Dual roles of modulatory calcineurin-interacting protein 1 in cardiac hypertrophy

Generation and Analysis of MCIP1^−/− Mice.

MCIP1 genomic clones were isolated from a murine bacterial artificial chromosome library (Research Genetics, Birmingham, AL). The MCIP1 targeting vector was designed to remove exons 5 and 6. The left arm of the targeting construct contained ≈2.5 kb of the intron between exons 4 and 5 including the splice acceptor site of exon 5. A β-galactosidase reporter gene was fused to the splice acceptor site of exon 5. An NcoI site was engineered at the beginning of the β-galactosidase reporter to assist in genotyping. The right arm of the targeting construct contained ≈4.5 kb of a region of the MCIP1 genomic locus consisting of the intron between exons 6 and 7 and most of exon 7. The targeting vector was linearized and electroporated into mouse embryonic stem cells of strain 129 origin. Properly targeted embryonic stem cell clones were identified by Southern blot analysis and injected into blastocysts isolated from C57BL6/J mice. Chimeras obtained from the blastocyst injections were bred with C57BL6 mice to obtain offspring with the targeted allele. All experiments presented were performed in the mixed C57BL6/129 background. The muscle creatine kinase (MCK)-constitutively active CnA (CnA*) mice have been described (20). For β-galactosidase staining, embryos from timed matings or adult tissues were fixed in PBS-buffered formalin and subsequently stained overnight with 5-bromo-4-chloro-3-indolyl β-D-galactoside. Northern blot analysis was performed by using total RNA isolated from the hearts of 6-week-old mice using Trizol reagent (Invitrogen).

Histology.

Hearts were fixed in PBS-buffered formalin, embedded in paraffin, and sectioned at 5 μm for histological examination. Sections were stained with Masson's trichrome and photographed under normal bright-field microscopic conditions.

Transverse Aortic Constriction (TAC).

The protocol for constricting the transverse aorta of mice is an established procedure for producing left ventricular (LV) hypertrophy. Adult mice 10–12 weeks of age were anesthetized with ketamine (87 mg/kg) and xylazine (13 mg/kg) administered i.p. After blunt dissection through the intercostal muscles, the aorta was identified and freed by additional blunt dissection. A 7.0 silk suture was placed around the great vessel and tied around a blunt needle of 27 gauge. The blunt needle then was removed rapidly, causing a tight constriction. The surgical incision was closed in two layers with an interrupted suture pattern. Seven days after surgery, mice were killed, and their hearts were removed rapidly for assessment of chamber weight and snap-frozen for RNA isolation. Noninvasive transthoracic echocardiograms were performed according to the methods described (21).

Isoproterenol Infusion.

Adult male mice 10–12 weeks of age were anesthetized by i.p. administration of Avertin. A miniosmotic pump (Alzet, Palo Alto, CA) containing isoproterenol or saline vehicle was inserted s.c. in the back of the animal. This pump delivered 28 μg/hour per 25 kg of body weight (BW) of isoproterenol in a 0.9% saline solution or saline solution alone. Mice were killed 7 days later for assessment of cardiac hypertrophy.

Calcineurin Activity Assay.

Cardiac tissue was minced finely and homogenized in lysis buffer (50 mM Tris, pH 7.4/5 mM ascorbic acid/protease inhibitors, Roche, Indianapolis) followed by further disruption by sonication. Lysates then were cleared by centrifugation. Free phosphate was removed by passing the lysate through a Micro Bio 6 desalting column (BioRad). Calcineurin activity was measured by using the synthetic phosphorylated RII peptide (Biomol, Plymouth Meeting, PA). The amount of free phosphate released was measured with molybdate dye solution (Promega). Calcineurin-specific phosphatase activity was measured as the amount inhibited by cyclosporin A. Western blots to detect calcineurin protein were performed with the above lysates by using a calcineurin A antibody (Transduction Laboratories, Lexington, KY).

Northern Blot Analysis.

Total RNA was isolated from cardiac ventricular tissue by using Trizol (Invitrogen). Four micrograms of total RNA was spotted onto dot blots, and 15 μg of total RNA was loaded in each gel lane. Gene-specific probes were hybridized by using QuikHyb (Ambion, Austin, TX). Measurements of bound probe were made by using a Storm PhosphorImager (Molecular Dynamics).

