provide visual assistance during placement of the electrodes on the oviduct filled with the colored solution.

Surgical exposure of oviduct and injection. Timing of the i-GONAD procedure is one of the important factors determining the efficiency of genome editing. As described above, our systematic analysis of time points led us to conclude that day 0.7 would be the most suitable time for performing the i-GONAD procedure18. Our observation was that the zygotes at day 0.7 are the most suitable because they are largely devoid of cumulus cells but are still floating in the ampulla.

Injection of solution into the oviduct is the most challenging step in the i-GONAD method. It may require some practice to master this technique. To make this procedure easier, it is important to expose the oviduct adequately and position it so that it can be grasped with forceps to easily pierce the oviductal wall with a capillary pipette (Box 1). We provide some tips and simple protocols in Boxes 2 and 3 for practicing and mastering the procedure.

We suggest drawing 1.5 μl of solution into the capillary tip before injection. In some animals or strains, the amount of solution that can be injected can be <1.5 μl. We observed that the oviducts of C57BL/6-strain mice are comparatively smaller than those of the MCH(ICR) strain. Our typical injection volume ranges from 1.0 to 1.5 μl of solution per oviduct.

In vivo electroporation. In our first report, we assessed the viability of embryos by isolating and culturing the embryos 1 d after the electroporation procedure from the animals subjected to GONAD17. We noticed that approximately half of the embryos seemed abnormal and did not develop properly, whereas most of the control embryos (from animals that were not subjected to GONAD) developed normally. This indicates that electroporation can affect embryo viability and result in diminished numbers of offspring. We think that electroporation conditions may need further optimization to prevent embryo loss while obtaining a high rate of genome-edited offspring. The optimization of current value may help with the reproducibility of i-GONAD; we advise the use of 100–200mA as a starting point.

Analysis of offspring

Genotyping of CRISPR genome-edited animal models in general is quite challenging. This is because of the potential for mosaicism and the possibility of many genome-editing outcomes, including non-homologous end-joining (NHEJ) indels and large deletions or insertions. Identification of desired insertion alleles among such mixtures can be a difficult task. A few different options are available for identification of simple gene disruptions caused by NHEJ-mediated indels; these include T7E1 and Cel1 or surveyor assay11, but the preferred method is sequencing of the target region. If a restriction endonuclease site exists adjacent to where the gRNA binds, targeted offspring can be identified using a restriction fragment length polymorphism (RFLP)-based assay. For short ssODN-mediated knock-in models, if the intended mutation does not result in gain or loss of a restriction enzyme site, consider including such changes (with additional substitutions) in the repair templates, which can aid in detecting gene-editing events quickly and easily11. The genotyping strategies for detecting longer (e.g., >1 kb) knock-in cassettes require multiple different sets of PCR reactions, which are described in detail in our previous protocol23. The targeted locus of the gene-edited animals should be thoroughly examined in the F1 generation by fully sequencing the region to ensure that the correctly targeted allele is segregated from the possible mixture of mosaic alleles.

Expertise needed to implement i-GONAD

i-GONAD can be readily implemented by researchers with some experience in traditional transgenic technologies, which are typically performed in every transgenic core facility. Because i-GONAD does not rely on the complex set of skills needed for ex vivo handling of embryos, and it does not require sophisticated equipment, the procedure can also be performed by anyone with skills in performing small-animal surgeries. The only major equipment needed would be an electroporator, a dissecting microscope and standard surgical equipment. The protocol provided in Box 2 (two-day protocol using fluorescent dye or a fluorescent gene marker) and the experimental procedure in Box 3 (Tyrosinase (Tyr) rescue experiment) should serve as quick recipes for any beginner who wishes to try the method. In our experience of offering more than a dozen workshops, many students and technicians have been able to learn the method within a few attempts.

Limitations of i-GONAD

Compared with microinjection-based traditional transgenesis approaches, i-GONAD requires much higher concentrations of CRISPR and repair template reagents18. The typical concentrations of CRISPR reagents (gRNA and Cas9 protein) used in microinjection mixes is 10–50 ng/µl, whereas the optimal concentrations of these reagents for i-GONAD is up to 100 times higher than that for microinjection. Although this is not a concern for generating gene-disrupted and even point mutation knock-in models, because the reagents needed can be commercially procured in lyophilized form (at high concentrations) and reconstituted (diluted) to the required concentrations, it can be a major limitation for inserting longer ssDNA cassettes. For example, ten times the concentration of a 1-kb sequence would be necessary to obtain a comparable molar concentration for a 100-base repair ssODN template sequence. Another minor limitation of i-GONAD is the poor birth rate in inbred strains, even though the overall efficiency of producing genome-edited offspring is similar across outbred, hybrid and inbred strains. The method will require more optimization to achieve higher birth rates in inbred mouse strains. Mosaicism is one of the common limitations of CRISPR-based genome-editing technologies. We observed a range of 36–65% mosaicism among the i-GONAD-generated offspring, as compared with a 43% mosaicism rate for the microinjection method in our experience.


Biological materials


Experimental procedures involving animals should be carried out according to relevant institutional and governmental regulations. All animal experiments discussed here were performed in accordance with institutional guidelines and were approved by the applicable institutional animal care and use committee (permit nos: 154014, 165009 and 171003 at Tokai University; 2017062 at Hamamatsu University; 17008 at Shigei Medical Research Institute; 2015-079 at Hokkaido University; and IACUC17-0008 and IACUC15-0014 at Meiji University).


Any mouse strain, including hybrid and inbred strains, can be used. Below is the list of mouse strains that we have tested thus far. The ages of mice we have tried are indicated, although we believe older mice can also be used, as long as they can breed satisfactorily.

  • Jcl:MCH(ICR) (hybrid strain; 10–20 weeks for females; 10–50 weeks for males; CLEA Japan)

  • B6D2F1/Slc (hybrid strain; 8 weeks for females; 10–12 weeks for males; Japan SLC)

  • C57BL/6JJcl (CLEA Japan), C57BL/6NCrSlc (Japan SLC) and C57BL/6NCrl (CLEA Japan) (inbred strain; 7–10 weeks for females; 10–24 weeks for males)

  • BALB/cAJcl (inbred strain; 10–18 weeks for females; 10–24 weeks for males; CLEA Japan)

  • C3H/HeSlc (inbred strain; 8 weeks for females; 10–12 weeks for males; Japan SLC)

  • DBA/2CrSlc (inbred strain; 8 weeks for females; 10–12 weeks for males; Japan SLC)



Listed below are the standard kits and reagents that are used in our laboratories. Comparable kits and reagents from other vendors can also be used in place of these.

  • Template plasmid (e.g., pP206 for knocking-in of the T2A-mCitrine cassette into the Pitx3 locus18; custom-synthesized from vendors such as GeneWiz)

  • Primer set (standard desalt grade) for amplification of dsDNA fragment for RNA synthesis (custom DNA oligos; Table 1; Thermo Fisher Scientific)

    Table 1 Oligos
  • KOD-Plus-Neo (Toyobo, cat. no. KOD-401) or any suitable high-fidelity PCR mix

  • TAE (50×; Nippon Gene, cat. no. 313-90035)

  • SeaKem ME agarose (Lonza, cat. no. 50011)

  • Gel loading buffer (6×; Thermo Fisher Scientific, cat. no. R0611)

  • GeneRuler DNA Ladder Mix (Thermo Fisher Scientific, cat. no. SM0331)

  • Ethidium bromide (Sigma-Aldrich, cat. no. E1510-10ml)


    Ethidium bromide is carcinogenic and mutagenic. Wear personal protective equipment when handling this reagent.

  • NucleoSpin Gel and PCR Clean-up Kit (TaKaRa, cat. no. 740609)

  • T7 RiboMax Express Large Scale RNA Production System (Promega, cat. no. P1320)

  • MEGAclear Kit (Ambion; Life Technologies, cat. no. AM1908)

  • SuperScript III reverse transcriptase (Life Technologies, cat. no. 18080051) or SuperScript IV reverse transcriptase (Life Technologies, cat. no. 18091050)

  • Primer for reverse transcription (Table 1; Thermo Fisher Scientific)


    We design primers using the standard criteria used for genomic PCR (e.g., melting temperature: 58–60 °C; nucleotide length: 18–23 bases).

  • Modified TE (Thermo Fisher Scientific, cat. no. 12090015)

  • Sodium acetate (3 M, pH 5.2; Nacalai Tesque, cat. no. 06893-24)

  • Ethanol (99.5 % (vol/vol); Wako, cat. no. 057-00456)

  • EmbryoMax injection buffer (EMD Millipore, cat. no. MR-095-10F), or equivalent.

Anesthetic agent

  • Medetomidine hydrochloride (1.0 mg/ml; Kyoritsu Seiyaku)

  • Midazolam (5.0 mg/ml; Dormicum Injection; Astellas Pharma, cat. no. 1124401A1052)

  • Butorphanol tartrate (5.0 mg/ml; Meiji-Seika-Pharma)

  • Atipamezole hydrochloride (5.0 mg/ml; Kyoritsu Seiyaku)

  • Normal saline (Otsuka Pharmaceutical Factory, cat. no. 1760)

Reagents for the genome-editing mixture

  • crRNA (10 nmol) and tracrRNA for Cas9 (Integrated DNA Technologies (IDT), custom synthesis and cat. no. 1072534, respectively)


    The crRNA is custom-synthesized for each specific gRNA (Table 2), whereas tracrRNA is universal.

    Table 2 crRNAs
  • Alt-R CRISPR-Cpf1 crRNA (10 nmol; IDT, Table 2)


    Alt-R CRISPR-Cpf1 crRNA is custom synthesized.

