A subscription to JoVE is required to view this content. Sign in or start your free trial.

Summary

The present protocol describes a single M213L mutation in Gja1 that retains full-length Connexin43 generation but prevents translation of the smaller GJA1-20k internally translated isoform.

Abstract

The CRISPR-Cas9 gene-editing system, based on genome repair mechanisms, enables the generation of gene-modified mouse models more quickly and easily relative to traditional homologous recombination. The CRISPR-Cas9 system is particularly attractive when a single-point mutation is desired. The gap junction protein, Connexin 43 (Cx43), is encoded by gene Gja1, which has a single coding exon and cannot be spliced. However, Gja1 produces not only full-length Cx43 protein but up to six N-terminus truncated isoforms by a process known as internal translation, the result of ribosomal translation initiation at internal AUG (Methionine) start sites. GJA1-20k is the most commonly generated truncated isoform of Cx43 initiated at the AUG codon at position 213 of Gja1 mRNA. Because residue 213 occurs at the end of the last transmembrane domain of Cx43, GJA1-20k is effectively the 20 kDa C-terminus tail of Cx43 as an independent protein. Previous investigators identified, in cells, that a critical role of GJA1-20k is to facilitate trafficking of full-length Cx43 gap junction hemichannels to the plasma membrane. To examine this phenomenon in vivo, a mutant mouse with a Gja1 point-mutation was generated that replaces the ATG (Methionine) at residue 213 with TTA (Leucine, M213L mutation). The result of M213L is that Gja1 mRNA and full-length Cx43 are still generated, yet the translation of Gja1-20k is significantly reduced. This report focuses on choosing the restriction enzyme site to develop a one amino acid mutated (Gja1^M213L/M213L) mouse model. This protocol describes genetically modified mice by the CRISPR-Cas9 system and rapid genotyping by combining PCR and restriction enzyme treatments.

Introduction

The full-length Connexin 43 (Cx43) and the N-terminus truncated isoform, GJA1-20k, encoded by the same GJA1 mRNA but utilize different start codons¹ to initiate translation. Cx43 translation occurs at the first AUG start codon, whereas GJA1-20k translation initiates at the AUG at residue 213. It was previously found that GJA1-20k has essential roles for full-length Cx43 trafficking, actin stabilization, and regulation of mitochondrial morphology in vitro^1,2,3.

To understand the role of GJA1-20k in vivo, a GJA1-20k "knock-out" mouse model was generated that retained the ability to create full-length Cx43. The approach was to use the CRISPR-Cas9 system to substitute the single residue at 213 from a Methionine (M) to a Leucine (L) (Gja1^M213L/M213L)⁴. An internal M to L mutation dramatically decreases the likelihood of internal translation occurring yet retains the translation and function of full-length protein⁴. Because the wild type (WT) and mutated allele have identical sizes and near-identical mRNA products, there is considerable difficulty in confirming genotype in the mice. DNA sequencing can identify the mutation but is too expensive and time-consuming for routine use. In general, several faster-genotyping methods have been established, such as Real-time Polymerase Chain Reaction (RT-PCR), minisequencing-ligation, high-resolution melting analysis, and tetra-primer amplification refractory mutation system PCR (ARMS-PCR)^5,6,7,8. Yet these alternative methods require multiple steps, unique resources, and/or several specific primer sets which can induce non-specific PCR products.

This protocol introduces a detailed gene targeting and editing approach by CRISPR-Cas9 to create a single amino acid mutation, and rapid genotyping is presented to confirm the mutation. Genotype identification involves the creative use of restriction enzymes utilizing only a single set of primers to identify the target gene. Readers are referred to Reference⁴ to observe the profound electrophysiological effect of sudden cardiac death caused by a one residue Gja1^M213L/M213L substitution mutation which still generates full-length protein yet fails to generate a smaller internally translated truncation isoform. This protocol will help make other mice models use a point mutation to decrease the internal translation of an isoform of interest while retaining the expression of the endogenous full-length protein.