Statistical Analysis.

Statistical significance was determined by the two-tailed Student's t test.

Generation of MCIP1^−/− Mice.

The mouse MCIP1 gene contains seven exons including four potential alternative first exons (Fig. 1A). Exons 1 and 4 contain separate promoter regions and seem to be the primary first exons used (22). We replaced exons 5 and 6, which correspond to the second and third exons of the MCIP1 mRNA expressed in the heart, with a β-galactosidase reporter fused to the splice acceptor site of exon 5 (Fig. 1A). Proper targeting and processing of the mutant mRNA creates a transcript with either exon 1 or 4 fused in frame to β-galactosidase. Because exons 1 and 4 each encode 29 amino acids of the MCIP1 protein, the targeted gene is expected to generate a fusion protein containing only the most amino-terminal portion of the MCIP1 protein, which does not interact with or inhibit calcineurin (data not shown).

Embryonic stem cells heterozygous for the targeted MCIP1 allele were injected into C57BL6 blastocysts to generate chimeric mice, which transmitted the mutant allele through the germ line. Southern blot analysis of offspring from chimeric mice confirmed proper targeting of the MCIP1 locus (Fig. 1B). Breeding of heterozygous mutant mice yielded MCIP1^−/− mice at the expected Mendelian frequency (Table 1). Homozygous mutants were viable and fertile and showed no obvious abnormalities except that their heart weight (HW)/BW ratios were slightly reduced relative to those of wild-type controls (P < 0.05; Table 1). Given the ability of MCIP1 to suppress the cardiac growth-promoting activity of calcineurin, this finding seemed paradoxical (see below). Because MCIP1 is also expressed in the brain and skeletal muscle and at lower levels in other tissues (11), we performed detailed histopathological analyses of tissues and organs of mutant mice. No abnormalities other than the slight reduction in HW were observed.

Northern blot analysis revealed no detectable MCIP1 mRNA in hearts of homozygous mutant animals (Fig. 2A). Western blot analysis also showed no detectable MCIP1 protein in the hearts or brains of mutant mice (data not shown and E. Bush, personal communication). There was no difference in expression of MCIP2 mRNA in hearts from wild-type and MCIP1^−/− mice (Fig. 2A), ruling out the possibility that it might be up-regulated to compensate for the absence of MCIP1.

The β-galactosidase reporter integrated into the MCIP1 locus was expressed in the same pattern as the endogenous MCIP1 gene during embryonic development. At embryonic day 10, β-galactosidase activity was detectable only in the developing heart (Fig. 2B), which is consistent with earlier published results showing MCIP1 expression primarily in the developing heart at embryonic days 9.5 and 10.5 (23). Examination of histological sections revealed that this staining was localized to the ventricular myocardium. We also examined reporter gene activity in various postnatal tissues of adult MCIP1^+/− animals. β-Galactosidase activity was present throughout the adult heart including the atria (Fig. 2C). Staining was also detected in skeletal muscle containing a high proportion of slow-twitch-type fibers including the soleus and was absent in primarily fast-twitch muscle groups such as white vastus. β-Galactosidase activity was also present in the adult brain, particularly in the hippocampus and dentate gyrus (data not shown).

Calcineurin Activity Is Decreased in MCIP1^−/− Mice.

Based on its up-regulation in response to calcineurin signaling, MCIP1 has been hypothesized to participate in a negative feedback loop to suppress calcineurin activity (19). Surprisingly, we observed a decrease of ≈45% in the level of calcineurin activity in cardiac extracts from MCIP1 null mice (Fig. 3A) that was not due to a decrease in calcineurin protein levels (Fig. 3B). These findings suggested that MCIP1 might be required for maximal activity of calcineurin. It should be pointed out that this calcineurin assay detects maximal potential calcineurin activity in the presence of optimal amounts of calcium and calmodulin. If there were differences in calcineurin activity in wild-type and mutant hearts in vivo due to, for example, differences in subcellular localization or association of the enzyme with positive or negative cofactors, it is unlikely such differences would be detected in this assay.

Exacerbated Response of MCIP1^−/− Mice to Calcineurin Overexpression.

Transgenic mice expressing an activated form of calcineurin (CnA*) in the heart under control of the α myosin heavy chain (αMHC) promoter develop marked cardiac hypertrophy leading to failure and death (2). We bred mice bearing an αMHC-CnA* transgene into the MCIP1^−/− background to determine whether cardiac hypertrophy would be influenced by the absence of MCIP1. However, few MCIP1^−/− animals harboring the CnA* transgene were obtained, preventing a careful examination of the phenotype. Therefore, we also crossed the MCIP1^−/− mice to transgenic mice expressing CnA* under the control of the MCK promoter (MCK-CnA*) (20) that develop cardiac hypertrophy, albeit to a lesser extent than the αMHC-CnA* animals (ref. 20 and Fig. 4A). The hypertrophic response to calcineurin overexpression in this model was exacerbated in the MCIP1^−/− background (Fig. 4A). Mutant mice harboring the MCK-CnA* transgene also developed severe cardiac fibrosis, with large areas of myocyte dropout surrounding the fibrotic zones (Fig. 4A). Terminal deoxynucleotidyltransferase-mediated dUTP end-labeling staining of heart sections revealed no increase in myocyte apoptosis in these hearts (data not shown). The increase of HW/BW ratios caused by the CnA* transgene was larger in the MCIP1^−/− animals than in wild-type littermates (49% vs. 26% increase; Fig. 4B), although this may be an underestimate due to the amount of myocyte dropout. Transthoracic echocardiography was performed to determine whether the lack of MCIP1 exacerbated the functional deficit resulting from calcineurin activation. M-mode echocardiographic analysis revealed dilation and decreased fractional shortening in the MCIP1^−/−, MCK-CnA* background compared with the MCIP1^+/−, MCK-CnA* background (Fig. 4C). The MCK-CnA* transgene is expressed in skeletal muscle as well as cardiac myocytes and has been shown to increase the percentage of type I fibers (20). The skeletal muscle fiber type composition of MCIP^−/−, MCK-CnA* remains to be determined.

Diminished Hypertrophic Response of MCIP1^−/− Mice to TAC.

We next examined the hypertrophic response of the MCIP1^−/− animals in a pressure-overload model of TAC. Unexpectedly, the MCIP1^−/− animals exhibited a smaller degree of hypertrophy than wild-type controls (16% vs. 29% increase in LV weight/BW; Fig. 5A). Echocardiographic measurements confirmed this finding (Table 2). Increases in LV and septal wall thickness in TAC mice were also smaller in the MCIP1^−/− animals compared with wild-type controls. Chamber diameter was similar in the wild-type and MCIP1^−/− animals, indicating that no dilation occurred.

Expression of atrial natriuretic factor (ANF), a fetal cardiac gene and marker of hypertrophy, correlated with the relative changes in LV weight/BW of wild-type and null mice (Fig. 5B). Namely, the increase in ANF expression in banded animals was greater in the wild-type mice compared with the MCIP1^−/− animals (2.8- vs. 1.3-fold). Interestingly, ANF expression was slightly higher in sham MCIP1^−/− animals as compared with wild-type sham controls.

Because MCIP1 expression is activated by calcineurin and expression of the calcineurin-responsive exon 4 isoform of MCIP1 is induced in response to pressure overload (17), we measured induction of MCIP1 gene expression as a measure of calcineurin activation in hearts after TAC. It is possible to measure the level of expression of MCIP1 isoforms initiated from exons 1 or 4, because the targeted allele contains both exons, which are spliced properly to the β-galactosidase reporter. Induction of the MCIP1 exon 4 isoform was greater in wild-type than homozygous null animals in response to aortic banding (3.8- vs. 1.5-fold; Fig. 5C). However, we cannot rule out the possibility that the targeted MCIP1 locus is somehow unresponsive for reasons not related to calcineurin activation or that the stability of the exon 4-LacZ fusion transcript does not allow for an accurate measurement of promoter activity. Expression of the MCIP2 gene did not change in banded animals in either the wild-type or MCIP1^−/− group (data not shown).

Diminished Hypertrophic Response of MCIP1^−/− Mice to Chronic Adrenergic Stimulation.

Chronic delivery of the β-adrenergic agonist, isoproterenol, through an osmotic minipump also results in hypertrophic growth. As shown in Fig. 6, MCIP1^−/− mice showed a diminished hypertrophic response to isoproterenol compared with wild-type controls (16% vs. 41% increase in LV weight/BW, P < 0.05).