  • Nuclease-free duplex buffer (IDT, cat. no. 11-05-01-12)

  • Repair ssDNA template (standard desalt ultramers from IDT, used as repair ssODN template) (e.g., ssODN for Tyr rescue experiment: ‘5′-TGTTTTATAATAGGACCTGCCAGTGCTCAGGCAACTTCATGGGTTTCAACTGCGGAAACTGTAAGTTTGGATTTGGGGGCCCAAATTGTACAGAGAAGCGAGTCTTGATTAGAAGAAACATTTTTGATTTG-3′’, Box 3), or long ssDNAs synthesized using the ivTRT protocol

  • Nuclease-free water (not diethyl pyrocarbonate (DEPC)-treated; Thermo Fisher Scientific, cat. no. AM9938)

  • Alt-R S.p. (Streptococcus pyogenes) Cas9 nuclease 3NLS (IDT, cat. no. 1074182)

  • Alt-R A.s. (Acidaminococcus sp.) Cpf1 (Cas12a) nuclease 2NLS (IDT, cat. no. 1076159)

  • Fast Green FCF (Nacalai Tesque, cat. no. 15939-54)

  • Opti-MEM medium (Thermo Fisher Scientific, cat. no. 31985062)

Reagents for the i-GONAD procedure

  • Ethanol for disinfection (Wako, cat. no. 324-00037)

  • PBS (10×, without Ca2+ and Mg2+, pH 7.2; Thermo Fisher Scientific, cat. no. 70013032)

  • Water (Sigma-Aldrich, cat. no. W3500)


  • Allele-In-One Mouse Tail Direct Lysis Buffer (Kurabo, cat. no. ABP-PP-MT01500)

  • Primers (standard desalt grade) for amplifying and sequencing the target region (Thermo Fisher Scientific, custom DNA oligos; Table 1)

  • TaKaRa Taq (dNTP is included in the kit; TaKaRa, cat. no. R001)

  • GC buffer (2×; TaKaRa, cat. no. 9154)

Reagents needed for the 2-d protocol

  • Tetramethylrhodamine-labeled dextran (3 kDa; 10 mg; Thermo Fisher Scientific, cat. no. D3307)

  • EGFP mRNA (can be purchased from vendors or in vitro–synthesized17); e.g., CleanCap EGFP mRNA (TriLink BioTechnologies, cat. no. L7601)

  • Trypan blue stain solution (0.5% (wt/vol); Nacalai Tesque, cat. no. 29853-34)

  • FBS (Nichirei Biosciences, cat. no. 12106C)



Equipment and instruments required for the i-GONAD procedure are shown in Fig. 2.

  • Thin-wall PCR tubes and caps (0.2 ml; eight-strip; Thermo Fisher Scientific, cat. no. 435440-Q)

    Fig. 2: Equipment and instruments required for i-GONAD procedure.

    a, Workbench displaying the major equipment, instruments and material needed for performing i-GONAD. be, Higher magnification of surgical instruments (b), dissecting microscope and mouth-pipetting device (c), and electroporators (CUY21EditII (d) and NEPA21 (e)).

  • Micropipettes (Eppendorf Research Plus, cat. no. 3120 000.909)

  • Microcentrifuge (Tomy, model no. MX-307 or equivalent)

  • Thermocycler (Eppendorf, model no. Mastercycler Nexus X2)

  • Microwave oven (Mitsubishi, cat. no. RR-M1)

  • Gel electrophoresis system (Pharmacia Biotech, model no. EPS300)

  • Gel documentation system (ATTO, cat. no. AE-6932GXES with UV transilluminator (model no. WUV-M20))

  • Spectrophotometer (Thermo Fisher Scientific, model no. NanoDrop One)

  • Microcentrifuge tubes (1.5 ml; BM Equipment, cat. no. NT-175)

  • RNase-free pipette tips (ZAP 200 µL Aerosol Filter Pipet Tip, Labcon, cat. no. 1059-965-008, and ZAPSILK Aerosol Pipet Tips 10 µL, Labcon, cat. no. 1173-965-008)

  • Latex gloves (AS ONE, cat. no. 1-8448-03)

  • Heat block (dry thermo unit; TAITEC, cat. no. DTU-2B)

  • PCR tubes (0.2 ml with flat caps; Genetics, cat. no. FG-021F)

  • Vortex mixer (AS ONE, model no. VM-96B)

  • Tuberculin syringes with needle (1 ml; Terumo, cat. no. SS-01T2613S)

  • Hot plate (Makami Denshi, cat. no. MP-916-N)

  • Forceps (Napox, cat. nos. A-10, A-15 and A-18; Inox, cat. no. 5)

  • Square Petri dish (Sigma-Aldrich, cat. no. Z617679)

  • Vinyl tape (19 mm × 10 m; Nichiban, cat. no. VT-195)

  • Scissors (Napox, cat. no. B-12)

  • Paper towels (JOINTEX, cat. no. N015J-S-35)

  • Tissue culture dishes (60-mm; Iwaki, cat. no. 3010-060)

  • Aorta clamp (Napox, cat. no. C-17-40-2)

  • Dissecting microscope (Olympus, model no. SZ11)

  • Stereomicroscope light source (Olympus, model no. KL1600 LED)

  • Microcap and associated glass tube (Drummond, cat. no. 1-000-2000)

  • Pipette tips, TipOne, blue graduated tip (1,000 μl; USA Scientific, cat. no. 1111-2021)

  • Pipette tips (200 μl; Watson Bio Lab, cat. no. 122-703C)

  • Silicon capillary tube (3 mm (o.d.); AS ONE, cat. no. 6-586-03)

  • Glass capillaries (Narishige, cat. no. GDC-1; Drummond, cat. no. 2-000-050; or Drummond, cat. no. 1-000-0300)

  • Micropipette puller (Sutter Instrument, model no. P-97/IVF, or Narishige, model no. PN-31)

  • Microscissors (Napox, cat. no. MB-53)

  • Kimwipes S-200 (Nippon Paper Crecia, cat. no. 62011)

  • Electroporator (BEX, model no. CUY21EditII; NepaGene, model no. NEPA21; or BTX, model no. T820)

  • Tweezer-type electrodes (BEX, cat. no. LF650P3 or NepaGene, cat. no. CUY652-3)

  • Absorbent cotton (AS ONE, cat. no. 40-500G)

  • Suture wound clips (MikRon 9-mm Autoclips; Becton Dickinson, cat. no. 427631)

  • Shaking heat block (ThermoMixer Comfort; Eppendorf)

  • Labo burner (Phoenix Dent, model no. APT-L) or alcohol lamp (Maruemu, cat. no. 0613-01)

  • Aluminum block (on ice; Ina-Optika, cat. no. ABR-24W-02)

  • Needle (30 gauge; Natsume, cat. no. KN-386)

  • Inverted fluorescence microscope (Olympus, model no. IX70) with adequate filter sets (e.g., Olympus, model nos. U-MNIBA (for EGFP) and U-MWIG (for tetramethylrhodamine))

  • Fluorescence illumination source (Olympus, model no. U-HGLGPS)

  • Dissecting microscope (Olympus, model no. SZX7)

Reagent setup

1× TAE

Mix 40 ml of 50× TAE with 1,960 ml of dH2O. The 1× solution can be stored at room temperature (RT; 25 °C) for up to 3 months.

1% (wt/vol) Agarose gel

Measure the required amount (% wt/vol) of agarose in 1× TAE. For example, measure 1 g of agarose in 100 ml of 1× TAE to prepare a 1% (wt/vol) agarose gel, which would be sufficient for pouring a 15 cm × 15-cm gel tray. Boil the solution in a microwave oven until the agarose is dissolved completely, cool to 50–60 °C, and pour onto the gel tray with a gel-comb. Make the gel immediately before the experiment.

70% (vol/vol) Ethanol

Mix 35 ml of 99.5% (vol/vol) ethanol and 15 ml of nuclease-free water in a 50-ml conical centrifuge tube to prepare 50 ml of 70% (vol/vol) ethanol; store it tightly sealed at RT for up to 6 months.

crRNA and tracrRNA (for Cas9)

Resuspend the crRNA and tracrRNA each to a 200 µM final concentration in nuclease-free duplex buffer and store at −80 °C for up to 18 months.

crRNA (for Cpf1)

Resuspend the crRNA to a 100 µM final concentration in nuclease-free duplex buffer and store at −80 °C for up to 18 months.

gRNA preparation

For Cas9-based genome editing, anneal crRNA and tracrRNAs by mixing equimolar ratios. For example, mix 10 μl of 200 µM crRNA and 10 μl of 200 µM tracrRNA and anneal in a thermocycler (94 °C for 2 min); then place at RT for ~10 min. The crRNA–tracrRNA complex (100 µM each) can be stored at −80 °C for up to 18 months. We call this crRNA–tracrRNA complex ‘gRNA’.

Ultramers for repair ssODN template

Resuspend ultramers to a final concentration of 10 µg/µl in nuclease-free water and store at −80 °C for up to 18 months.

Fast Green FCF

Resuspend 0.1 g of Fast Green FCF in 50 ml of Opti-MEM to a concentration of 0.2% (wt/vol). Make 1-ml aliquots in microcentrifuge tubes and store at −20 °C or −80 °C for up to 18 months.

Anesthetic agents

Combine 0.75 ml of medetomidine hydrochloride (1.0 mg/ml) with 0.8 ml of midazolam (5.0 mg/ml), 1.0 ml of butorphanol tartrate (5.0 mg/ml) and 7.45 ml of normal saline. Combine 0.15 ml of atipamezole hydrochloride (5.0 mg/ml) with 9.85 ml of normal saline. Store these solutions at 4 °C for up to 1 month.

