A Murine Myh6MerCreMer Knock-In Allele Specifically Mediates Temporal Genetic Deletion in Cardiomyocytes after Tamoxifen Induction

A mouse model that mediates temporal, specific, and efficient myocardial deletion with Cre-LoxP technology will be a valuable tool to determine the function of genes during heart formation. Mhy6 encodes a cardiac muscle specific protein: alpha-myosin heavy chain. Here, we generated a new Myh6-MerCreMer (Myh6MerCreMer/+) inducible Cre knock-in mouse by inserting a MerCreMer cassette into the Myh6 start codon. By crossing knock-in mice with Rosa26 reporter lines, we found the Myh6MerCreMer/+ mice mediate complete Cre-LoxP recombination in cardiomyocytes after tamoxifen induction. X-gal staining and immunohistochemistry analysis revealed that Myh6-driven Cre recombinase was specifically activated in cardiomyocytes at embryonic and adult stages. Furthermore, echocardiography showed that Myh6MerCreMer/+ mice maintained normal cardiac structure and function before and after tamoxifen administration. These results suggest that the new Myh6MerCreMer/+ mouse can serve as a robust tool to dissect the roles of genes in heart development and function. Additionally, myocardial progeny during heart development and after cardiac injury can be traced using this mouse line.

Myh6 (α-MHC, MYHC and MYHCA) encodes the cardiac muscle specific protein alphamyosin heavy chain and is critical for heart development [22][23][24]. MYH6 mutations in humans cause atrial septal defect as well as dilated and hypertrophic cardiomyopathy [22][23][24]. The alpha-myosin heavy chain is dynamically expressed in cardiomyocytes during heart formation [25]. In this study, we created Myh6 MerCreMer/+ knock-in mice by inserting the MerCreMer cassette into the Myh6 start codon. Myh6-driven Cre recombinase was specifically activated in cardiomyocytes after tamoxifen induction at embryonic and adult stages. Thus, the Myh6 MerCreMer/+ knock-in mouse model may be a useful instrument in the temporal genetic deletion of genes of interest in cardiomyocytes in addition to tracing myocardial lineage during development and after cardiac injury.

Materials and Methods Animals
Myh6 MerCreMer/+ knock-in mice were generated by gene targeting. A MerCreMer-FRT-Neo-FRT cassette was inserted 6 bp upstream of the start codon of Myh6 (with disruption of endogenous ATG). The targeting construct contains a MerCreMer-FRT-Neo-FRT cassette flanked by 5' and 3' homologous arms (Fig 1). A linearized construct was transfected into mouse embryonic stem (ES) cells. Positive ES cells were identified by long-range PCR (Roche) with a primer external to the homologous arms and a primer located in the MerCreMer-FRT-Neo-FRT cassette. PCR fragments were verified by DNA sequencing.
Chimeric mice derived from the positive ES cells were crossed with Black Swiss wild type mice to obtain Myh6 MerCreMer-Neo/+ mice. The Myh6 MerCreMer/+ allele was obtained by crossing Myh6 MerCreMer-Neo/+ mice with FLPe mice [26]. R26R lacZ/+ and R26R tdTomato/+ mice were obtained from the Jackson Laboratory [27,28]. Myh6 MerCreMer/+ mice were genotyped by PCR of DNA isolated from mouse tails using the following primers: GCAGGCACTTTACATAGAG TCCTG (Forward, 5'!3'); GTTCAGCATCCAACAAGGCACTGA (Reverse, 5'!3'). Mice were euthanized through cervical dislocation for collecting embryonic and postnatal tissues. Animal husbandry procedures were approved by the Institutional Animal Care and Use Committee at Icahn School of Medicine at Mount Sinai (LA09-00494) and are in compliance with NIH guidelines (PHS Animal Welfare Assurance A3111-01).

