Homology-directed repair of an MYBPC3 gene mutation in a rat model of hypertrophic cardiomyopathy

Variants in myosin-binding protein C3 (MYBPC3) gene are a main cause of hypertrophic cardiomyopathy (HCM), accounting for 30% to 40% of the total number of HCM mutations. Gene editing represents a potential permanent cure for HCM. The aim of this study was to investigate whether genome editing of MYBPC3 using the CRISPR/Cas9 system in vivo could rescue the phenotype of rats with HCM. We generated a rat model of HCM (“1098hom”) that carried an Mybpc3 premature termination codon mutation (p.W1098x) discovered in a human HCM pedigree. On postnatal day 3, the CRISPR/Cas9 system was introduced into rat pups by a single dose of AAV9 particles to correct the variant using homology-directed repair (HDR). Analysis was performed 6 months after AAV9 injection. The 1098hom rats didn’t express MYBPC3 protein and developed an HCM phenotype with increased ventricular wall thickness and diminished cardiac function. Importantly, CRISPR HDR genome editing corrected 3.56% of total mutations, restored MYBPC3 protein expression by 2.12%, and normalized the HCM phenotype of 1098hom rats. Our work demonstrates that the HDR strategy is a promising approach for treating HCM associated with MYBPC3 mutation, and that CRISPR technology has great potential for treating hereditary heart diseases.


INTRODUCTION
Hypertrophic cardiomyopathy (HCM) is a hereditary disease with a global prevalence of approximately 1 in 500 [1]. It is characterized by left ventricular hypertrophy, myocardial hypercontractility, reduced compliance, myofibrillar disarray, and fibrosis. Symptoms in patients with HCM range from mild chest pain, syncope, and dyspnea to severe sudden cardiac death [2]. HCM is a major cause of sudden cardiac death in young adults, especially young athletes [3]. Current therapy for HCM includes pharmacological intervention represented by β-blockers or non-dihydropyridine calciumchannel blockers, surgical septal myectomy, alcohol septal ablation, and radiofrequency ablation. These treatments provide some relief, but not a permanent cure.
HCM is predominantly caused by mutations in sarcomere protein encoding genes [4]. Up to 1500 mutations in 11 genes have been reported as potential causes of HCM. The gene coding cardiac myosin-binding protein C, myosin-binding protein C3 (MYBPC3), is a major pathogenic gene, accounting for 30% to 40% of all HCM mutations [5]. The MYBPC3 gene encodes a 144-kDa protein located in the C-zone of the A-band of the sarcomere duplex [6]. MYBPC3 protein promotes the structural integrity of the sarcomere by linking myosin, titin and actin. It also regulates cardiac contractility in response to adrenergic stimulation [7]. Variants in MYBPC3 mainly lead to a premature termination codon (PTC), resulting in a defective mRNA unable to express full-length MYBPC3 protein with normal titin and/or the major myosinbinding sites. Patients carrying these variants suffered impaired sarcomere function and developed an HCM phenotype. Studies have shown that correction of MYBPC3 mutations can restore sarcomere function in induced pluripotent stem cell-derived cardiomyocytes. However, in vivo studies of correcting MYBPC3 mutations to treat HCM have not been reported [1].
The CRISPR (clustered regularly interspaced short palindromic repeats)/Cas9 system has been proven to be a powerful tool for genetic engineering. To date, the CRISPR/Cas9 system have been implemented as gene therapy for many disease models, such as hearing loss, liver disease and muscle dystrophy [8][9][10]. However, its role in correcting MYBPC3 mutations in vivo has not been elucidated. In this study, we generated a novel rat model of HCM (referred to as "1098hom") that carries the p.W1098x PTC mutation discovered in an HCM pedigree. By employing a CRISPR/Cas9 based homology-directed repair (HDR) strategy in diseased pups, we partially restored MYBPC3 protein expression and alleviated cardiac function. Our work demonstrates that the HDR genome editing strategy is a promising approach for treating HCM associated with MYBPC3 mutation, and that this CRISPR-based technique has great potential for treating hereditary heart diseases.

MATERIALS AND METHODS Human subjects
The study was approved by the Ethics Review Board of Tongji Hospital and Tongji Medical College. It complied with the principles of the Declaration of Helsinki. Written informed consent was obtained from individual subjects of the HCM pedigree in this study.