1× PBS

Add 55.5 ml of 10× PBS to a 500-ml bottle of distilled water and mix well. 1× PBS can be stored at RT for at least 1 year. Prepare a 60-mm tissue culture dish containing 1× PBS before in vivo electroporation is performed.

Tetramethylrhodamine-labeled dextran

Add 5 ml of 1× PBS to a vial containing 10 mg of 3-kDa tetramethylrhodamine-labeled dextran and dissolve the tetramethylrhodamine-labeled dextran to obtain a 2 µg/µl concentration. Make 125-μl aliquots in microcentrifuge tubes and store at−20 °C or −80 °C; tetramethylrhodamine-labeled dextran aliquots remain usable for at least a year.


Add 5 ml of sterile 1× PBS to a vial containing 0.15 ml of FBS to obtain a final concentration of 3% (vol/vol). PBS–FBS can be stored at 4 °C for at least a year.

Equipment setup

Square Petri dish table for anchoring the anesthetized mouse

Anchor the plate and its lid with vinyl tape (Fig. 2a).

Paper towel pieces

Cut paper towels into 2–3 cm2 pieces and place them in a 60-mm tissue culture dish (Fig. 2b).

Preparation of injection capillaries using commercial needle pullers

The glass capillary is similar to the type used for DNA microinjection into zygotes in a transgenic mouse production experiment. Following are the conditions for making injection capillaries with commercially available instruments. Sutter P-97/IVF settings for a Narishige GDC-1 glass capillary: heat, 790; pull, 35; velocity, 50; pressure, 500; time, 100. Narishige PN-31 settings for a Drummond 1-000-0300 glass capillary: heater, 80; magnet-sub, 30; magnet-main, 60–80 (dependent on the season and/or humidity). The glass capillary tubes used for the i-GONAD procedure need not be of high quality (as required for embryo microinjections). See Box 1 for the preparation of capillaries for the i-GONAD procedure without using the needle puller equipment.

Mouth pipette for intraoviductal injection

Many types of mouth-pipetting devices are used by the transgenic research community. The one described here is assembled using an aerosol barrier filter pipette tip, silicon tubing, a 1,000-μl pipette tip, a glass tube containing a microcap and a glass capillary (Supplementary Fig. 1a). A filter pipette tip is inserted into one end of the silicon tubing, and the pointy end of a 1,000 μl pipette tip is inserted into the other end. The wider end of the 1,000 μl pipette tip is fitted with a glass tube, which is fitted with a microcap on its other end.

Kimwipe towel pieces

Cut a Kimwipe towel into ~3 mm × 5-mm pieces and place them in a 60-mm tissue culture dish (Fig. 2b).

Mouth pipette for embryo handling

This apparatus is used for transferring embryos from one dish to another. This is similar to the mouth pipette used for intraoviductal injection, except for the glass capillary. We use hand-made, pipettes that are finely drawn from glass capillary tubes (Drummond, cat. no. 1-000-2000) by flaming in the middle and pulling the ends. The tips are flamed briefly to smooth the sharp ends.


gRNA design

Timing 30 min

  1. 1

    Design gRNA(s) at the target region of interest. Candidate gRNA sequences can be designed using several online algorithms (e.g., CRISPOR27, CHOPCHOP28 and Breaking-Cas29). Select the target site as close as possible to the desired insertion site. To avoid off-target editing, target sites should contain two or more nucleotide mismatches to other genomic regions.

Repair ssDNA template design

Timing 1–3 h


A repair DNA template is designed if precise editing such as targeted insertions or replacement based on the HDR pathway (but not an NHEJ-based indel mutation) is required. These repair DNA templates are often supplied as ssDNA.

  1. 2

    Design the repair ssDNA template so that the ssDNA contains the insertion/replacement sequence flanked by 50- to 100-base homology arms on each side. When long ssDNA is synthesized in-house by ivTRT, a T7 RNA polymerase promoter sequence should be included immediately upstream of one of the homology arms. An example architecture for an ssDNA design would be T7 promoter– left arm–DNA of interest–right arm. The principles of design for long repair ssDNA templates are described in greater detail in Miura et al.23.

Preparation of repair ssDNA template

Timing 2–5 d

  1. 3

    Lyophilized repair ssDNA template can be purchased as an ssODN (e.g., ultramer from IDT) (option A). Long ssDNAs (e.g., longer than 201 bases) can also be synthesized in-house (option B).

    1. (A)

      Repair ssODN template

      1. (i)

        Order an ssODN (up to 200 bases long) from a commercial vendor. A synthesis scale of 20 nmol (standard desalt grade) is sufficient for many i-GONAD experiments.

    2. (B)

      Long repair ssDNA template


      Here, we describe the modifications to our published procedure for synthesizing long ssDNAs23 needed to obtain larger amounts of repair DNA template, which is essential to the success of i-GONAD.

      1. (i)

        Preparation of dsDNA template. Here, we describe a protocol to prepare dsDNA by PCR using plasmid DNA as PCR template. The resulting PCR product will be used as a template for RNA synthesis. The plasmid DNA can be obtained by standard cloning or custom DNA synthesis from commercial vendors (e.g., GeneWiz). Template for RNA synthesis is generated by PCR using KOD-Plus-Neo DNA polymerase. Reactions are scaled up to obtain higher amounts of DNA. The reaction described below is 3× scale (3 × 50 µl) of the manufacturer’s protocol. Mix the reaction in a PCR tube by pipetting and a short spin (2,000g for 5 s) in a microcentrifuge at RT. Primers used for amplifying PCR products can be standard desalt grade from any commercial vendor.

        Component Amount per reaction (µl) Final concentration
        Template plasmid DNA (1–2 ng/µl) 3 3–6 ng
        10× PCR buffer for KOD-Plus-Neo 15
        dNTPs mixture (2 mM) 15 0.2 mM
        MgSO4 (25 mM) 9 1.5 mM
        F Primer (20 µM; Table 1) 3 0.4 µM
        R Primer (20 µM; Table 1) 3 0.4 µM
        KOD-Plus-Neo enzyme (1 U/µl) 3 0.02 U/µl
        Water 99  
        Total 150  

        Critical step

        Use a high-fidelity Taq polymerase for generating a PCR template.

      2. (ii)

        Perform PCR in a thermocycler with the following conditions:

        Cycle no. Denature Anneal/extend
        1 95 °C for 1 min  
        2–31 95 °C for 15 s 68 °C for 3 min
        32   68 °C for 7 min
        Hold at 4 °C   
      3. (iii)

        After the PCR reaction, verify the PCR product by agarose gel electrophoresis. Load 3 µl of the PCR product mixed with 1 µl of 6× gel loading buffer, along with 2 µl of DNA markers (100 ng/µl; GeneRuler DNA Ladder Mix), onto a 1% (wt/vol) agarose gel in 1× TAE. Run at 135 V for 30 min. Analyze the gel using a gel documentation system. After confirming the successful amplification of the PCR reaction, purify the remaining PCR product (147 µl) using a NucleoSpin Gel and PCR Clean-up Kit (column purification).

      4. (iv)

        Add 300 µl of Buffer NT1 (included in the kit) to 147 µl of PCR product and mix well.

      5. (v)

        Transfer the sample (447 µl) to a column and centrifuge for 30 s at 11,000g at RT.

      6. (vi)

        Discard the flow-through; add 700 µl of Buffer NT3 (included in the kit) to the column and centrifuge for 30 s at 11,000g at RT.

      7. (vii)

        Repeat washing with Buffer NT3 (above step).

      8. (viii)

        Centrifuge for 1 min at 11,000g at RT to completely remove the Buffer NT3.

      9. (ix)

        Place the column into a new microcentrifuge tube, add 15 µl of Buffer NE (included in the kit) and incubate at RT for 1 min.

      10. (x)

        Centrifuge for 1 min at 11,000g at RT (first elution).

      11. (xi)

        Prepare for a second elution by adding 15 µl of Buffer NE and incubate at RT for 1 min.

      12. (xii)

        Centrifuge for 1 min at 11,000g at RT (second elution; total elution volume will be 30 µl).

      13. (xiii)

        Use 1 µl of the elution to estimate the concentration with a NanoDrop spectrophotometer.

      14. (xiv)

        After the purification, verify the purified PCR product by agarose gel electrophoresis. Load 1 µl of sample after addition of 4 µl of Milli-Q water and 2 µl of 6× gel loading buffer, along with DNA markers (GeneRuler DNA Ladder Mix), onto a 1% (wt/vol) agarose gel in 1× TAE. Run at 135 V for 30 min. Store the samples at −20 °C.

        Pause point

        The samples can be stored at −20 °C for up to 1 year.

      15. (xv)

        RNA synthesis using T7 RiboMax Express. RNA is synthesized according to the manufacturer’s instructions (T7 RiboMax Express Large Scale RNA Production System). Add the following reagents to a PCR tube or microcentrifuge tube and mix by pipetting and give a short spin (2,000g for 5 s at RT). 1 µg of the template DNA is used for in vitro transcription.

        Component Amount per reaction
        RiboMAX Express T7 2× Buffer 10 µl
        Linear DNA template DNA (1 µg) x µla
        Nuclease-free water Up to 20 µl
        Enzyme mix, T7 Express 2 µl
        Total 20 µl
        1. a1 µg of the template from Step 3B(xii) is used (adjust the volume according to the DNA concentration).

        Critical step

        Avoid repeated freeze–thaw of buffers. Thaw all the reagents on ice and be sure to use RNase-free tips and gloves in all the RNA synthesis steps.