Protocol

`All animal care and study protocols were approved by the Institutional Animal Care and Use Committees of Cedars-Sinai Medical Center and the University of Utah. C57BL/6J female mice obtained from commercial sources at 8-9 weeks of age (see Table of Materials) were used for the experiments.

1. Preparation for gene targeting

1. Select the guide target sequence around the targeted mutation site coding M213 using CRISPOR web algorithm^4,9,10 (see Table of Materials).
   NOTE: A 20-base guide sequence (ATTCAGAGCGAGAGACACCA), in the opposite strand from the GJA1 coding sequence, with an MIT score¹¹ of 62 was selected in which the potential cleavage site is located after 16 bases downstream of the codon to be mutated (ATG) (Figure 1; the whole crRNA sequence is shown in Table 1).
2. Synthesize the crRNA and tracrRNA that interact with Cas9 as guide RNA¹².
   NOTE: Although the MIT score of the guide RNA is 62, there is only one potential off-target mutation with four mismatches more than 12 bases away from the PAM. Therefore, this guide RNA is considered safe.
3. Design a donor oligo complementary to the guide sequence to introduce a single amino acid substitution (ATG to TTA; M213L) with 60 and 48 bases homology arms in the 5'- and 3'-sides, respectively.
   NOTE: This donor oligo includes a point mutation for disruption of the PAM (AGG to ACG) by introducing a silent mutation (TCC to TCG; S217S) to avoid re-editing after CRISPR homology-directed repair (HDR)¹³ for the introduction of the intended mutations (Figure 1 and Table 1).
4. Use NaCl (200 mM final concentration) to precipitate 10 µg of oligos to minimize salt carry-over into pronuclear microinjection buffer (10 mM of Tris-HCl, pH 7.5; 0.1 mM of EDTA, and 100 mM of NaCl without spermine and spermidine¹⁴).
   NOTE: Do not wash the DNA pellet with 70% ethanol as the pellet can be dissolved into the solution.
5. Mix 50 ng/µL of donor oligo, 60 ng/µL of crRNA/tracrRNA mix (1:1 molar ratio) (from step 1.2), and 50 ng/µL of eSpCas9 protein (see Table of Materials) to make CRISPR mixture in final volume 20 µL (Table 2). Decrease the concentration of the donor oligo if high toxicity is observed (e.g., 25 ng/µL).

2. Induction of superovulation, harvesting eggs, pronuclear microinjection of CRISPR mix, and blastocyst screening

NOTE: This procedure follows a previously published general protocol¹⁴.

