miR-221 and -222 target CACNA1C and KCNJ5 leading to altered cardiac ion channel expression and current density

Abstract

MicroRNAs (miRs) contribute to different aspects of cardiovascular pathology, among others cardiac hypertrophy and atrial fibrillation. The aim of our study was to evaluate the impact of miR-221/222 on cardiac electrical remodeling. Cardiac miR expression was analyzed in a mouse model with altered electrocardiography parameters and severe heart hypertrophy. Next generation sequencing revealed 14 differentially expressed miRs in hypertrophic hearts, with miR-221 and -222 being the strongest regulated miR-cluster. This increase was restricted to cardiomyocytes and not observed in cardiac fibroblasts. Additionally, we evaluated the change of miR-221/222 in vivo in two models of pharmacologically induced heart hypertrophy (angiotensin II, isoprenaline), thereby demonstrating a stimulus-induced increase in miR-221/222 in vivo by angiotensin II but not by isoprenaline. Whole transcriptome analysis by RNA-seq and qRT-PCR validation revealed an enriched number of downregulated mRNAs coding for proteins located in the T-tubule, which are also predicted targets for miR-221/222. Among those, mRNAs were the L-type Ca²⁺ channel subunits as well as potassium channel subunits. We confirmed that both miRs target the 3'-untranslated regions of Cacna1c and Kcnj5. Furthermore, enhanced expression of these miRs reduced L-type Ca²⁺ channel and Kcnj5 channel abundance and function, which was analyzed by whole-cell patch clamp recordings or Western blot and flux measurements, respectively. miR-221 and -222 contribute to the regulation of L-type Ca²⁺ channels as well as Kcnj5 channels and, therefore, potentially contribute to disturbed cardiac excitation generation and propagation. Future studies will have to evaluate the pathophysiological and clinical relevance of aberrant miR-221/222 expression for electrical remodeling.

Keywords: Angiotensin II; Cardiomyocytes; Electrical remodeling; Heart hypertrophy.

Fig. 1

miRNA expression in hearts of mice with severe heart hypertrophy. a miRNA expression was analyzed by next generation sequencing in hearts of WT and KO mice. b In the hearts of KO animals, 14 miRNAs were differentially expressed compared to WT animals (thresholds: ≥ 100 RPM in WT for downregulated miR; ≥ 100 RPM in KO for upregulated miR; fold change ≥ |1.5|; p < 0.01, for detailed information see Supplementary File S1, N = 6 animals/group)

Fig. 2

Altered miR-221/222 expression in hearts of mice with severe heart hypertrophy is due to enhanced miRNA expression in cardiomyocytes. miR-221 (right panels) and miR-222 (left panels) expression were evaluated by TaqMan qRT-PCR (a, relative change compared to WT) and droplet digital PCR (b) in hearts of wild type and knockout animals in an additional cohort. N = 30 animals/group. The increase in miR-221 and -222 could be observed in isolated cardiomyocytes (c, N = 18 animals/group relative change compared to WT). Pri-miR-221/222 was increased in mice with heart hypertrophy (d, whole heart samples N = 30 animals/group, relative change compared to WT). No significant difference in the amount of miR-221 and -222 could be detected in cardiac fibroblasts from wild type or knockout animals (e, N = 10 animals/group, relative change compared to WT). To evaluate if the increase in miR-221 and -222 expression correlates with the expression of a validated cardiac target for these miRNAs, we analyzed p27 mRNA amount (N = 30 animals/group, relative change compared to WT) and protein content in hearts from WT and KO animals (f, N = 19–20 animals/group). Panel f shows a representative Western blot image

Fig. 3

Heart hypertrophy alone is not sufficient to increase cardiac miR-221/222 expression. Infusion of AII or iso induced a comparable increase in HW/TL (a, b) and cardiomyocyte diameter (c, d) compared to control animals. While AII treatment increased miR-221 (e) and miR-222 (f) amount in the hearts of the animals, iso induced no biological relevant change in those miRNAs (g, h) (AII: N = 12–15 animals/group, iso: 5–6 animals/group, relative change compared to control)

Fig. 4

GO-term cluster analysis revealed an enrichment of mRNAs being downregulated and a target for miR-221/222 involved in cation channel complexes and T-tubule function. a Comparison of genes determined by whole transcriptome sequencing, being downregulated in mice with heart hypertrophy to target genes for miR-221/222 revealed an enrichment of genes involved in cardiac action potential (p = 5.30E − 8, g:Profiler, N = 6/group), with a threshold of p ≤ 0.01, FDR < 0.05. b Real-time qRT-PCR with heart samples from an additional cohort of mice w/o heart hypertrophy confirmed these findings (N = 30 animals/group, relative change compared to WT)

Fig. 5

miR-221/222 reduce L-type Ca²⁺ channel current and Kcnj5 (GIRK4) protein amount by targeting Cacna1c or Kcnj5 3′-UTR. a To evaluate if the miRNAs bind to the 3′-UTR dual luciferase constructs containing the 3′-UTR from the L-type Ca²⁺ channel subunits, the potassium channel subunits, the seed sequence or an empty vector was transfected in HEK293 cells either with or without scramble or miR-221 mimic. miR-221 mimic reduced the luciferase activity of the Cacna1c-II, Cacna1c-III, and the Kcnj5 construct. b While miR-222 mimic reduced the luciferase activity for Cacnb2, Cacna1c-I, Cacna1c-III, Kcnj5, and Kcnd2 significantly (N = 4–9 experiments/group). c HL-1 cells were transfected either with scrambled or mimics for miR-221. After 48 h, the mRNA for Cacnb2 and Cacna1c was reduced (N = 5–6 wells/group, relative change compared to scramble). d To confirm the effect on L-type Ca²⁺ channel, we performed patch clamp analysis in HL-1 cells transfected with scramble, miR-221 or miR-222 mimics. miR-221 and -222 mimic reduced the I_Ca,L current (n = 19–46 cells/group, N = 3–5 experiments). Representative current tracings for control and mimic are given. e Western blot analysis for GIRK1, GIRK4, and Kcnd2 was performed in HL-1 cells treated either with scrambled (control) or miR-221 mimics (N = 3 per group). f HL-1 cells were transfected either with scramble, miR-221 or miR-222 mimics and GIRK4-dependent ion flux was measured by fluorescence changes of a thallium-sensitive dye. In HL-1 cells transfected with miR-221/222 mimics, the area under the curve and thereby the ion flux over time were significantly reduced compared to control cells. (N = 4 experiments, n = 3 wells)

Fig. 6

Cacna1c mRNA correlates with the change in miR-221/222 expression in AII-treated animals. To test if the increase in miRNAs is correlated to the changes in Cacna1c and Kcnj5 expression, we performed real-time qRT-PCR in the two mouse models with pharmacologically induced heart hypertrophy. While in AII-treated animals, Cacna1c was downregulated, there was no change in Cacna1c in isoprenaline-treated animals (a) AII: N = 12–15 animals/group, b isoprenaline: 5–6 animals/group (relative change compared to control)