      16. (xvi)

        Incubate the reaction at 37 °C in a thermocycler for 30 min.

      17. (xvii)

        Add 1 µl of RQ1 RNase-Free DNase (included in the kit), mix well, and incubate at 37 °C in a thermocycler for 15 min to eliminate the DNA template.

      18. (xviii)

        Purification of RNA using a MEGAclear Kit. RNA is purified using a MEGAclear Kit according to manufacturer’s instruction. First, preheat a dry heat block to 70 °C and pre-warm the elution solution (from the MEGAclear Kit).

      19. (xix)

        Add 80 µl of elution solution to the sample from Step 3B(xvii) (total volume to 101 µl) and mix by gentle pipetting.

      20. (xx)

        Add 350 µl of binding solution (from the MEGAclear Kit) and mix by gentle pipetting.

      21. (xxi)

        Add 250 µl of >99.5% (vol/vol) ethanol and mix by gentle pipetting.

      22. (xxii)

        Transfer the sample to the column and centrifuge (21,000g, 1 min, RT).

      23. (xxiii)

        Discard the flow-through and re-insert the column into the microcentrifuge tube.

      24. (xxiv)

        Add 500 µl of washing solution (from the MEGAclear Kit) to the column and centrifuge (21,000g, 1 min, RT).

      25. (xxv)

        Discard the flow-through and repeat washing (as in Step 3B(xxiii) to 3B(xxiv)).

      26. (xxvi)

        Discard the flow-through and centrifuge (21,000g, 30 s, RT) to completely remove all traces of ethanol.

      27. (xxvii)

        Insert the column into a new 1.5-ml microcentrifuge tube and add 50 µl of pre-warmed elution solution from Step 3B(xviii) directly to the column bed, incubate for 10 min, and centrifuge at 21,000g, for 1 min, at RT. A second elution is performed by addition of 50 µl of pre-warmed elution solution and centrifuging at 21,000g, for 1 min, at RT to recover more RNA (total = 100 µl).

      28. (xxviii)

        Check the quality and concentration of the RNA by analyzing 1 µl of the sample in a NanoDrop spectrophotometer. Typically, the concentration of the RNA will be 0.8–1.4 µg/µl.

      29. (xxix)

        Confirm the quality of the RNA by 1% (wt/vol) agarose gel electrophoresis, along with DNA markers (GeneRuler DNA Ladder Mix) in 1× TAE (Fig. 3a). Store the samples at −80 °C until use.

        Fig. 3: Examples of 1% (wt/vol) agarose gels showing electrophoresis.

        a,b, Electrophoresis of RNA (a) and cDNA (b) synthesized using the ivTRT protocol. Lane 1: ~0.9-kb-long cassette consisting of the mCitrine fluorescence gene and homology arms for the Mmp9 locus21; Lane 2: a similar cassette containing homology arms for the Pitx3 locus18. The RNA and cDNA were electrophoresed along with a dsDNA marker (M). The sizes of the major bands of the marker are shown on the left side of the gel image. The same dsDNA ladder marker was used for both gels.


        Pause point

        The samples can be stored at −80 °C for up to 1 year.

      30. (xxx)

        Synthesis of cDNA from RNA. cDNA is synthesized according to the manufacturer’s instructions (Superscript III Reverse Transcription protocol). Add the following reagents to a microcentrifuge tube, mix by pipetting and give a short spin (2,000g for 5 s at RT). About 50 µg of the purified RNA is used as a template for cDNA synthesis.

        Component Amount per reaction (µl) Final concentration
        RNA template from Step 3B(xxvii) x 50 µg
        Reverse transcription primer (100 µM; Table 1) 5 5 µM
        dNTP mix (10 mM) 10 1 mM
        DEPC-treated water Up to 100  
        Total 100a  
        1. aReactions can be scaled up to obtain a higher quantity of cDNA. This reaction is a 10× scale (10 × 10 µl) of the manufacturer’s protocol.


        Be sure to use RNase-free tips and gloves in all the cDNA synthesis and purification steps.

      31. (xxxi)

        Mix the reaction by pipetting and a short spin (2,000g for 5 s at RT) in a centrifuge.

      32. (xxxii)

        Divide the sample into five 20-µl aliquots in 0.2-ml PCR tubes with flat caps (all labeled ‘tube A’).

      33. (xxxiii)

        Incubate each tube at 65 °C for 5 min in a thermocycler.

      34. (xxxiv)

        Immediately place the tubes (all tube A) on ice, wait for at least 1 min and then proceed to cDNA synthesis. Mix the following reagents (included in the kit) in a new tube (tube B).

        Component Amount per reaction (µl)
        10× Reverse transcription buffer 20
        0.1 M DTT 10
        25 mM MgCl2 40
        RNaseOUT (40 U/µl) 10
        SuperScript III reverse transcriptase (200 U/µl) 10
        Total 90


        DTT is harmful when inhaled, and it may cause irritation to the skin and eyes. Wear gloves and personal protective equipment when working with DTT.

      35. (xxxv)

        Add 18 µl of the solution from tube B to each tube A. Mix well and incubate the reaction at 50 °C for 50 min in a thermocycler. Stop the reaction by heat-inactivating at 85 °C for 5 min. Cool the reaction to RT.

      36. (xxxvi)

        Add 2 µl of RNaseH to each tube A, mix by pipetting and incubate at 37 °C for 20 min in a thermocycler.

      37. (xxxvii)

        Combine all the samples in a microcentrifuge tube (total 200 µl). Store the samples at −20 °C or −80 °C until use.

        Pause point

        The samples can be stored at −20 °C or −80 °C for up to 1 year.

      38. (xxxviii)

        Purify the cDNA product using a NucleoSpin Gel and PCR Clean-up Kit (column purification). First, add 380 µl of Buffer NT1 to 200 µl of sample and mix well.

      39. (xxxix)

        Transfer the sample (580 µl) to a column and centrifuge for 30 s at 11,000g at RT.

      40. (xl)

        Discard the flow-through, add 700 µl of Buffer NT3 to the column and centrifuge for 30 s at 11,000g at RT.

      41. (xli)

        Repeat the washing with Buffer NT3 as described in Step 3B(xl) above.

      42. (xlii)

        Centrifuge for 1 min at 11,000g at RT to completely remove the Buffer NT3.

      43. (xliii)

        Place the column into a new microcentrifuge tube, add 30 µl of Buffer NE, and incubate at RT for 1 min.

      44. (xliv)

        Centrifuge for 1 min at 11,000g at RT (1st elution).

      45. (xlv)

        Place the column into the tube, add 30 µl of Buffer NE and incubate at RT for 1 min.

      46. (xlvi)

        Centrifuge for 1 min at 11,000g at RT (second elution; total elution volume will be 60 µl).

      47. (xlvii)

        Ethanol precipitation of cDNA. Add 40 µl of modified TE (10 mM Tris-HCl, 0.1 mM EDTA), 10 µl of 3 M sodium acetate (pH 5.2) and 250 µl of 99.5% (vol/vol) ethanol to 60 µl of sample. Vortex and centrifuge at 21,000g, 4 °C, for 10 min. Discard the supernatant, add 130 µl of 70% (vol/vol) ethanol and centrifuge (21,000g, 4 °C, 2 min).

      48. (xlviii)

        Completely remove the supernatant and air-dry the pellet for 5–10 min. Dissolve the pellet in 4.2 µl of injection buffer.

      49. (xlix)

        Transfer 0.5 µl of cDNA to a new tube and dilute it with 4.5 µl of injection buffer (total 5 µl). Use 1 µl for concentration determination with a NanoDrop spectrophotometer (expected concentration of original solution is typically 1.4–2.3 µg/µl) and the remaining 4 µl for agarose gel electrophoresis after adding 2 µl of 6× gel loading buffer (Fig. 3b).


      50. (l)

        Store the remaining sample (3.7 µl) at −20 °C or −80 °C for future use.

        Pause point

        The sample can be stored at −20 °C or −80 °C for up to 1 year.

Preparation of pregnant female mice

Timing 2 d

  1. 4

    House mice in cages on a 12:12 light cycle (light on at 7:00 am, light off at 7:00 pm). Male mice are maintained in individual cages and female mice are group-housed (5–6 per cage).

  2. 5

    Breed the adequate-estrus female mice with males overnight (from 5:00 pm). To do so, transfer 1–2 estrus females per male mouse cage.

    Critical step

    Choose females showing signs of estrus for breeding. The typical signs of estrus include vaginal tissues that appear light pink and are less moist, and a relatively larger vaginal opening.

  3. 6

    Examine the females next morning at ~9:00 am for vaginal plugs. Separate the plugged females from the males, place them into new cages and house them until they are used for the i-GONAD procedure at 4:00 pm.


Preparation of genome-editing mixture

Timing 1 h

  1. 7

    Mix the components described in the table below (e.g., 10 µl mix can be used for three females) in a 1.5-ml tube on ice. Keep the solution on ice.

    Component For creating knockout For creating knock-in of small DNA fragmenta For creating knock-in of longer DNA fragmentb Final concentration
    gRNA (100 µM; Reagent setup, Table 2) 3 µl 3 µl 1.8 µl 30 µM
    Repair ssDNA template
     ssODN (10 µg/µl, 150–300 µM)   2 µl   2 µg/µl (30–60 µM)
     ssDNA (1.4–2.3 µg/µl, 5–30 µM)    3.6 µl 0.8–1.4 µg/µl (3–20 µM)
    Cas9 or Cpf1 protein (10 mg/ml, 61 µM for Cas9 and 63 µM for Cpf1) 1 µl 1 µl 0.6 µl 1 mg/ml (6.1 µM for Cas9 and 6.3 µM for Cpf1)
    Fast Green FCF, optional, to facilitate visualization of solution (0.2% (wt/vol)) (1 µl) (1 µl)   0.02% (wt/vol)
    Opti-MEM Up to 10 µl Up to 10 µl   
    Total 10 µl 10 µl 6 µl  
    1. aUp to 200 bases, including homology arms. bFrom 201 to 1,000 bases, including homology arms.