1. Inject 5 IU of PMSG (pregnant mare's serum gonadotropin) (see Table of Materials) in 100 µL of sterile water to 5-10 mice at 8 weeks of age by an intraperitoneal approach at about 3:00 p.m. (day 1).
2. Inject 5 IU of hCG (human chorionic gonadotropin) (see Table of Materials) in 100 µL of sterile water to the egg donors by an intraperitoneal approach 46-48 h after PMSG injection (day 3). Set up 1:1 breeding with stud males.
3. Harvest fertilized eggs in M2 medium with hyaluronidase, wash three times in 100 µL of M2 medium around 10:00 a.m., then keep in mWM¹⁵ or KSOM¹⁶ medium (day 4) (see Table of Materials).
4. Introduce the CRISPR mixture (step 1.3) into 1-cell stage embryos harvested from the mice by pronuclear microinjection¹⁴ in 100 µL of M2 covered with paraffin oil in a plastic dish on an inverted microscope with a contrast-enhancing optics in early to late afternoon (Video 1).
   NOTE: Pronuclear microinjection of the CRISPR mixture was performed on 1-cell stage embryos at 14-16 h postcoitum (p.c.). The injected embryos were cultured in a 5% CO₂ incubator for a few hours and surgically transferred at 18-20 h p.c. into recipient ICR mice that had a copulation plug in the same morning.
5. Transfer 20-30 2-cell embryos into each recipient to produce recombinant animals.
   NOTE: Follow the general protocol¹⁴ or culture those embryos in mWM or KSOM medium for 4 days to validate the efficiency and examine toxicity. Toxicity of the CRISPR mixture is acceptable when at least one-third of the injected embryos reach early to the fully expanded blastocyst stage with multiple recombinants among them. At the same time, more than 90% of unmanipulated embryos reach the same blastocyst stage in a parallel culture. Follow the step 2.6-2.9 for blastcyst screening which is not essential but can be done if preferred to quantify success rate of HDR.
6. Individually pick up early to fully expanded blastocysts with 2 µL of culture medium using a plugged fine pipette tip with a micropipetter into single PCR tubes (day 8; Video 2).
7. Lyse blastocysts in 8 µL of digestion mixture (same as step 3.2; see Table of Materials) and process at 75 °C for 10 min, 95 °C for 5 min, then cool down to 4 °C for storage. Use 2 µL of the lysate as the template for PCR in a total 15 µL PCR mixture (step 3).
8. Examine the efficacy of the HDR by agarose gel electrophoresis¹⁷ of the PCR samples following restriction digestion with NlaIII (step 3-5).
9. Confirm the status of targeted recombination by sequencing the 668 bp amplicon¹⁷ in the founders and subsequent progeny to confirm the integrity of the mutation.

3. DNA extraction

NOTE: Mice at postnatal day 10 were used for this experiment due to sudden death of GJA1-20k knock-out mouse around 2-4 weeks after birth⁴. A more general protocol also can be applied following previously published report¹⁸.

1. Cut 1-3 mm of the toe or tail tip with clean scissors and transfer it to a 0.2 mL 8-Strip PCR tube. To avoid contamination among the toe or tail samples, use pre-cleaned scissors or one blade per mouse or clean prior to each sample using 70% ethanol or 10% bleach. If the tail bleeds post sampling, apply brief pressure with gauze to stop the bleed. Store the tail samples at -20 °C until the extraction for up to about a week.
2. Add tissue lysis solution (see Table of Materials) in a 100 µL/tube and mix well, followed by spin-down with a tabletop mini-centrifuge (2,200 x g for 10 s at room temperature). Ensure the tail samples are submerged in the solution.
3. Set the tubes onto a thermal cycler set by the following program; 75 °C for 10 min (tissue lysis), 95 °C for 5 min (inactivation), and 4 °C (holding). Store the tissue lysate at 4 °C for up to a week if PCRs cannot be performed immediately.

4. DNA amplification by PCR

1. Prepare a PCR solution containing 5 µL of nuclease-free H₂O, 0.75 µL of 10 µM forward and reverse Primers, and 7.5 µL of PCR master mix (see Table of Materials) per sample into new PCR tube (see Table 1 for primer sequences). If there are several samples, scale contents to aliquot 14 µL per tube.
2. Add 1 µL of the tissue lysate to the PCR solution prepared in step 3.1. Be careful not to touch the tail.
3. Mix well and spin down with a tabletop mini-centrifuge (2,200 x g for 10 s at room temperature).
4. Set the tubes onto a thermal cycler set by the below-mentioned program. Store the PCR products at 4 °C for 1-2 months or ~1 year at -20 °C.
   ​NOTE: 95 °C for 3 min (step 1, Initial Denaturation), 95 °C for 15 s (step 2, Denaturation), 60 °C for 15 s (step 2, Annealing), 72 °C for 45 s (step 2, Extension), repeat step 2 for 35 cycles, 72 °C for 10 min (step 3, Final Extension), and 4 °C (step 4, Holding).