X-Gal and Trichrome Staining
Whole mouse embryos and hearts were harvested at the indicated time points. β-galactosidase activity was assessed by X-gal staining as described previously [30]. Mouse embryos or hearts were dissected in PBS and fixed in 4% paraformaldehyde/PBS for 30 min at 4°C. The fixed samples were then washed twice with PBS, followed by staining in X-gal solution overnight at room temperature [31]. For cryosections, cardiac samples were embedded in Optimal Cutting Temperature (OCT) compound (Tissue-Tek) on dry ice and were cut to 10 μm in thickness. Frozen sections were stained in X-gal solution at 37°C overnight. After two washes in PBS, cardiac sections were mounted in permount medium and the images were then captured under a Leica microscope. Cardiac fibrosis was examined by Masson's trichrome Stain Kit (Sigma-Aldrich) using the manufacturer's protocol.

Immunofluorescence
Mouse hearts were fixed in 4% paraformaldehyde/PBS on ice for 30 min. They were then embedded in OCT using standard procedure. Cryosections were cut to 6 μm in thickness for immunofluorescence. Tissue sections were washed with PBS and blocked with 10% goat normal serum for 30 min at room temperature. The tissues were subsequently incubated with either primary antibody anti-mouse cardiac Troponin T (1:200, Thermo scientific), antimouse αSMA (1:100, Sigma), or anti-mouse PECAM (CD31) (1:100, BD Biosciences), for 1 hour at room temperature. Slides were washed three times in PBS, followed by incubation with Alexa Flour 488 conjugated secondary antibodies (1:500; Invitrogen) for 45 min at room temperature. Sections were counterstained with diamidino-2-phenylindole (DAPI) and photographed under a fluorescence microscope.

Echocardiography Analysis
Echocardiography (Echo) was performed with Vevo2100 system (FujiFilm VisualSonics Inc.) with methods described previously [32]. Briefly, left ventricular cardiac function and structure were assessed from short axis B-mode and M-mode images at the level of the papillary muscles. M-mode recordings were used to measure ventricular wall thickness and chamber dimensions. VisualSonics Vevo 2100 V1.5.0 software (Visualsonics; Toronto, Canada) was used to measure echocardiographic parameters, including diastolic and systolic left ventricular internal diameter (LVID, short axis B-mode), anterior wall thickness (LVAW, short axis M-mode), posterior wall thickness (LVPW, short axis M-mode). Using left ventricular systolic and diastolic chamber dimensions, short-axis ejection fraction (LVEF) and fractional shortening (LVFS) were calculated within the software using standard formulas [32]. Heart rate (HR) and body temperature were monitored throughout image acquisition. All echocardiographic data acquisition and analysis were performed by two independent examiners blinded to the experimental groups. All quantitative data are expressed as mean ± SDEV. Differences between two groups were analyzed by Student's t test using GraphPad Prism 6 software. P<0.05 was considered statistically significant.

Generation of Myh6 MerCreMer/+ Knock-In Mouse Line
To achieve temporal inactivation of genes of interest in cardiomyocytes, we generated the Myh6 MerCreMer/+ knock-in mouse line by targeting the Myh6 locus. A targeting vector containing the MerCreMer-FRT-Neo-FRT cassette flanked by homologous arms was electroporated into mouse ES cells. After homologous recombination, MerCreMer-FRT-Neo-FRT was inserted into the start codon of Myh6 ( Fig 1A). The recombinant bands 5.9-kb (5' end) and 4.3-kb (3' end) were amplified by long-range PCR with primers external to the arms and primers in the MerCreMer-FRT-Neo-FRT cassette ( Fig 1B). Myh6 MerCreMer-Neo/+ mice derived from the positive ES cells were crossed with Flippase deleter mice [26] to obtain Myh6 MerCreMer/+ mice ( Fig  1A). Myh6 MerCreMer/+ is a knock-in/knock-out allele for Myh6 and those mice developed normally without any evident defects in appearance or behavior in Black Swiss background.