    Critical step

    The genome-editing mixture is prepared on the day of the i-GONAD experiment. As a loss of solution due to leakage during the injection step (Step 20 below) is possible, we recommend that beginners prepare an excess amount of genome-editing mixture (e.g., twice the amount needed). Use of filtered tips is recommended because the solution contains RNA.

  2. 8

    Incubate the tube containing the genome-editing mixture at 37 °C for 10 min just before performing the i-GONAD procedure. The mixture can be kept at RT until the end of the procedure (e.g., 3 h).

Surgical procedure to expose ovary, oviduct and part of the uterus

Timing 10–20 min

  1. 9

    Measure the body weights of the plugged female mice to calculate the appropriate dose of anesthetic agent and inject the anesthetic agent via the i.p. or s.c. route (0.1 ml/10 g) using a 1-ml tuberculin syringe with a needle. Transfer the mouse back to the cage, and place the cage on a hot plate (40 °C) and watch for signs of anesthesia, such as unconsciousness (e.g., 5 min).

  2. 10

    Confirm anesthesia by pinching the mouse’s toes with an A-10 forceps. If the mouse does not respond (indicative of deep anesthesia), place it on a square Petri dish and anchor the legs using strips of vinyl tape.

  3. 11

    Spray ethanol for disinfection onto the dorsal skin of the anesthetized mouse. Make a dorsal incision (10–15 mm long) at the central portion of the back with scissors (Fig. 4a, Supplementary Video 1).

    Fig. 4: Diagrams showing the surgical steps to expose ovary/oviduct/uterus (Steps 10–16).

    a, An anesthetized mouse secured with vinyl tape is shown; a dorsal incision at the central portion of the back of the anesthetized mouse is made with scissors. b, Grabbing the muscle layer with an A-10 forceps (with left hand) and making an incision. c, Pulling the ovary outside of the incision after grasping the adipose tissue with an A-15 forceps (right hand). d, Placing the exposed ovary, oviduct and part of the uterus on a small piece of paper towel. e, Turning and/or positioning of the oviduct by moving the aorta clamp, where it will be clearly observable under the dissecting microscope. Choices 1 and 2 are the two options for sites of intraoviductal injection (Fig. 6).

  4. 12

    Place a small piece of paper towel (~1.5–2.0 cm2) on the skin immediately next to the incision (Supplementary Video 1).

  5. 13

    Grab the muscle layer on the left side with an A-10 forceps and lift it. Then make an incision (spanning 5–7 mm) in the muscle layer with scissors (Fig. 4b).

  6. 14

    Grasp the adipose tissue surrounding the ovary with the A-15 forceps and pull the tissue outside of the incision (Fig. 4c).

  7. 15

    Place the exposed ovary, oviduct and part of the uterus on the paper towel and anchor the adipose tissue with an aorta clamp in order to prevent the tissues from returning to their original positions (Fig. 4d,e).

  8. 16

    Next, place the ampulla (a portion of the oviduct showing oviductal expansion and containing the fertilized zygotes) in position by moving (rotating) the aorta clamp so that the ampulla can be clearly observed under a dissecting microscope (Fig. 4e).

Intraoviductal injection of the genome-editing solution

Timing 2–5 min


A beginner researcher can perform the i-GONAD procedure (Steps 17–31) on ~2–4 pregnant mice in one session. An experienced researcher can perform the procedure on as many as eight females in one session.

  1. 17

    Using a micropipette, take 1.5 µl of the solution from the bottom of the tube containing the genome-editing mixture (from Step 8) and place the solution on the inner wall of the upper part of the tube (Fig. 5a–c) (Supplementary Video 2).

    Fig. 5: Diagrams showing loading of genome-editing reagents into an injection needle (Steps 17 and 18).

    a, A 1.5-ml tube containing the genome-editing mixture. b, Aspiration of part of the solution from the bottom of the tube into a micropipette. c, Placement of the solution on the inner wall of the tube, close to its opening. d, Loading of the solution from the tube into the capillary.

  2. 18

    Using the mouth pipette (Supplementary Fig. 1a) fitted with a capillary tube (Box 1) at the end, draw the solution from this spot while watching the droplet being drawn under the microscope (Fig. 5d; Supplementary Video 2).

  3. 19

    Using an aorta clamp, position the surgically exposed ampulla of the mouse in the center of the field of view, as shown in Fig. 4e.

  4. 20

    Holding an A-1 or no. 5 forceps in one hand, gently grasp the oviduct. With the other hand, insert the capillary into the lumen of the oviduct by piercing the oviductal wall (choice 1 and 2 of Figs. 4e and 6). Inject the sample solution. Note that holding of the oviduct with the forceps may not be necessary at this stage, as shown in the right-hand figure of choice 1 in Fig. 6, although holding may help prevent backflow of the injected solution toward the ovary. The quantity of the solution injected will be sufficient if the sample solution fills the ampulla. If the tip is incorrectly inserted, the solution may flow toward the ovary (instead of flowing toward the uterus, where the embryos are normally present at this stage of pregnancy) (Supplementary Videos 3 and 4).

    Fig. 6: Diagrams showing intraoviductal injection (Step 20).

    First choice (Choice 1; upper panel) for injection is the portion located between the ampulla and the infundibulum. The image on the right side of the arrow represents the flow of the genome-editing mixture. The second choice (Choice 2; lower panel) for injection is just upstream of the ampulla. Note the difference between the angle of rotation of the oviduct in Choice 1 and that in Choice 2.

    Critical step

    This step is critical to the success of the i-GONAD procedure. A researcher proficient in traditional transgenic technologies with experience in handling oviducts, such as oviductal embryo transfer techniques, can learn the i-GONAD procedure very quickly. However, beginners may need to practice on a few mice to master the technique. We suggest two options for injection sites for intraoviductal injection. The preferred site is located between the ampulla and infundibulum, labeled ‘Choice 1’ in Figs. 4e and 6. This region is relatively whitish (paler) and less fragile than the other parts of the oviduct and, therefore, it is relatively safer for grasping the oviduct with forceps. The second site of choice is located at a portion just upstream of the ampulla, labeled ‘Choice 2’ in Figs. 4e and 6. If the researcher is not comfortable with choice 1, the mouse (secured on the dish) can be rotated clockwise approximately a quarter turn (~90°) to position the oviductal tissue for choice 2 (see dotted semi-circular arrow in Fig. 6). This region is somewhat more fragile than the choice 1 spot, and therefore requires more careful handling of the tissue. The success of injection can be assessed by including Fast Green FCF in the mixture, which serves as a visible indicator (Supplementary Fig. 2a).


  5. 21

    Withdraw the capillary after injecting the solution, then proceed immediately to Step 22.

    Critical step

    Because the fertilized eggs are surrounded by a viscous layer of cumulus cells, the Cas9/CRISPR/ssODN solution may not effectively reach to the surface of the zygotes. To achieve uniform dispersion of the injected solution within the oviductal lumen, we suggest tightly clamping the oviduct distal to the injected site with forceps and gently squeezing the ampulla for ~50–100 times (using another forceps) while watching under the microscope for movement of the colored solution (Supplementary Video 5). Although we have not compared the gene-editing efficiencies when squeezing was performed (versus no squeezing), we think that squeezing can ensure distribution of the injected solution throughout the ampulla, eventually dispersing though cumulus cells to reach the surface of the zygotes.

In vivo electroporation

Timing 1–3 min

  1. 22

    Cover the oviduct with a piece of wet Kimwipe towel soaked in 1× PBS immediately after the injection of solution into the oviductal lumen (Fig. 7a, Supplementary Video 6).

    Fig. 7: Diagrams showing in vivo electroporation (Steps 22–26).

    a, Covering the oviduct with a piece of wet Kimwipe towel. b, Wetting of the electrodes in PBS. c, Placement of electrodes on the Kimwipe-covered oviduct to grasp the tissue for electroporation (left) and electroporation of the oviduct (right). d, Wiping of the inner surface of the electrodes with absorbent cotton soaked in 70% (vol/vol) ethanol to clean it for use on the next oviduct.

  2. 23

    Dip the tweezer-type electrodes in 1× PBS transiently (Fig. 7b).

  3. 24

    Hold the Kimwipe towel-covered oviduct between the two electrodes (Fig. 7c). This is a critical step that requires correct placement of a piece of Kimwipe towel to prevent tissue injury caused by electric shock.

    Critical step

    Use of Fast Green FCF can help in finding the correct position at which to apply the electrodes for grasping the tissue (Supplementary Fig. 2b).

  4. 25

    Apply square-wave pulses using the electroporator. The successful delivery of pulses can be ensured by (i) a slight convulsion of the mouse itself and (ii) generation of fine small bubbles between the electrodes (these can be observed under the dissecting microscope) (Fig. 7c, Supplementary Video 6). The suggested electroporation conditions for different commercial electroporators are as follows:

    • CUY21EditII: square (mA), (+/−), Pd V (voltage of the pulses): 60 V or 80 V, Pd A (current of the pulses): 100–150 mA, Pd on: 5.00 ms, Pd off: 50 ms, Pd N (pulse cycle): 3, decay: 10%, decaytype: Log

    • NEPA21: poring pulse (Pp): 50 V, 5 ms, 50 ms interval, 3 pulses, 10% decay (± pulse orientation) and transfer pulse (Tp): 10 V, 50 ms, 50 ms interval, 3 pulses, 40% decay (± pulse orientation)

    • T820: 50 V, 5 ms, 1-s interval, 8 square-wave pulses


  5. 26

    Remove any blood or debris from the electrodes by gently wiping them with an absorbent cotton soaked in 70% (vol/vol) ethanol (Fig. 7d).