5. Incubation with restriction enzyme

1. Prepare the enzyme solution containing 7 µL of nuclease-free H₂O, 2 µL of 10x CutSmart buffer, and 1 µL of NlaIII restriction enzyme (see Table of Materials). If there are several samples, multiply each content to make the mixture and aliquot 10 µL per tube.
2. Add 10 µL of PCR product obtained in step 3 to the enzyme solution per tube.
3. Mix well and spin down with a tabletop mini-centrifuge (2,200 x g for 10 s at room temperature).
4. Set the tubes onto a thermal cycler or heat block set by the below-mentioned program. Store the product after the incubation at 4 °C for 1-2 months or ~1 year at -20 °C.
   ​NOTE: 37 °C for 16 h (at least 2 h; short incubation time may result in insufficient cleavage) and 4 °C for holding.

6. DNA band detection

1. Prepare a 1.5% agarose gel containing DNA gel stain for electrophoresis.
   1. Add 0.75 g agarose in 50 mL of 1x TAE buffer followed by mix and heat in microwave until agarose dissolves completely. After cooling down, add 5 µL of DNA stain and gently mix.
   2. Pour into the gel mold (25 mL per mold) and allow the gel to solidify. To obtain better resolution and separation, increase the agarose concentration to 2.5%-4% as needed.
2. Load 10 µL of digested PCR product to the well. Mix 6x loading buffer, if necessary.
3. Run the gel with 100 V for 35 min.
   NOTE: These parameters may need optimization.
4. Image under UV light (302 nm wavelength).

Representative Results

The CRISPR/Cas9 gene-editing system produces an ATG to TTA mutation and a silent TCC to TCG mutation at 56,264,279 to 56,264,281 and at 56,264,291 to 56,264,293 on mouse chromosome 10, or at 869 to 871 and 881 to 883 on Gja1 mRNA, respectively. Those mutation results in a Methionine 213 to Leucine (M213L) mutation on GJA1 protein and the TTC to TTG mutation disrupts a nearby PAM sequence to avoid undesired gene edition (Figure 1). The details of the mutation are described in a previous report⁴.

To analyze proper genotypes, the PCR products need to be digested by restriction enzyme "NlaIII" before agarose gel electrophoresis (Figure 2, step 4). As shown in Figure 2A, wild-type (WT) allele PCR products will be cut into two different base pair products (227 bp and 441 bp). In contrast, mutant allele PCR products will be intact (668 bp) due to the lack of NlaIII recognition site by mutation (Figure 2B). General agarose electrophoresis can be performed using the digested PCR products (step 5). Specific bands in the agarose gel indicate each genotype (WT, wild type; Het, heterozygous; Hm, homozygous). Due to the restriction mentioned above, each genotype results in several bands of different sizes (Figure 2C). The bands will be two bands in WT (227 bp and 441 bp), three bands in Het (227 bp, 441 bp, and 668 bp), and one band in Hm (668 bp), respectively. According to the gel imaging using a single restriction enzyme, proper genotypes of the mice can be distinguished.

In our experiment, twenty-one founders were initially produced, of which nine were heterozygous or mosaic animals. Two out of the nine founders died before growing up to breeding age, most likely due to high-content mosaicism of bi-allelic mutant somatic cells. Two independent mutant lines were subsequently established from the remaining seven founders by backcrossing to wild-type C57BL/6J mice for at least two generations. Because the targeted mutation is lethal when homozygous, we had to maintain the line as heterozygous for further diluting possible off-target mutations. Homozygous experimental animals were produced by intercrossing subsequent generations.