Cardiac Cre Recombination Is Achieved in Myh6 MerCreMer/+ Mice After Tamoxifen Induction
To determine the efficiency and the specificity of Cre recombination in Myh6 MerCreMer/+ mice, we crossed Myh6 MerCreMer/+ mice with the R26R lacZ/+ reporter line to generate Myh6 MerCreMer/+ ; R26R lacZ/+ double heterozygous animals. β-galactosidase is expressed when Cre translocates from the cytoplasm to the nucleus after tamoxifen induction. To examine Myh6-mediated inducible Cre activity during gestation, tamoxifen was administered to pregnant dams with dosage of 0.05 mg/g body weight/day at embryonic days (E) 11.5, E14.5, and E16.5 by intraperitoneal injection for two consecutive days [29]. Whole embryos or hearts were harvested at E13.5, E16.5, and neonate (P0) to evaluate β-galactosidase activity, respectively. We observed robust X-gal staining in the hearts at E13.5, E16.5 and P0, but no staining was detected in other regions or organs (Fig 2A-2C). This revealed that tamoxifen effectively induces recombination in the hearts at embryonic stages. Furthermore, with three consecutive days of tamoxifen injection at 0.1 mg/g body weight/day, we were also able to detect robust X-gal staining on the postnatal hearts at P30 and P60 (Fig 2D-2E). Of note, X-gal staining was not detected in the aorta or pulmonary artery on Myh6 MerCreMer/+ ;R26R lacZ/+ hearts at embryonic and adult stages (arrowheads in Fig 2A2,2B1,2C1,2D1 and 2E1). Further examination of cardiac sections revealed that β-galactosidase activity encompasses atrial and ventricular walls, as well as septal regions (notched arrows in Fig 2C2,2C3,2D2,2D3,2E2 and 2E3), but not valves (unnotched arrows in Fig 2C4,2D4 and 2E4). Moreover, very few X-gal positive cells were detected in Myh6 MerCreMer/+ ;R26R lacZ/+ hearts at the adult stage when tamoxifen was not injected (Fig 3), suggesting that Myh6 MerCreMer/+ mice have very minimal Cre leakiness. Collectively, these observations indicated that the Myh6 MerCreMer/+ allele mediates highly efficient Cre-LoxP recombination in embryonic and postnatal hearts after tamoxifen induction.

Myh6 MerCreMer/+ Introduces Specific Recombination in Cardiomyocytes
Next, we attempted to determine whether recombination mediated by Myh6 MerCreMer/+ is limited to cardiomyocytes. R26R tdTomato/+ reporter mice were crossed with Myh6 MerCreMer/+ mice to obtain Myh6 MerCreMer/+ ;R26R tdTomato/+ double heterozygous animals. Consistent with the Xgal staining results, high levels of tdTomato expression were observed on Myh6 MerCreMer/+ ; R26R tdTomato/+ hearts at E13.5, E16.5, P0, and P60 following tamoxifen induction (Fig 4A). To examine whether the tdTomato-positive cells were cardiomyocytes, cardiac sections were made for immunofluorescence. We observed that at E13.5 and E16.5, even though robust tdTomato expression was detected in both atrial and ventricular walls, atrial fluorescent signals appear much stronger than that of ventricles. The trabeculated myocardium also displayed stronger fluorescence than the compact myocardium in the ventricles (arrows and arrowheads  in Fig 4B2 and 4C2). This may reflect differential Myh6 expression level in cardiac compartments at embryonic stages as previously described [33]. From birth to adulthood, tdTomato signals were evenly distributed in the ventricle (notched arrows in Fig 4D2 and 4E2). We further performed immuostaining with anti-cardiac troponin T (cTnT/Tnnt2, marker for cardiomyocytes), anti-alpha smooth muscle actin (αSMA, marker for vascular smooth muscle cells), and anti-PECAM (marker for endothelial cells) [34][35][36][37]. The results showed that tdTomato was fully co-localized with cTnT in the heart at E16.5-P60 (notched arrows in Fig  4B4,4C4,4D4 and 4E4), but not with αSMA or PECAM in the coronary smooth muscle cells (unnotched arrows in Fig 4F4 and 4G4) or endothelial/endocardial cells (unnotched arrows in Fig 4H2 and 4I2). Additionally, tdTomato was not detected in valves (arrowheads in Fig 4C4  and 4I2). This indicates that Myh6 MerCreMer/+ specifically drives Cre recombination in cardiomyocytes.

Myh6 MerCreMer/+ Hearts Display Normal Structure and Function before and after Tamoxifen Induction
Given that Myh6 MerCreMer/+ is a knock-in/knock-out allele for Myh6, it is important to know whether the cardiac function and structure were adversely affected by MerCreMer insertion. We performed transthoracic echocardiography on Myh6 MerCreMer/+ and wild type littermate mice at P60-90 (n = 10 for each group). Left ventricular short-axis measurements showed no change in cardiac structure and function between Myh6 MerCreMer/+ and wild type littermate mice (Fig 5A and S1 and S2 Videos). There was no significant difference in cardiac chamber dimensions, wall thicknesses, fractional shortening, or ejection fraction in Myh6 MerCreMer/+ mice when compared to their wild type littermates (Fig 5B, Table 1), suggesting MerCreMer insertion into the Myh6 start codon had no effect on cardiac development and function. Furthermore, to determine whether tamoxifen administration had any impact on cardiac function, we performed echocardiography on Myh6 MerCreMer/+ and wild type littermates five weeks after the final injection (0.1 mg/g body weight/day for 3 days). No significant difference was found between Myh6 MerCreMer/+ and littermate controls (n = 6 for each group, Fig 5, Table 1 and S3 and S4 Videos). Subsequent TUNEL and trichrome staining demonstrated that tamoxifen does not lead to myocardial apoptosis or fibrosis on Myh6 MerCreMer/+ hearts one and five weeks after injection (Fig 6). In addition, we attempted to determine the minimum effective tamoxifen dosage to minimize any potential cardiac toxicity. With 0.05 mg/g body weight/day for three days, the adult Myh6 MerCreMer/+ hearts still exhibited sufficient recombination one month after tamoxifen injection (Fig 7).