  6. 27

    Remove the Kimwipe piece, the aorta clamp and the paper towel in this order (Fig. 8a–c).

    Fig. 8: Diagrams showing steps after in vivo electroporation (Steps 27–30).

    ac, Removal of the Kimwipe piece (a), the aorta clamp (b) and the paper towel (c) after electroporation. d, Returning the tissues (ovary, oviduct, uterus and adipose tissue) to their original positions. e, Closing the incision with suture wound clips.

  7. 28

    Return the tissues (ovary, oviduct, uterus and adipose tissue) to their original positions with forceps (Fig. 8d).

i-GONAD procedure on the other oviduct

Timing 3–7 min

  1. 29

    Repeat Steps 12–28 to perform the i-GONAD procedure on the other oviduct. The other oviduct can be accessed using the same skin incision. However, another incision of the muscle layer is required to expose the other oviduct.

Post-surgical procedure

Timing 15–30 min; 2–3 min hands-on

  1. 30

    After completing the i-GONAD procedure on both oviducts, hold the skin at the dorsal incision and close it with suture wound clips (Fig. 8e).

  2. 31

    Inject atipamezole hydrochloride via the i.p. or s.c. route (0.1 ml/10 g). Keep the mouse in the cage placed on the hot plate (40 °C) until it is awake (e.g., 10 min after injection of atipamezole hydrochloride), then proceed to Step 32.

    Critical step

    Provide mice with postsurgical analgesia according to your institution’s protocol.

Post-surgical housing

Timing 2–50 d

  1. 32

    House the i-GONAD-treated mice until the analysis of offspring either at embryonic stages or live offspring per the experimental design.

    Critical step

    We generally group-house females after surgery (and separate them just 1–2 d before the expected delivery to keep track of offspring from different females. Although we do not have a clear observation of single versus group housing, we presume single housing may reduce some stress caused by environmental/husbandry issues in some facilities.


Sampling and genotype analysis of offspring

Timing 2–4 d

  1. 33

    Recover the offspring from i-GONAD-treated females either at a fetal stage or at weaning of live-born mice per experimental design and analyze for genotype (or phenotypic changes, if the gene-editing experiment is expected to elicit a phenotype).

  2. 34

    Collect tissue samples (tip of tail/limb (2–3 mm long) from fetuses or ~2-mm ear pieces from newborn mice) in 1.5-ml microcentrifuge tubes.

  3. 35

    Add 40 µl of Allele-In-One Mouse Tail Direct Lysis Buffer and incubate the tube at 55 °C on a shaking heat block for >3 h or overnight.

  4. 36

    Incubate the tubes at 85 °C on a heat block for 45 min. Store the tissue lysates at −20 °C until use.

    Pause point

    The tissue lysates can be stored at −20 °C for years.

  5. 37

    Centrifuge at 21,000g for 2 min at RT and use for PCR reactions.

  6. 38

    Add PCR reagents, including primers, as shown in the table below, and perform PCR to amplify the target region of interest.

    Reagent Volume per reaction (µl) Volume for 10 reactions (µl) Final concentration
    2× GC buffer 5 50
    2.5 mM dNTP 0.8 8 200 µM
    F Primer (20 µM; Table 1) 0.25 2.5 0.5 µM
    R Primer (20 µM; Table 1) 0.25 2.5 0.5 µM
    TaKaRa Taq (5 U/µl) 0.025 0.25 0.125 U/10 µl
    Water 2.675 26.75  
    Genomic DNA (Step 37) 1   e.g., 100 ng
    Total 10   

    Critical step

    Here, we describe a standard PCR genotyping assay involving amplification of up to 800-bp fragments. Note that multiple sets of PCR/genotyping need to be performed, depending on the type of editing. See Miura et al.23 for an extensive description of genotyping of alleles generated using Easi-CRISPR.

  7. 39

    Run the PCR reactions in a thermocycler, using the following standard PCR conditions.

    Cycle no. Denature Anneal Extend
    1 95 °C for 5 min   
    2–36 95 °C for 45 s 58 °C for 30 s 72 °C for 1 min
    37 72 °C for 5 min   
    Hold at 4 °C    
  8. 40

    Run PCR products on a 1% (wt/vol) agarose gel electrophoresis along with control PCRs and DNA markers (GeneRuler DNA Ladder Mix) at 135 V for 30–60 min. Stain the gel with ethidium bromide (0.5 µg/ml) for 20 min. Analyze the gels using a gel documentation system.


    Ethidium bromide is carcinogenic and mutagenic. Wear personal protective equipment when handling this reagent.


  9. 41

    Sequence the PCR products to confirm the genotypes32.



Troubleshooting advice can be found in Table 3.

Table 3 Troubleshooting table


The approximate time needed for each step is given below. In general, the time required for the actual i-GONAD procedure (Steps 9–31) depends on the number of mice used. We generally prepare 3–5 female mice for an i-GONAD session, and all the procedures are completed within ~2–2.5 h (e.g., 4:00 pm to 6:30 pm), including some extra time for the experimental setup and for cleanup. The time needed to complete the i-GONAD procedure per mouse also depends on the skill level of the experimenter; it can range from 10 to 30 min per mouse.

  • Step 1, gRNA design: 30 min

  • Step 2, repair ssDNA template design: 1–3 h

  • Step 3, preparation of repair ssDNA template: 2–5 d

  • Steps 4–6, preparation of pregnant female mice: 2 d

  • Steps 7 and 8, preparation of genome-editing mixture: 1 h

  • Steps 9–16, surgical exposure of ovary, oviduct and part of the uterus: 10–20 min

  • Steps 17–21, intraoviductal injection of the genome-editing solution: 2–5 min

  • Steps 22–28, in vivo electroporation: 1–3 min

  • Step 29, i-GONAD procedure on the opposite oviduct: 3–7 min

  • Steps 30–31, post-surgical procedure: 15–30 min; 2–3 min hands-on time

  • Step 32, post-surgical housing: 2–50 d

  • Steps 33–41, sampling and genotype analysis of offspring: 2–4 d

  • Box 2, protocol for establishing i-GONAD in a new lab: 2 h

Anticipated results

The anticipated results for each of the major experimental stages of i-GONAD are described below.

Long ssDNA preparation

dsDNA template

We provided protocols for its preparation using PCR because the ssDNAs we have used in i-GONAD so far were prepared by this method, although plasmid DNA can also be used as the dsDNA template for RNA synthesis after linearization. The approximate yield of DNA from the PCR reaction (3× scale) should be 9–15 µg, depending on the target sequence and PCR conditions.

RNA synthesis

The amount of RNA synthesized is dependent on the kit used, but we obtain ~80–140 µg of RNA when using the T7 RiboMax Express Large Scale RNA Production System. RNA can be visualized on an agarose gel as a major single band. Minor bands that represent RNA secondary structures may sometimes appear.

cDNA synthesis

Use of a high enough concentration of cDNA (ssDNA) in the genome-editing solution is important for the success of i-GONAD. The anticipated amount of cDNA (long ssDNA) at a 10× scale when using SuperScript III or SuperScript IV reverse transcriptase is 5–8 µg, which can be used for two pregnant females (a total of four oviducts). Because gel purification of ssDNA is not performed, there is a possibility that the sample contains various lengths of ssDNA derived from incomplete synthesis (which may appear as a smear on a gel). Coexistence of such partially synthesized ssDNA can be problematic because it may cause unintended repair outcomes that can hamper insertion of full-length molecules. Therefore, the synthesized ssDNA should appear as a single band in the gel after the gel electrophoresis (Fig. 3; Table 3).

Preparation of pregnant females

In our experience, 52.5% (21/40) of C57BL/6J females assessed to be in estrus on the basis of smear examination had vaginal plugs after mating with C57BL/6J males. This efficiency may be improved further if experienced females (animals that have delivered pups previously) are used. If plugged females are identified correctly, we generally notice that >90% of plugged MCH(ICR) females retain pregnancy following the i-GONAD procedure.

Concentration of Cas9 and Cpf1

We suggest using a Cas9 protein concentration of 1 µg/µl for i-GONAD experiments, which is what we used in most of our experiments when we established the method. We tested lower Cas9 concentrations, using Tyr locus gene editing as an example, at 50, 100 and 500 ng/µl. We observed >40% knock-in efficiency in the Tyr locus when Cas9 protein was at 100 ng/µl (Table 4). However, 50 ng/µl Cas9 produced lower genome-editing efficiency, suggesting that at least 100 ng/µl Cas9 is required to obtain desirable levels of gene editing. Similarly, lower concentrations of Cpf1 may also work, even though we have tried only a 1 µg/µl concentration of Cpf1 so far.

Table 4 Tyrosinase rescue efficiencies in various concentrations of Cas9 protein


Mosaicism levels were generally lower when Cas9 protein was used in the form of RNP than when Cas9 mRNA was used (36–65% of mosaic pups/fetuses were obtained with Cas9 RNP, whereas 80% was obtained with Cas9 mRNA)18. Mosaicism can be further lowered, in theory, if i-GONAD can be performed at an earlier time point. However, considering that our experiments at day 0.4 of pregnancy produced lower gene-editing efficiencies (because zygotes are within tightly covered cumulus cells that presumably reduce electroporation efficiency), we suggest that day 0.7 of pregnancy offers the optimal time point for producing best gene-editing efficiencies while keeping mosaicism levels low. Regardless, mosaicism is not a major concern because founder animals generally transmit edited alleles to the next generation.