It is worth noting that relatively high embryonal toxicity with the donor oligos at 50 ng/µL was experienced for injection of the donor oligos into the embryos. However, a donor concentration of 12.5 ng/µL didn't induce the targeted recombination in the resulting pups. Additionally, it was found that added ethanol precipitation of the oligos helped remove contaminants and allowed induction of efficient, targeted genome editing. Note that washing the DNA pellet with 70% ethanol dissolves the pellet into a solution. Therefore, NaCl was used at 200 mM final concentration to precipitate 10 µg of oligos to minimize salt carry-over into pronuclear microinjection buffer. Test injection with blastocyst analysis was attempted using 50 or 25 ng/µL of the donor oligos with ethanol precipitation. The success rate of each condition was 76.9% (10 of 13 samples with 50 ng/µL) and 25.0% (3 of 12 samples with 25 ng/µL) (Figure 3A–C). Although 50 ng/µL of oligo with ethanol precipitation was still somewhat toxic, many eggs needed to be injected, and recombinants were produced at a high success rate. Therefore, 50 ng/µL of the donor oligos were finally injected with ethanol precipitation and obtained founder animals with a 50.0% success rate (16 of 32 pups had recombination including 21 survivors and 11 death after birth). Two of the founder animals having germline transmission of uniform mutation or monoallelic or identical mutation with stable transmission confirmed by sequencing were used to establish independent lines. The consensus is that backcrossing two generations with wild-type animals and using enhanced-fidelity Cas9 protein eliminates major off-targeting events. The number of such events is not statistically distinguishable from the background rate of de novo mutations^19,20.

Animals with undesired genotypes (wild type or improper recombinants) used for experiments and at the study's conclusion were euthanized by CO₂ inhalation followed by cervical dislocation according to the IACUC protocol.

Figure 1: Schematic representation for the gene targeting by CRISPR/Cas9. (A) The targeted sequences for homologs recombination. Single-letter amino acids are shown between the sequences. Colored highlights indicate homology arm (Gray), Mutation codon (Green), PAM sequence (Orange), and gRNA target sequence (Blue). The blue arrow indicates gRNA. NlaIII restriction site is underlined by dotted line. (B) The schematic sequences of WT and mutant allele and donor oligo for the point mutation resulted in M213L mutation. TCC to TCG is a silent mutation for disruption of undesired PAM. Please click here to view a larger version of this figure.

Figure 2: Schematic PCR products and restriction site of WT and the mutant allele. (A,B) Scheme of the PCR and the digestion by NlaIII. The primers amplify the 668 bp of Gja1 DNA around the mutation site. The WT allele PCR products were digested to two different sizes by NlaIII. In contrast, the mutant allele is NOT digested and maintains intact PCR products. (C) The representative image of the agarose gel indicates WT and mutant allele. Each mouse genotype had a specific band size; WT (lane 1, 3, and 5), Het (lane 2), and Hm (lane 4). Please click here to view a larger version of this figure.

Figure 3: The recombination efficiency in two different doses of donor oligos in the CRISPR mixture. (A) Genotyping results of morulae or blastocysts injected with different doses (25 or 50 ng/µL) of donor oligos showing the wild type (WT; lane #10, 12, and 14), heterozygous/mosaic (het/mosaic; lane # 1, 2, 3, 6, 8, 9, 11, and 13), homozygous (hom; lane #4 and 5) or not amplified (lane #7).Wild Type positive control and H₂O as negative control were loaded between lane #17 and #18. (B–D) The pie charts represent the recombination efficiency of CRISPR mixtures in two different doses of donor oligos (B,C) and that of founder animals injected with 50 ng/µL of donor oligos (D). Note that 11 out of 32 animals born were either killed by the mothers or died before weaning. Samples that failed to amplify in PCR were excluded in the calculation of recombination efficiency. The individual genotype rate from all samples is indicated in Table 3 and Table 4. Please click here to view a larger version of this figure.

Name	Sequences			Comments
crRNA	5' AUUCAGAGCGAGAGACACCAGU UUUAGAGCUAUGCUGUUUUG 3'			Underline indicates CRSPR target sequence
Oligo donar DNA	5' CCCCACCAGGTGGACTGCTTCCTCTCACG TCCCACGGAGAAAACCATCTTCATCATCTTCT TACTGGTGGTGTCGTTGGTGTCTCTCGCTCT GAATATCATTGAGCTCTTCTATGTCTTCTTC 3'			Underline indicates mutated codons. Fist 60 bp before TTA mutaion and 48 bp after TCG mutaion are homology arm
Genotyping primer forward	5' TGGGATTGAAGAACACGGCA 3'
Genotyping primer Reverse	5' CCACGATAGCTAAGGGCTGG 3'

Table 1: Sequence information.