Animal experiment
The Mybpc3-p.W1098x SD rats were generated by the Beijing Laboratory Animal Research Center of Chinese Academy of Medical Sciences. Male homozygous p.W1098x mutant rat puppies and male wild type rat puppies of 3-day-old were used in this study. The rats were treated and raised till 6month-old and sacrificed. There were 8 wild type rats allocated to the WT group, 6 homozygous p.W1098x littermates to the HDR group and 13 homozygous p.W1098x littermates to the control group randomly. According to the sample size calculations [11], for a power of 0.80, an alpha value of 0.05, the estimating sample size was 3.98 per group. Thus the present sample size was considered sufficient. The investigators were not blind to the groups. And the rats of all groups were equally raised. All animal experiments complied with the "Guide for the Care and Use of Laboratory Animals" published by the United States National Institutes of Health (NIH Publication No. 85-23, revised 1996). This study was approved by the Institutional Animal Research Committee of Tongji Medical College. Experimental rats were housed at the animal care center of Tongji Medical College at 25°C with 12 h/12 h light/dark cycles. Adequate water and food supply were provided throughout the duration of the study. Genotypes of the rats were confirmed by Sanger sequencing with validation primers (Supplementary Methods).

Cell culture
The rat fibroblast cells 208 F were purchased from Obio Technology (Shanghai) Corp.,Ltd. We amplified the genome sequence of the target gene via PCR and confirmed the detailed sequence information using Sanger sequencing. The 208 F cells were cultured in MEM supplemented with 10% FBS. Transfection was performed with 3 µl Lipofectamine 2000 (Life Technologies, Carlsbad, California, USA) and 2 µg of total plasmid each well of a 12-well culture plate.

gRNA design and plasmid cloning
Three guide RNAs (gRNA) were designed using the CCTop tool (https:// crispr.cos.uni-heidelberg.de/) [12,13]. These gRNAs were cloned into the px458 vector (a plasmid expressing spCas9 and gRNA backbone sequence, bought from Addgene, RRID: Addgene_48138). For in vitro study, 208 F cells were transfected with the 3 gRNA specific px458 vectors, and prepared for subsequent T7E1 assay and Sanger sequencing. For in vivo study, the gRNA-donor sequence and spCas9 sequence were cloned into two separate AAV plasmids and processed to two AAV9 viruses. Rat pups were injected with a single dosage of the mixture of gRNA-donor-AAV9 virus and spCas9-AAV9 virus. Cloning details and AAV9 production were previously described [14].

Echocardiography
Vevo 1100 Imaging System (Visual Sonics Inc., Toronto, Canada) was used for echocardiography examination. Rats were anesthetized with isoflurane (21% in oxygen). Echocardiographic parameters were obtained with a 30 MHz transducer.

Hemodynamics
After anesthetization, a pressure-volume catheter (Millar 1.4Fr, SPR 835, Millar Instruments, USA) was inserted into the left ventricle through the right carotid artery. After stabilization, the readout signals were analyzed using the PVAN software (Millar Instruments, Houston, Texas, USA).

Western blot
Rat myocardium were homogenized using Protein or IP lysate buffer (Beyotime Technology, Shanghai, China) containing 1:100 protease inhibitor and 1:100 phosphatase inhibitor. Protein concentrations were determined using the Bicinchoninic Acid Assay Kit (Boster, Wuhan, China). After denaturation, tissue lysates were resolved by SDS-PAGE, transferred to nitrocellulose membrane, and blocked with 5% BSA in TBST. After incubation with primary antibody for 12-16 h at 4°C and incubation with secondary antibody for 2 h at room temperature, the membrane was developed with ECL system (Advansta, California, USA). The antibodies were MYBPC3 (sc-137237, Santa Cruz Biotechnology, Dallas, Texas, USA) and GAPDH (AC002, ABclonal Technology, Wuhan, China).

RNA extraction and qRT-PCR
Total RNA was extracted using TRIzol (Invitrogen, Carlsbad, California, USA). Reverse-transcription was performed using MultiScribe system (ABI, Waltham, Massachusetts, USA) for complementary DNA (cDNA). 7900HT Fast Real-Time PCR System was used for real-time PCR. The primer sequences were listed in Supplementary Methods.

Isolation of genomic DNA
Genomic DNA were isolated from the cultured cells or myocardium using TIANamp Genomic DNA Kit (DP304, TIANGEN BIOTECH, Beijing, China) according to the manufacturer's protocols. Briefly, the myocardium tissue was cut into pieces and digested with Proteinase K in buffer GA and GB for 10 min at 70°C. Genomic DNA was sedimented with ethyl alcohol and retained in a spin column. Wash the column with buffer GD and PW. Elute the column with buffer TE and centrifuge. Genomic DNA was thus obtained.

Deep sequencing
The primer details for deep sequencing were listed in Supplementary Table  S4. Genomic DNA was extracted and was PCR amplified by locus-specific primers targeting on-or off-target sites using AlleleID version 6 (PREMIER Biosoft, San Francisco, California, USA). A pair of 8 base unique barcodes were added to 5' end of the forward and reverse primers of each sample. The PCR products were purified using Universal DNA Purification Kit (DP214, TIANGEN BIOTECH, Beijing, China). All barcoded amplicons in an equal molar ratio were pooled together and processed on the Illumina MiSeq platform by 2 × 250 paired-end sequencing.

Statistics
Each dot in the figures represented a biological repeat and was technically repeated 3 times. SPSS version 20 (IBM SPSS, Beijing, China) and GraphPad Prism version 5 (GraphPad, Boston, Massachusetts, USA) were used for analyzing. Data were expressed as mean ± standard error of the mean. The comparison of unpaired data following normal distribution and homogeneity of variance between two groups was analyzed by unpaired Student's t test. Comparisons among multiple groups were analyzed by one-way analysis of variance with Newman-Keul's post hoc test. Biostatistics significance was accepted at p < 0.05.

RESULTS
Generation of HCM rat model and designation of gene editing strategy The p.W1098x variant in MYBPC3 was discovered in a Chinese Han HCM pedigree (Fig. 1a). This three-generation pedigree consists of two family members who presented with HCM. The proband (II:2) is a 33-year-old man diagnosed with HCM eighteen years ago. He suffered from chest pain and syncope. Echocardiography revealed a ventricular septum thickness of 16 mm. High-throughput sequencing identified a c.3293G>A mutation causing the p.W1098x variant in the proband. Subsequent investigation revealed that the proband's mother (I:2) is also an HCM patient carrying the same variant. The p.W1098x variant is located within the 30th exon and converts the tryptophan (Trp, W) 1098 codon to a premature termination codon (Fig. 1b). While there have been a few sporadic reports of HCM cases with this variant [7], there is no functional evidence for this variation in ClinVar database.
To explore the efficacy of genetic engineering for the treatment of a MYBPC3 missense variant associated HCM, we generated the p.W1098x rat model (Fig. 1c). Different from human, heterozygous rats with a p.W1098x allele and a wild-type allele did not exhibit an HCM phenotype (Supplementary Table S1). However, homozygous rats carrying both mutant p.W1098x alleles developed cardiac hypertrophy and reduced cardiac function, reproducing the phenotype of the HCM patients ( Fig. 1d-g). Therefore, homozygous p.W1098x mutant rats (1098hom) were used as the HCM model in this study.
An HDR genome editing strategy was designed to correct the stop codon (TGA) back to the original TGG (Trp, W) codon. To achieve efficient genome editing, we designed 3 gRNAs to target exon 30 (Fig. 2a). These gRNAs were cloned into plasmid with a SpCas9 backbone and transfected into rat fibroblast 208 F cells. Genomic DNA was extracted from 208 F cells, and target PCR products were generated for the T7E1 assay and Sanger sequencing. T7E1 endonuclease detected mismatched doublestranded DNA and cut it into fragments. As shown in Fig. 2b and Supplementary Fig. S1, gRNA1 showed the highest editing activity and was selected for in vivo gene editing studies. As shown in Fig. 2c dual AAV9 strategy were employed, with one virus carrying the gRNA1 plus donor sequence, and the other virus carrying the SpCas9 sequence. The unedited BbsI clone site was used as a negative control sequence. Five synonymous mutations were introduced into the donor sequence at the gRNA target site to avoid repeated editing (Fig. 2d). Detailed information of gRNA targeting and donor sequence were provided in Supplementary Methods.

Alleviation of cardiac hypertrophy and cardiac function by HDR treatment
Rat pups were divided into 3 groups: wild type (WT), 1098hom rats receiving control treatment (1098hom + Con), and 1098hom rats receiving HDR treatment (1098hom + HDR). AAV9 viruses were injected retro-orbitally at postnatal day 3. The rats were equally fed and raised for 6 months without any other interventions. Six The G to A mutation causing the p.W1098x variant leading to a Tryptophan codon converting to a premature termination codon. c Sanger sequencing validating this variant in homozygous 1098hom rats. d No detectable MYBPC3 protein in 1098hom rats. e Cardiac M-mode echocardiography tracings of a 1098hom rat and its wild-type littermate at 6-month of age. f Ejection fraction was reduced in 1098hom rats. g 1098hom rats had thickened ventricular wall. *p < 0.05. Data were expressed as mean ± standard error of the mean. months later, echocardiographic, hemodynamic and heart histochemical analysis were performed. As shown in Fig. 3, control 1098hom rats had hypertrophic hearts and thickened ventricular wall, as assessed by heart/body weight ratio and left anterior ventricular wall thickness. HDR treatment alleviated the cardiac hypertrophy of 1098hom rats.
We next assessed cardiac function by echocardiography and in vivo hemodynamics. As shown in Fig. 4, cardiac function was impaired in control 1098hom rats. Ejection fraction (EF) was 61% in control 1098hom rats compared with 78% in the WT group. Fractional shortening (FS), max dP/dt and min dP/dt were also decreased in control 1098hom rats. HDR treatment partially restored the impaired cardiac function (Fig. 4a- d and Supplementary Table S2). In addition, atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP), biomarkers of heart failure that were elevated in control 1098hom rat hearts, were normalized by  b Images of whole mount hearts, c H&E staining, and d M-mode echocardiography recordings showed cardiac hypertrophy in control 1098hom rats; whereas HDR treatment alleviated cardiac hypertrophy in 1098hom rats. e The heart weight/body weight ratio (HW/BW) was increased in control 1098hom rats. HDR treatment reduced the HW/BW ratio. f Left ventricular anterior wall thickness at diastole (LVAW; d) was increased in control 1098hom rats, and was alleviated by HDR treatment. WT group: n = 8; 1098hom + Con: n = 13; 1098hom + HDR: n = 6; *p < 0.05. Data were expressed as mean ± standard error of the mean.
HDR editing (Fig. 4e, f). Sirius red staining of heart section revealed increased cardiac fibrosis in control 1098hom rats, which was attenuated by HDR treatment. There was no significant difference in cardiac hypertrophy and cardiac function between 1098hom rats with and without control AAV treatment (Supplementary  Table S3). Together, these observations revealed that HDR treatment partially improved the HCM phenotype in 1098hom rats.
HDR treatment partially corrected p.W1098x mutation and restored MYBPC3 protein expression The functional and histological studies described above revealed therapeutic effects of the HDR treatment. To verify the efficacy of genome editing by HDR strategy, we performed Sanger sequencing on rat heart tissue samples, as shown in Fig. 5a. A considerable number of alleles was edited according to the donor template, as indicated by the changes in peak pattern. Of interest to us, the p.W1098x mutation was also partially corrected by A > G conversion.
To further evaluate the precise editing efficiency, we performed deep sequencing, as shown in Fig. 5b, c. Figure 5b showed the proportion of base substitution of 5 synonymous mutations and the p.W1098x position were about 25%. However, the precise genome editing efficiency was much lower. As shown in Fig. 5c, though insertions/deletions were found in 20.6% of all reads, the proportion of precise HDR editing was 3.56%. Our data was consistent with previous studies, indicating the high efficiency of NHEJ and low efficiency of precise editing by HDR strategy [16].
In addition, we assessed the frequency of off-target edits. We predicted the potential off-target sites of gRNA1 using the CCTop tool (https://crispr.cos.uni-heidelberg.de/) and amplified the sequence of top 10 potential off-target sites. Deep sequence revealed that no significant off-target effect was detected (Supplementary Tables S4, S5, and Supplementary Fig. S2).
We also performed immunostaining of the myocardium to detect MYBPC3 protein expression. As shown in Fig. 6a, b, control 1098hom rats did not express detectable levels of MYBPC3, while in rats with HDR treatment, 6.6% of the cardiac myocytes expressed MYBPC3 protein. Immunoblotting revealed MYBPC3 protein expression in the myocardium of HDR treated 1098hom rats (Fig. 6c, d). Together, these data indicate that the HDR strategy partially corrected the p.W1098x mutation and restored MYBPC3 protein expression.

DISCUSSION
In this study, we first demonstrated that the p.W1098x variant in MYBPC3 causes HCM by establishing a 1098hom rat model. Next, we used a CRSIPR/Cas9 strategy to correct the p.W1098x variant using HDR, which partially restored MYBPC3 protein expression levels. Furthermore, HDR treatment also attenuated cardiac hypertrophy and cardiac function in 1098hom rats. Our study suggests that CRISPR/Cas9 may be a potential treatment for HCM. This is the first study in vivo to reveal the therapeutic potential of HDR in treating inherited heart disease.
The HDR pathway is an ideal way of precise gene engineering. With efficient HDR, it is theoretically possible to program the genome with any desired change. Prior studies have shown that HDR has great potential for the treatment of genetic diseases such as hereditary tyrosinemia [17], OTC deficiency [9], and inherited blindness [18]. For myopathy treatment, nanoparticle delivery of Cas9 ribonucleoprotein and donor DNA template corrected 5.4% of the dystrophin gene and improved animal strength in the mdx mouse model of Duchenne's muscular dystrophy [19]. Collectively, Fig. 4 HDR treatment improved cardiac function in 1098hom rats. HDR treatment improved (a) ejection fraction (EF), (b) fractional shortening (FS), (c) max dp/dt, and (d) min dp/dt in 1098hom rats. e, f Biomarkers ANP and BNP were up-regulated in control 1098hom rats, and were down-regulated by HDR treatment. g, h Sirius red staining revealed increased fibrosis in control 1098hom rats, whereas HDR treatment alleviated cardiac fibrosis. *p < 0.05. Data were expressed as mean ± standard error of the mean.
these studies demonstrated the potential of an HDR strategy for disease treatment.
The biggest obstacle to the application of HDR is its low efficiency in postmitotic cell types. Since most somatic cells are nondividing, the efficiency of HDR repair is naturally low. Adult human heart, which is thought to be postmitotic, has only 1% of its cardiomyocyte renewed at the age of 20-year-old and 0.3% at the age of 70-year-old, as measured by the integration tests of 14 C [20]. The neonate heart, on the other hand, retains its ability to regenerate for some time after birth. Mouse cardiomyocytes exit the cell cycle after about a week of birth, while human cardiomyocyte regeneration continues for years [21]. Retention of cardiomyocyte regeneration provides a time window for therapeutic gene editing of postnatal hearts. In this study, the CRISPR/Cas9 HDR strategy was used to restore the expression of MYBPC3 protein by correcting 3.56% of the p.W1098x variant as evaluated by deep sequence. It is interesting that such a small portion of corrected cardiomyocytes would improve cardiac function, which is consistent with previous studies. For instance, expression of 1.1 ± 1.1% and 3.2 ± 2.4% of dystrophin by cardiomyocytes improved cardiac symptoms in a mouse model of muscular dystrophy [22]. Low levels of dystrophin (3-15%) are sufficient to delay the onset of cardiomyopathy [23]. Disruption of about 2-6% of mutant alleles restores heart morphology and function in PRKAG2 transgenic mice with glycogen-storage cardiomyopathy [24]. These findings suggest that relatively low editing efficiencies are sufficient to produce therapeutic effects.
The major cause of HCM are variants in genes encoding sarcomere components, such as the MYBPC3 and MYH7 genes [25]. Individuals carrying homozygous, double or triple mutations have more severe disease progression [26]. At present, there is no effective treatment for HCM caused by truncating variants in MYBPC3. Our study has established an HCM rat model of the homozygous p.W1098x variants in Mybpc3. The heart function of these 1098hom rats was restored by CRISPR/Cas9 genome editing, indicating that CRISPR/Cas9 is a good tool for the treatment of HCM. A limitation of this study is the inability of gRNA to distinguish between the mutant and wild-type alleles. In patients who carry a pathogenic sequence that is heterozygotes, it is important to target the mutant allele while maintaining the integrity of the wild-type allele. This might be more like the real world, where disease-causing allele could not be specifically targeted by current Cas9 variants. For homozygous mutations, HDR strategy would be a better choice.
In conclusion, we applied the CRSIPR/Cas9 HDR strategy to correct a truncating Mybpc3 variant, restored MYBPC3 protein expression, and improve cardiac function of HCM rats. Our work demonstrates that HDR strategy is a promising therapy for the treatment of HCM associated with MYBPC3 mutations, and that Fig. 6 HDR treatment partially restored MYBPC3 protein expression. a Immunofluorescence of heart sections. Green, MYBPC3 protein; blue, cell nucleus. Scale bars, 50 µm. b Quantification of the percentage of MYBPC3 positive cells by immunofluorescence. Control 1098hom rats did not express detectable levels of MYBPC3, while in rats with HDR treatment, 6.6% of the cardiac myocytes expressed MYBPC3 protein. c, d Immuno blotting and semi quantification of control 1098hom rats and 1098hom rats with HDR treatment. HDR treatment restored MYBPC3 protein expression by 2.12%. *p < 0.05. Data were expressed as mean ± standard error of the mean.