Injection of solution into the oviduct

Bleeding can occur occasionally during surgical exposure of the oviductal tissues. This can interfere with finding a suitable position on the oviduct to inject the solution. If bleeding occurs, the blood can be removed with a piece of paper towel before attempting injection of gene-editing solution.

Sometimes, the injection process can produce air bubbles in the oviductal lumen, especially after all the solution is administered. Although it is desirable to avoid blowing bubbles, the presence of bubbles may not affect the i-GONAD efficiency because we have obtained genome-edited fetuses even in those experiments in which air bubbles were inadvertently administered.

Electroporation conditions

We tested five electroporation conditions using Tyr gene correction in the MCH(ICR) strain as an example. Two animals were used for each electroporation condition (50, 100, 200, 300 or 400 mA). As shown in Supplementary Fig. 3, 50 mA did not generate genome-edited offspring, even though we recovered as many as 31 fetuses from the two females. The genome-editing efficiency reached 81% (13/16) at 100 mA and 100% from 200 mA onward (10/10 in 200 mA, 2/2 in 300 mA, 1/1 in 400 mA). However, the yield of fetuses was lower when the current was increased. On the basis of these results, we conclude that electroporation currents of 100–200 mA produce the most optimal results in the MCH(ICR) strain. Electroporation conditions can vary from strain to strain in regard to the number of offspring obtained. For example, we observed a narrow range of current (100 mA versus 150 mA) in which the offspring number was reduced by more than fourfold in the C57BL/6J strain (Table 5). We observed that a 100-mA current produces the most consistent and optimal results in the C57BL/6 strain. We obtained an overall efficiency of 96% indel mutations in the Tyr locus using this electroporation condition (Table 6). Therefore, we suggest that electroporation conditions may require optimization for each mouse strain. We suggest a range from 50-mA to 200-mA current as a starting point for any given strain.

Table 5 Effect of current value on embryo recovery in C57BL/6J mice
Table 6 i-GONAD efficiency in C57BL/6J mice

Genome-editing efficiencies

The genome-editing efficiencies we obtained with the i-GONAD method typically ranged from 35 to 100%, 15 to 60% and 7 to 15% for generating simple indel, ssODN knock-in and long ssDNA knock-in models, respectively18. Note that the data for indels and ssODN knock-ins were derived from a large number of experiments performed on several loci (five loci and >85 mice for indels; four loci and 65 mice for ssODN knock-ins), whereas only two loci were tested using five females for long ssDNA knock-ins. In general, success with i-GONAD depends on gRNA cleavage efficiency and the successful delivery of CRISPR components.

Cas9 protein (RNP) formulations produce higher gene-editing efficiencies than Cas9 mRNA (e.g., 97% versus 31%, as observed for Foxe3 locus gene editing). Although the overall efficiencies are somewhat similar across many strain backgrounds, some inbred mouse strains require the use of more females for the i-GONAD procedure to generate gene-edited mice because of the smaller litter sizes in those inbred strains. The i-GONAD efficiencies across many mouse strains are presented in our recent paper18. Of note, the i-GONAD method could be particularly useful for some strains in which artificial reproductive technologies (ARTs) and/or generation of transgenic mice with conventional methods are challenging. The i-GONAD method has the potential to generate genome-edited mice in the BALB/c strain, which is known to be a challenging strain for the generation of transgenic mice18,33.

Genome-editing efficiency in i-GONAD is comparable to that obtained by zygote injection. For example, ssODN knock-in efficiency in the Tyr locus using Cas9 protein was 49% and 52% using i-GONAD and zygote injection approaches, respectively. Cpf1 can also be used in i-GONAD, yielding 58% ssODN knock-in efficiencies in the Tyr rescue experiment with the MCH(ICR) strain18.

Reporting Summary

Further information on research design is available in the Nature Research Reporting Summary linked to this article.

Data availability

All the data generated and analyzed in this study are included in the tables, figures and supplementary material.


  1. 1.

    Gordon, J. & Ruddle, F. Integration and stable germ line transmission of genes injected into mouse pronuclei. Science 214, 1244–1246 (1981).

    CAS  Article  Google Scholar 

  2. 2.

    Evans, M. J. & Kaufman, M. H. Establishment in culture of pluripotential cells from mouse embryos. Nature 292, 154–156 (1981).

    CAS  Article  Google Scholar 

  3. 3.

    Smithies, O., Gregg, R. G., Boggs, S. S., Koralewski, M. A. & Kucherlapati, R. S. Insertion of DNA sequences into the human chromosomal β-globin locus by homologous recombination. Nature 317, 230–234 (1985).

    CAS  Article  Google Scholar 

  4. 4.

    Thomas, K. High frequency targeting of genes to specific sites in the mammalian genome. Cell 44, 419–428 (1986).

    CAS  Article  Google Scholar 

  5. 5.

    Doyle, A., McGarry, M. P., Lee, N. A. & Lee, J. J. The construction of transgenic and gene knockout/knockin mouse models of human disease. Transgenic Res. 21, 327–349 (2012).

    CAS  Article  Google Scholar 

  6. 6.

    Skarnes, W. C. et al. A conditional knockout resource for the genome-wide study of mouse gene function. Nature 474, 337–342 (2011).

    CAS  Article  Google Scholar 

  7. 7.

    Geurts, A. M. et al. Knockout rats via embryo microinjection of zinc-finger nucleases. Science 325, 433 (2009).

    CAS  Article  Google Scholar 

  8. 8.

    Bogdanove, A. J. & Voytas, D. F. TAL effectors: customizable proteins for DNA targeting. Science 333, 1843–1846 (2011).

    CAS  Article  Google Scholar 

  9. 9.

    Hsu, P. D., Lander, E. S. & Zhang, F. Development and applications of CRISPR-Cas9 for genome engineering. Cell 157, 1262–1278 (2014).

    CAS  Article  Google Scholar 

  10. 10.

    Gurumurthy, C. B. et al. GONAD: a novel CRISPR/Cas9 genome editing method that does not require ex vivo handling of embryos. Curr. Protoc. Hum. Genet. 88, 15.8 (2016).

  11. 11.

    Harms, D. W. et al. Mouse genome editing using the CRISPR/Cas system. Curr. Protoc. Hum. Genet. 83, 15.7 (2014).

    Google Scholar 

  12. 12.

    Teixeira, M. et al. Electroporation of mice zygotes with dual guide RNA/Cas9 complexes for simple and efficient cloning-free genome editing. Sci. Rep. 8, 474 (2018).

    Article  Google Scholar 

  13. 13.

    Tröder, S. E. et al. An optimized electroporation approach for efficient CRISPR/Cas9 genome editing in murine zygotes. PLoS ONE 13, e0196891 (2018).

    Article  Google Scholar 

  14. 14.

    Chen, S., Lee, B., Lee, A. Y. F., Modzelewski, A. J. & He, L. Highly efficient mouse genome editing by CRISPR ribonucleoprotein electroporation of zygotes. J. Biol. Chem. 291, 14457–14467 (2016).

    CAS  Article  Google Scholar 

  15. 15.

    Qin, W. et al. Efficient CRISPR/cas9-mediated genome editing in mice by zygote electroporation of nuclease. Genetics 200, 423–430 (2015).

    CAS  Article  Google Scholar 

  16. 16.

    Hashimoto, M., Yamashita, Y. & Takemoto, T. Electroporation of Cas9 protein/sgRNA into early pronuclear zygotes generates non-mosaic mutants in the mouse. Dev. Biol. 418, 1–9 (2016).

    CAS  Article  Google Scholar 

  17. 17.

    Takahashi, G. et al. GONAD: Genome-editing via Oviductal Nucleic Acids Delivery system: a novel microinjection independent genome engineering method in mice. Sci. Rep. 5, 11406 (2015).

    Article  Google Scholar 

  18. 18.

    Ohtsuka, M. et al. i-GONAD: a robust method for in situ germline genome engineering using CRISPR nucleases. Genome Biol. 19, 25 (2018).

    Article  Google Scholar 

  19. 19.

    Ohtsuka, M. et al. Pronuclear injection-based mouse targeted transgenesis for reproducible and highly efficient transgene expression. Nucleic Acids Res. 38, e198 (2010).

    Article  Google Scholar 

  20. 20.

    Behringer, R., Gertsenstein, M., Nagy, K. V. & Nagy, A. Manipulating the Mouse Embryo: A Laboratory Manual, 4th edn. (Cold Spring Harbor Laboratory Press, 2014).

  21. 21.

    Quadros, R. M. et al. Easi-CRISPR: a robust method for one-step generation of mice carrying conditional and insertion alleles using long ssDNA donors and CRISPR ribonucleoproteins. Genome Biol. 18, 92 (2017).

    Article  Google Scholar 

  22. 22.

    Miura, H., Gurumurthy, C. B., Sato, T., Sato, M. & Ohtsuka, M. CRISPR/Cas9-based generation of knockdown mice by intronic insertion of artificial microRNA using longer single-stranded DNA. Sci. Rep. 5, 12799 (2015).

    CAS  Article  Google Scholar 

  23. 23.

    Miura, H., Quadros, R. M., Gurumurthy, C. B. & Ohtsuka, M. Easi-CRISPR for creating knock-in and conditional knockout mouse models using long ssDNA donors. Nat. Protoc. 13, 195–215 (2018).

    CAS  Article  Google Scholar 

  24. 24.