Component	Stock solution	Amount	Final Concentration
Donor oligo	1µg/µL in Rnase-free H2O	1.0 µL	50 ng/µL
GJA1 crRNA	1µg/µL in Rnase-free H2O	0.4 µL	20 ng/µL
tracrRNA	1µg/µL in Rnase-free H2O	0.8 µL	40 ng/µL
eSpCas9 protein	1µg/µL in Rnase-free H2O	1.0 µL	50 ng/µL
RNAse-free injection buffer	0.1 mM EDTA pH 8.0	16.8 µL
	10 mM Tris-HCl pH 7.5
	100 mM NaCl
Total volume		20.0 µL

Table 2: Recipe of the CRISPR mixture.

Lane # (Figure 3)	Donor oligo conc.	Status	Not amplified	Wild type	Het/mosaic	Hom	Note
1	50 ng/µl	Morula			Het/mosaic		Deletion
2	50 ng/µl	Morula			Het/mosaic
3	50 ng/µl	Morula			Het/mosaic
4	50 ng/µl	Morula				Hom
5	50 ng/µl	Morula				Hom
6	50 ng/µl	Morula			Het/mosaic		Deletion
7	50 ng/µl	Morula	N/A
8	50 ng/µl	Morula			Het/mosaic
9	50 ng/µl	Morula			Het/mosaic
10	50 ng/µl	Morula		Wild type
11	50 ng/µl	Morula			Het/mosaic
12	50 ng/µl	Morula		Wild type
13	50 ng/µl	Morula			Het/mosaic		Insertion
14	50 ng/µl	Morula		Wild type
			–	3/13 (23.1%)	8/13 (61.5%)	2/13 (15.4%)
15	25 ng/µl	Morula	N/A
16	25 ng/µl	Morula	N/A
17	25 ng/µl	Morula	N/A
18	25 ng/µl	Morula	N/A
19	25 ng/µl	Morula		Wild type
20	25 ng/µl	Morula	N/A
21	25 ng/µl	Morula		Wild type
22	25 ng/µl	Morula		Wild type
23	25 ng/µl	Blastocyst	N/A
24	25 ng/µl	Blastocyst		Wild type
25	25 ng/µl	Blastocyst			Het/mosaic
26	25 ng/µl	Blastocyst		Wild type
27	25 ng/µl	Blastocyst	N/A
28	25 ng/µl	Blastocyst		Wild type
29	25 ng/µl	Blastocyst		Wild type
30	25 ng/µl	Blastocyst			Het/mosaic		Insertion/deletion
31	25 ng/µl	Blastocyst		Wild type
32	25 ng/µl	Blastocyst			Het/mosaic		Insertion
33	25 ng/µl	Blastocyst		Wild type
			–	9/12 (75.0%)	3/12 (25.0%)	0 (0.0%)

Table 3: Genotyping results from Morula and Blastocyst.

Animal #	Donor oligo conc.	Status	Not amplified	Wild type	Het/mosaic	Hom	Note
d1	50 ng/µl	Dead pup		Wild type
d2	50 ng/µl	Dead pup			Het/mosaic
d3	50 ng/µl	Dead pup			Het/mosaic		Deletion
d4	50 ng/µl	Dead pup		Wild type
d5	50 ng/µl	Dead pup			Het/mosaic
d6	50 ng/µl	Dead pup		Wild type
d7	50 ng/µl	Dead pup			Het/mosaic
d8	50 ng/µl	Dead pup			Het/mosaic
34	50 ng/µl	Live pup			Het/mosaic
35	50 ng/µl	Live pup		Wild type
36	50 ng/µl	Live pup		Wild type
37	50 ng/µl	Live pup			Het/mosaic		used as independent line
38	50 ng/µl	Live pup		Wild type
39	50 ng/µl	Live pup		Wild type
40	50 ng/µl	Live pup		Wild type
41	50 ng/µl	Live pup			Het/mosaic
42	50 ng/µl	Live pup		Wild type
43	50 ng/µl	Live pup		Wild type
44	50 ng/µl	Live pup		Wild type
45	50 ng/µl	Live pup		Wild type
d1	50 ng/µl	Dead pup				Hom
d2	50 ng/µl	Dead pup		Wild type
d3	50 ng/µl	Dead pup			Het/mosaic
46	50 ng/µl	Live pup			Het/mosaic
47	50 ng/µl	Live pup			Het/mosaic		Insertion
48	50 ng/µl	Live pup			Het/mosaic		Insertion
49	50 ng/µl	Live pup			Het/mosaic
50	50 ng/µl	Live pup		Wild type
51	50 ng/µl	Live pup			Het/mosaic		used as independent line
52	50 ng/µl	Live pup		Wild type
53	50 ng/µl	Live pup			Het/mosaic		Insertion/deletion
54	50 ng/µl	Live pup		Wild type
				16/32 (50.0%)	15/32 (46.9%)	1/32 (3.1%)

Table 4: Genotyping results from founder animals.

Video 1: CRISPR injection to the embryos. The CRISPR mixture is injected into 1-cell stage embryos described in step 2.4. Please click here to download this Video.

Video 2: Blastocysts pick up. The representative video showing blastocysts pick up in step 2.6. Please click here to download this Video.

Discussion

A gene-modified mouse model is a common approach for understanding gene function. However, since the GJA1-20k internally translated isoform is translated from the same Gja1 mRNA as full-length Cx43, a creative strategy was devised to retain full-length Cx43 expression yet suppress GJA1-20k expression. The approach is based on a mutation of the internal start codon of GJA1-20k. With a single point mutation, M213 was switched to L on Gja1 mRNA, which succeeded in suppressing GJA1-20k expression but retained full-length Cx43 expression⁴.

However, the point-mutation of a single internal translation start site presented another difficulty: determining the mouse genotypes. A single residue substitution is too small to make specific primers. Additionally, the WT and mutation alleles have identical base pairs. One more step was added to the genotyping protocol to solve these problems. Since the start codon of GJA1-20k (targeted mutation site) is recognized by the restriction enzyme, NlaIII, the site not to be recognized by NlaIII was mutated (Figure 1; CATG to CTTA). In addition, the primers to generate a PCR product were designed that does not have any other NlaIII appropriate site. A proper genotyping protocol was established for the point-mutated mouse line by adding the enzyme digestion step with the restriction enzyme. Generally, enzyme digestion is done by short-time incubation (0.5-2.0 h at 37 °C); overnight incubation (~16 h) is recommended to completely digest the whole PCR product. The introduction of mutation in the target sequence results in the misrecognition between WT and mutant alleles. Therefore, overnight incubation for sufficient digestion in this protocol is recommended unless the restriction enzyme of choice has the star activity²¹.

The current project aimed to disrupt the second translation initiation site by altering ATG (Methionine) to TTA (Leucine). The NlaIII is a convenient restriction enzyme for rapid screening as it recognizes CATG in the targeting site. Cytosine (C) is often associated with the KOZAK sequence immediately before the translation initiation site. The NcoI (C/CATGG) and NdeI (CA/TATG) are other versatile restriction enzymes for screening in this region if alternative codons need to be utilized for the creation/deletion of those restriction sites. If the introduction of silent mutations is not permissive for the restriction site-based screening, sequencing of individual samples is needed^22,23. In this project, targeted recombination was quite efficient.

This M213L mutant mouse line enables us to analyze the function of GJA1-20k. Recently, using this mouse line, it was identified that GJA1-20k is essential to the trafficking and formation of full-length Cx43 gap junctions in the heart⁴. Interestingly, the homozygous M213L mutation (Gja1^M213L/M213L), because it can not generate GJA1-20k, always results in sudden death 2-4 weeks after the birth because the respective hearts do not form gap junctions at the cardiac intercalated disc. In contrast, a heterozygous M213L mutation reveals close to normal cardiac function, and the mice have a similar life span to those of WT animals⁴. Previous in vitro study has also shown that M213L mutation disrupts proper gap junction formation¹. Moreover, this gap junction disruption is rescued by exogenous GJA1-20k transfection. Based on normal gap junction formation in heterozygous M213L mutant mouse heart under basal condition⁴ and the in vitro study¹, it can be concluded that full-length Cx43 with M213L mutation maintains proper function as a gap junction protein.

In addition to the essential role of GJA1-20k in Cx43 trafficking, GJA1-20k enriches mitochondrial outer membrane, inducing protective mitochondrial fission and biogenesis^3,24,25. In terms of mitochondrial function, mice with heterozygous mutation of GJA1-20k (Gja1^M213L/WT) have extreme sensitivity to ischemia/reperfusion injury. When exposed to ischemia/reperfusion, adult Gja1^M213L/WT mice have a much larger infarct size and more disrupted mitochondria than their WT counterparts²⁵.

In summary, the new mouse line containing the M213L mutation is helpful for the analysis of internally translated GJA1-20k. It is to be noted that both full-length Cx43 and the smaller GJA1-20k are expressed in all mammalian organ systems. Future studies are expected to explore the role of extracardiac GJA1-20k as well.

Materials

0.2 mL 8-Strip PCR tube	Thomas Scientific	1148A28
0.5 M EDTA pH 8.0	Invitrogen	15575-038	For pronuclear micorinjection buffer
10x CutSmart buffer	New England Biolabs	B7204S	Digestion buffer for NlaIII
1 M Tris-HCl pH 7.5	Invitrogen	15567-027	For pronuclear micorinjection buffer
50x TAE Buffer	Invitrogen	24710-030
96-well Thermal cycler	Applied Biosystems	Model # 9902
Agarose	Sigma-Aldrich	A9539
C57BL/6J	The Jackson Laboratory	664	mouse strain
Chemidoc MP imaging system	Bio-rad		To take gel image by 302 nm UV light
eSpCas9 protein	MilliporeSigma	ESPCAS9PRO-50UG
Gel Lading Dye Purple (6x)	New England Biolabs	B7024S	Loading buffer for electrophoresis
Gloves
hCG (human chorionic gonadotropin)	MilliporeSigma	23-073-425G
Hyaluronidase	MilliporeSigma	H3884-100MG
Image Lab software	Bio-rad		To analyze gel image
Inverted stereo microscope whit plasDIC	Zeiss	Axiovert A1
KAPA Mouse Genotyping Kits	Roche Diagnostics	KK7352	For tissue or cell lysis, DNA extraction, and PCR master mix
KSOM	LifeGlobal	ZEKS-050	embryo culture medium
Lab coart
M2 medium	MilliporeSigma	MR015D
NlaIII	New England Biolabs	R0125S	To digest PCR product
Nuclease-Free Water	Ambion	AM9937	To dilute reagents
Paraffin oil	Nacalai USA	2613785
Plugged 20 μL fine pipette tip	FisherScientific	02-707-171	To pick up blastocytes
PMSG (pregnant mare’s serum gonadotropin)	ProSpec-Tany	HOR-272
Sodium Chloride	Invitrogen	AM9760G	For pronuclear micorinjection buffer
SYBR safe DNA Gel Stain	Invitrogen	S33102	To detect bands in gel
tracrRNA	Dharmacon	U-002000-120	Guid RNA