Discussion
In this study, we described generation and characterization of a new Myh6 MerCreMer/+ knock-in mouse model. Myh6 MerCreMer/+ mice develop normally without cardiac functional defects. Short-term tamoxifen treatment resulted in efficient Cre recombination in cardiomyocytes. This new Myh6 MerCreMer/+ animal is a useful tool for deletion of genes of interest in myocardium with Cre-LoxP technology at desirable stages. A few mouse lines were created previously for inducible genetic deletion in myocardium with tetracycline and MerCreMer systems [11,12]. The tetracycline-inducible system requires two transgenes: a reverse tetracycline-controlled transactivator (rtTA) directed by a rat cTnT promoter and a Cre recombinase driven by a tetracycline responsive promoter (TetO), thereby making breeding scenarios complicated [11]. Another limitation of the tetracycline-inducible system is potential leakiness [38]. The α-MHC-MerCreMer transgenic mouse line was generated and had been used widely for gene inactivation in the myocardium [12,[39][40][41][42]. However, a few studies showed that the α-MHC-MerCreMer mouse line displayed myocardial fibrosis and cardiac dysfunction due to Cre-induced DNA damage and myocardial apoptosis after tamoxifen induction [15][16][17][18][19][20][21]. In this new Myh6 MerCreMer/+ mouse, Cre recombination is strongly activated within cardiomyocytes following tamoxifen induction. This animal exhibited relatively good tolerance to tamoxifen (no myocardial fibrosis or apoptosis) and displayed normal cardiac structure and function after appropriate induction. Moreover, a low dosage of tamoxifen (0.05 mg/g body weight for 3 days) also introduces robust and specific recombination in the cardiomyocytes at adult stage.
Myh6 MerCreMer/+ is a heterozygous null for Myh6 (Myh6 +/− ). Myh6 +/− animals were shown to have cardiac functional defects with sarcomeric structural alterations and fibrosis [43]. However, by performing trichrome staining and echocardiography, we did not detect effects on Myh6 MerCreMer/+ hearts (2-3 months old, n = 10, Fig 5, Table 1, and S1 and S2 Videos). The discrepancy could be due to the difference in the gene targeting strategy and/or a genetic divergence between these animals: in Myh6 MerCreMer/+ mice, the MerCreMer-FRT-Neo-FRT cassette Tamoxifen has little effect on Myh6 MerCreMer/+ hearts. Apoptosis and fibrosis assays on Myh6 MerCreMer/+ mouse hearts in one week (A-F) and five weeks (G-L) after tamoxifen injection (0.01 mg/g body weight for 3 days). A few apoptotic cells were observed on both R26R tdTomato/+ and Myh6 MerCreMer/+ ; R26R tdTomato/+ hearts after one week (arrows in A,B, P60!P67), and very few detected after five weeks (G,H, P60!P95). Tamoxifen mediates robust recombination on Myh6 MerCreMer/+ ;R26R tdTomato/+ hearts (D,J), but not on R26R tdTomato/+ control hearts (C,I). Myocardial fibrosis was not detected on R26R tdTomato/+ or Myh6 MerCreMer/+ ;R26R tdTomato/+ hearts by trichrome staining after one (E,F) and five weeks (K,L). Image in the upright corner of E is from an unrelated study, showing positive trichrome staining and cardiac fibrosis (arrows). This staining was performed in parallel with sample in E.
doi:10.1371/journal.pone.0133472.g006 was inserted into the ATG locus and the Neomycin sequence was removed by Flippase deleter mice. In the Myh6 +/− mice examined by Jones et al [43], the pgk-Neo-polyA cassette was targeted into the Myh6 locus with a deletion of an approximately 2-kb fragment of the Myh6 gene. The deleted sequence includes the first three exons, the 5' untranslated region, and the initiating methionine codon [43]. It is uncertain whether the existing Neomycin cassette in this strong myocardial locus has any negative effects on cardiac function. Moreover, it is important to note that the physiologic and pathologic phenotypes in Myh6 +/− mice are not completely penetrant [43]. Myh6 MerCreMer/+ animals in this study are in hybrid background (Black Swiss), and genetic and epigenetic variations could potentially be important factors for Myh6 +/− heart function [43]. In the future, it will be of interest to determine whether the inbred background of Myh6 MerCreMer/+ mice has any impact on their cardiac performance.
As mentioned before, tamoxifen injection into α-MHC-MerCreMer transgenic line could lead to severe toxicity to the heart [15][16][17][18][19][20][21]. Mice with three doses of tamoxifen at 0.03-0.09 mg/g body weight/day displayed cardiac fibrosis and dysfunction, with 10-50% mortality within one week [20,21]. In this study, we found that with three doses of tamoxifen at 0.1 mg/g body weight/day, the Myh6 MerCreMer/+ mice appeared normal in cardiac function and structure and no lethality was observed. No myocardial fibrosis or apoptosis was found in Myh6 MerCreMer/+ mice after one and five weeks of administration (Figs 4 and 5). This may be explained by the genetic difference between α-MHC-MerCreMer and Myh6 MerCreMer/+ mice. α-MHC-MerCreMer is a transgenic line and each myocardial cell may have multiple copies of MerCreMer (note each MerCreMer cassette has its own α-MHC promoter) [44]. In contrast, Myh6 MerCreMer/+ knock-in animals only have one copy of MerCreMer in their genome. Therefore, MerCreMer expression in α-MHC-MerCreMer myocardial cells might be much higher than that in Myh6 MerCreMer/+ myocardial cells. Under certain dosage of tamoxifen induction (e.g., 0.1 mg/g body weight/day for three days), the high level MerCreMer expression may lead to excessive amount of Cre translocation to nuclei which in turn may induce DNA damage and cell death in the cardiomyocytes. Myh6 MerCreMer/+ cardiomyocytes have lower MerCreMer expression and do not have a large amount of Cre translocation under this dosage.
The major application of this new Myh6 MerCreMer/+ mouse model will be the temporal disruption of genes of interest in cardiomyocytes in vivo. Given that almost all the myocardial cells robustly express Cre after tamoxifen induction, this inducible Cre mouse line can also be applied to determine myocardial lineage during development and after cardiac injury.
Supporting Information S1 Video. Echocardiography analysis of wild type mouse without tamoxifen treatment.