    Kobayashi, T. et al. Successful production of genome-edited rats by the rGONAD method. BMC Biotechnol. 18, 19 (2018).

    Article  Google Scholar 

  25. 25.

    Takabayashi, S. et al. i-GONAD (improved genome-editing via oviductal nucleic acids delivery), a convenient in vivo tool to produce genome-edited rats. Sci. Rep. 8, 12059 (2018).

    Article  Google Scholar 

  26. 26.

    Miano, J. M., Zhu, Q. M. & Lowenstein, C. J. A CRISPR path to engineering new genetic mouse models for cardiovascular research. Arterioscler. Thromb. Vasc. Biol. 36, 1058–1075 (2016).

    CAS  Article  Google Scholar 

  27. 27.

    Haeussler, M. et al. Evaluation of off-target and on-target scoring algorithms and integration into the guide RNA selection tool CRISPOR. Genome Biol. 17, 148 (2016).

    Article  Google Scholar 

  28. 28.

    Labun, K., Montague, T. G., Gagnon, J. A., Thyme, S. B. & Valen, E. CHOPCHOP v2: a web tool for the next generation of CRISPR genome engineering. Nucleic Acids Res. 44, W272–W276 (2016).

    CAS  Article  Google Scholar 

  29. 29.

    Oliveros, J. C. et al. Breaking-Cas—interactive design of guide RNAs for CRISPR-Cas experiments for ENSEMBL genomes. Nucleic Acids Res. 44, W267–W271 (2016).

    CAS  Article  Google Scholar 

  30. 30.

    Aida, T. et al. Cloning-free CRISPR/Cas system facilitates functional cassette knock-in in mice. Genome Biol. 16, 87 (2015).

    Article  Google Scholar 

  31. 31.

    Morrow, D. A. Prognostic value of serial B-type natriuretic peptide testing during follow-up of patients with unstable coronary artery disease. JAMA 294, 2866–2871 (2005).

    CAS  Article  Google Scholar 

  32. 32.

    Ma, H. & Difazio, S. An efficient method for purification of PCR products for sequencing. Biotechniques 44, 921–923 (2008).

    CAS  Article  Google Scholar 

  33. 33.

    Dinkel, A. et al. Efficient generation of transgenic BALB/c mice using BALB/c embryonic stem cells. J. Immunol. Methods 223, 255–260 (1999).

    CAS  Article  Google Scholar 

  34. 34.

    Dong, L., Lv, L. B. & Lai, R. [Molecular cloning of Tupaia belangeri chinensis neuropeptide Y and homology comparison with other analogues from primates]. Zool. Res. 33, 75–78 (2012).

    CAS  Article  Google Scholar 

Download references


We thank Y. Ishikawa (Tokai University) for preparing pregnant MCH(ICR) mice and BEX Co. Ltd for advising us about electroporation optimization. This work was supported by the 2014 Tokai University School of Medicine Research Aid, the Research and Study Project of Tokai University General Research Organization, the 2016–2017 Tokai University School of Medicine Project Research, the Research Aid from the Institute of Medical Sciences inTokai University, the MEXT‐Supported Program for the Strategic Research Foundation at Private Universities 2015–2019, and a Grant-in-Aid for Challenging Exploratory Research (15K14371) from JSPS to M.O. AsCpf1 protein was a gift fromIDT. We thank J. M. Miano (University of Rochester) and G. Burgio (Australian National University) for their helpful comments on the manuscript.

Author information




M.O., C.B.G. and M.S. conceived the idea of i-GONAD. M.O., M.S. and M.A.I. were involved in testing the 2-d protocol. H.M. established the long ssDNA preparation system. M.I. and N.K. tested the hand-made capillary-based i-GONAD method. M.O., A.N. and S.O. optimized i-GONAD in the C57BL/6 strain. S.N. tested the squeezing steps in the i-GONAD procedure. M.O., S.T. and M.M. performed and optimized i-GONAD in several mouse strains and electroporators. C.B.G., M.S., M.I., M.A.I., S.N., H.M. and M.O. contributed to the writing of the manuscript.

Corresponding authors

Correspondence to Channabasavaiah B. Gurumurthy or Masato Ohtsuka.

Ethics declarations

Competing interests

C.B.G., M.S. and M.O. have filed a patent application relating to the work described in this paper with International Patent Application no. PCT/US2018/047748, filed August 23, 2017 (Methods and compositions for in situ germline genome engineering). Tokai University and BEX Co. Ltd. applied for a patent describing the electroporation condition using CUY21Edit II on application number 2017–233100 (filed December 5, 2017). M.O. is an inventor of the patent.

Additional information

Peer review information: Nature Protocols thanks Izuho Hatada, Michael Wiles and other anonymous reviewer(s) for their contribution to the peer review of this work.

Publisher’s note: Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Related links

Key references using this protocol

Ohtsuka, M. et al. Genome Biol. 19, 25 (2018):

Takahashi, G. et al. Sci. Rep. 5, 11406 (2015):

Miura, H., Quadros, R. M., Gurumurthy, C. B. & Ohtsuka, M. Nat. Protoc. 13, 195–215 (2018):

Integrated supplementary information

Supplementary Fig. 1 Preparation of devises used for intraoviductal injection.

(a) Cartoon showing various parts of a mouth pipetting devise. (b) Cartoon showing glass capillaries needed for intraoviductal injection. Drawings of a capillary tip before and after it was cut with the microscissors.

Supplementary Fig. 2 Dissecting microscopic view of oviduct after injection of solution containing Fast Green FCF.

Images of the oviduct photographed immediately after injection (a), and after covering it with a wet Kimwipe towel (b).

Supplementary Fig. 3 Genome-editing efficiency and number of fetuses retrieved after performing i-GONAD with different electroporation currents.

Tyr rescue experiment is shown as an example. Two MCH(ICR) females were used for each current value. Gray bar indicates a total number of mid-gestation fetuses recovered from two females. Yellow bar indicates a total number of genome modified fetuses (includes both knock-in and indel). Blue bar indicates a total number of fetuses with pigmentation (successful knock-in). Green line indicates the % of genome edited fetuses. The data of this experiment are available from the corresponding author upon request.

Supplementary Fig. 4 Injection of solution using hand-made glass capillary.

(a) Dissecting microscopic view of injection (see also Supplementary Video 3). (b) Tip of hand-made glass capillary.

Supplementary Fig. 5 Cartoons showing collection of 2-cell-stage embryos from the i-GONAD-treated mouse (procedure 1 of 2).

(a) Making a midline incision at the ventral surface of the euthanized mouse. (b) Opening the ventral skin. (c) Opening the ventral muscle. (d) Pulling out of both ovaries/oviducts/uteri and trimming off the portion of the uterus. (e) Placing of the dissected ovaries/oviducts/uteri and adding a small amount of 1x PBS onto the tissues. (f) Removing of the associated blood on ovary/oviduct/uterus by placing them onto a paper towel. (g) Removing the mesenterium (a membranous fold) associated with the uterus and oviduct. (h) Removing the adipose tissue associated with an ovary. (i) Removing the ovary using microscissors under a dissecting microscope. R, right; L, left.

Supplementary Fig. 6 Cartoons showing collection of 2-cell-stage embryos from the i-GONAD-treated mouse (procedure 2 of 2).

(a) and (b) Cutting of the junctional portion between oviduct and uterus using microscissors. (c) Inserting the 30-G needle attached with a 1-ml disposable plastic syringe into the oviductal lumen. (d) Flushing the contents while watching the flowing out of embryos. (e) Removing the oviduct and collecting the released embryos using an egg handling pipette. The picture above the drops (top right) is the side view of the embryo handling procedure in the drop. Washing of the collected embryos by passing them through small drops, and (f) transferring them into another fresh drop ready for observing under an inverted fluorescence microscope. Embryos obtained from four females (#1 to #4) are shown as an example. R, embryos from right oviduct; L, embryos from left oviduct. Right most image shows a side view of microscopic observation.

Supplementary Fig. 7 Fluorescence in the preimplantation mouse embryos after i-GONAD.

(a) - (c) Two-cell embryos collected from the MCH(ICR) females that were subjected to i-GONAD procedure following injection of a solution containing tetramethylrhodamine-labeled dextran 3kDa (red), EGFP mRNA (green). Note that the embryos are observed for both red and green fluorescence. (d) to (f) Two-cell embryos collected from the MCH(ICR) females similarly treated except the electroporation step showing no fluorescence in any of the embryos. Scale bar = 100 µM.

Supplementary information

Supplementary Information

Supplementary Figs. 1–7

Reporting Summary

Supplementary Video 1

Surgical exposure of an oviduct.

Supplementary Video 2

Loading solution into a capillary needle.

Supplementary Video 3

Injecting solution into an oviduct by using a hand-made capillary with injection choice 1 (Fig. 6).

Supplementary Video 4

Injecting solution into an oviduct and then performing electroporation on the oviduct with an NEPA21 electroporator and CUY652-3 electrodes (2-d protocol) using injection choice 1 (Fig. 6).

Supplementary Video 5

Squeezing of the ampulla to disperse the injected solution uniformly within the lumen.

Supplementary Video 6

Injecting solution into an oviduct and then performing electroporation on the oviduct with a CUY21EditII electroporator and LF650P3 electrodes using injection choice 1 (Fig. 6).

Supplementary Video 7

Preparing a glass micropipette with a flame.

Supplementary Video 8

Cutting the tip of a glass capillary.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Gurumurthy, C.B., Sato, M., Nakamura, A. et al. Creation of CRISPR-based germline-genome-engineered mice without ex vivo handling of zygotes by i-GONAD. Nat Protoc 14, 2452–2482 (2019).

Download citation

Further reading


By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.


Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing