miR-222 is necessary for exercise-induced cardiac growth and protects against pathological cardiac remodeling

Abstract

Exercise induces physiological cardiac growth and protects the heart against pathological remodeling. Recent work suggests exercise also enhances the heart's capacity for repair, which could be important for regenerative therapies. While microRNAs are important in certain cardiac pathologies, less is known about their functional roles in exercise-induced cardiac phenotypes. We profiled cardiac microRNA expression in two distinct models of exercise and found microRNA-222 (miR-222) was upregulated in both. Downstream miR-222 targets modulating cardiomyocyte phenotypes were identified, including HIPK1 and HMBOX1. Inhibition of miR-222 in vivo completely blocked cardiac and cardiomyocyte growth in response to exercise while reducing markers of cardiomyocyte proliferation. Importantly, mice with inducible cardiomyocyte miR-222 expression were resistant to adverse cardiac remodeling and dysfunction after ischemic injury. These studies implicate miR-222 as necessary for exercise-induced cardiomyocyte growth and proliferation in the adult mammalian heart and show that it is sufficient to protect the heart against adverse remodeling.

Figure 1. MicroRNAs are differentially regulated by exercise

A. and B. Heart weight/body weight (HW/BW) and heart weight/tibia length (HW/TL) ratios of sedentary control (control) and swum (swim) mice. n=9 mice per group. C. and D. HW/BW and HW/TL ratios of sedentary control (control) and voluntary-wheel running (run) mice. n=4 mice per group. E. Heat map of 40 differentially regulated miRNAs concordantly altered in hearts from swimming and running mice compared to sedentary controls. n=3 with 3 mouse hearts per pool. F. qRT-PCR analysis of the identified 40 differentially regulated miRNAs in hearts from separate cohorts. n=5 hearts per group. G. qRT-PCR analysis of miR-222 expression in isolated adult cardiomycytes and non-cardiomycytes from voluntary wheel-running mice. n=4 mouse hearts per group. H. qRT-PCR analysis of miR-222 levels in serum from heart failure patients before (con) and after acute exercise. Data are shown as fold change of miR-222 expression normalized to spiked cel-miR-39. n=28. *p<0.05, #p<0.05, **p<0.01 versus respective control using Student’s test. Data shown as mean±SEM.

Figure 2. miR-222 regulates cardiomyocyte hypertrophy, proliferation and apoptosis in vitro

By using miR-222 gain- and loss-of-function in NRVMs, miR-222’s effects on cardiac cell size, proliferation, and hypertrophic markers were examined. A and B: Immunohistochemical staining for sarcomeric α-actinin followed by quantification of cardiomyocyte area as described in methods. Cells were transfected with control or miR-222 precursor in A and with control antimiR (ctl-anti) or antimiR-222 (anti-222) in B. At least 200 cells were quantified in each group. These data demonstrate miR-222 is necessary and sufficient to induce cardiomyocyte hypertrophy. C. Quantification of EdU and Ki67 staining, as well as cell number from primary NRVMs transfected with control precursor (ctl-pre) or miR-222 precursor (pre-222). D. Quantification of EdU and Ki67 staining, and cell number from NRVMs transfected with control antimiR (ctl-anti) or antimiR-222 (anti-222). These data demonstrate miR-222 is necessary and sufficient to induce proliferation of NRVMs. E. Flow cytometry analysis of TUNEL staining in cardiomyocytes treated with control precursor (ctl-pre) or miR-222 precursor (pre-222) or control antimiR (ctl-anti) or antimiR-222 (anti-222). Cardiomyocye apoptosis was induced by serum deprivation (SD) or serum deprivation plus doxorubicin (SD+doxorubicin). These data demonstrate miR-222 inhibits cardiomyocyte apoptosis. F. QRT-PCR for markers of cardiomyocyte hypertrophy and/or pathology in NRVMs treated with control (ctl-pre) or miR-222 precursor (pre-222). These data demonstrate that miR-222 induces a physiological pattern of gene expression. Data are shown as mean±SEM fold-change of gene expression normalized to U6 and reflect at least three independent experiments. Scale bar: 100 μm. *p<0.05, **p<0.01 versus respective control using Student’s test.

Figure 3. miR-222 targets in cardiomyocytes

A and B. qRT-PCR and immunoblotting were used to analyze RNA and protein levels of the four putative miR-222 targets in neonate cardiomyocytes treated with control precursor (ctl-pre), miR-222 precursor (pre-222), control antimiR (ctl-anti), or antimiR-222 (anti-222), respectively. Data are shown as fold-change in gene expression normalized to U6 in (A). These data demonstrate that miR-222 decreases RNA and protein levels for all four targets in primary cardiomyocytes. C. Luciferase assays of COS7 cells co-transfected with control precursor (ctl-pre) or miR-222 precursor (pre-222) and reporter plasmids containing 3′UTR wild-type or mutated miR-222 binding sites for each of the putative target genes. These data demonstrate all three candidates are direct targets of miR-222. D. Immunoblotting of the four genes in neonate cardiomyocytes transfected with control siRNA (ctl-si) or siRNAs of p27 (sip27), Hmbox1 (siHmbox1), Hipk2 (siHipk2) and Hipk1 (siHipk1) demonstrate effective knock-down for all. HSP90 was used as a loading control. E. Flow cytometry for EdU in neonate cardiomyocytes transfected with indicated siRNAs demonstrates that knockdown of p27 or HIPK1 increases EdU incorporation in cultured cardiomyocytes, consistent with proliferation. F and G. Neonate cardiomyocytes cultures were stained for sarcomeric α-actinin to identify cardiomyocytes, and cardiomyocyte number and area were quantified. Knockdown of p27 and HIPK1 increases cardiomyocyte proliferation (F) while Hmbox1 knockdown increases cardiomyocytes size (G). At least 200 cells or 30 images were quantified in each group. Data represent the mean±SEM from at least three independent experiments. Scale bar: 100 μm. *p<0.05, **p<0.01 versus respective control using Student’s test.

Figure 4. miR-222 is necessary for exercise-induced cardiac growth

A. Sedentary or swum mice were intravenously injected with LNA-antimiR-222 or control LNA-antimiR for 3 weeks prior to quantification of cardiac miR-222 to demonstrate effective reductions of mature miR-222 not pre-miR-222. B and C. HW/BW and HW/TL ratios are shown from sedentary control (sed, n=15) and swum (swim, n=16) mice without injection, sedentary mice injected with control LNA-antimiR (sed ctl-anti, n=7) or LNA-antimiR-222 (sed anti-222, n=7), and swum mice injected with control LNA-antimiR (swim ctl-anti, n=7) or LNA-antimiR-222 (swim anti-222, n=6). These studies demonstrate that LNA-antimiR-222 completely blocks cardiac hypertrophy in response to swimming. D. Quantification of cardiomyocyte area from heart sections stained with wheat germ agglutinin (WGA) (n=5–6, ~500 cells per animal) demonstrate that LNA-antimiR-222 also blocks exercise-induced cardiomyocyte hypertrophy. Immunohistochemical staining for phospho-histoneH3 (pHH3) E and Ki67 (F) from heart sections and quantification (n=5–6 hearts in each group) demonstrate that LNA-antimiR-222 reduces markers of cardiomyocyte proliferation. G. Western blot results show the effect of LNA-antimiR-222 on miR-222 targets in hearts after exercise. Scale bar: 100 μm in D and 50 μm in E and F. Error bars stand for standard errors. *p<0.05, versus sed ctl-anti or sedentary control; #p<0.05 versus swim ctl-anti using One-way ANOVA.

Figure 5. Cardiac-specific expression of miR-222 protects against cardiac remodeling and dysfunction after ischemic injury

A. Tiphenyltetrazolium chloride (TTC) staining to delineate infarct area and fluorescent microsphere distribution to define the area-at-risk (AAR) in hearts from tTA single (miR-222−/tTA+) and double transgenic (miR-222+/tTA+) mice, 24 hours after reperfusion after ischemia, demonstrate no difference in initial infarction in miR-222 expressing hearts. Representative photographs of TTC staining and fluorescent microsphere distribution (bottom) of medial sections of cardiac tissues are shown (n = 6–7 in each group). Scale bar: 1000 μm. B. Cardiac fractional shortening and left ventricular internal dimension in systole (LVIDs) as measured by transthoracic echocardiography in tTA single (miR-222−/tTA+) and double transgenic (miR-222+/tTA+) mice at baseline, 24 hours or 6 weeks after ischemic injury (n = 8–9 mice in each group). These data demonstrate similar cardiac dysfunction at 24 hours but better function and less dilation in miR-222 expressing hearts at 6 weeks. C. Masson trichrome staining (n=6–7 hearts in each group) demonstrates less fibrosis in miR-222-expressing double transgenic hearts at 6 weeks after ischemic injury. Scale bars are shown. D and E. Immunofluorescence demonstrates increased EdU incorporation in cardiomyocytes in miR-222-expressing hearts 6 weeks after ischemic injury but reduced EdU incorporation in non-cardiomyocytes (n=5 animals in each group, ~2500 cells counted per animal) (D) and marker of cardiomyocyte proliferation phospho-histoneH3 (pHH3) a increases in miR-222-expressing hearts 1 week after ischemic injury in cardiomyocytes (n=4 hearts in each group) (E). F. TUNEL staining demonstrates reduced apoptosis in cardiomyocytes in miR-222-expressing hearts 1 week after ischemia injury (n=4 hearts in each group). G. Immunoblotting shows reduced expression of HIPK1, HIPK2, HMBOX1 and p27 in miR-222-expressing hearts 1 week after ischemic injury (n=4 hearts in each group). Scale bar: 100 μm. Data shown as mean±SEM. *p<0.05, **p<0.01 versus respective control using Student’s test.