Exploring miRNA-mRNA regulatory network in cardiac pathology in Na⁺/H⁺ exchanger isoform 1 transgenic mice

Abstract

Numerous studies have demonstrated that Na⁺/H⁺ exchanger isoform 1 (NHE1) is elevated in myocardial diseases and its effect is detrimental. To better understand the involvement of NHE1, we have previously studied cardiac-specific NHE1 transgenic mice and shown that these mice develop cardiac hypertrophy, interstitial fibrosis, and cardiac dysfunction. The purpose of current study was to identify microRNAs and their mRNA targets involved in NHE1-mediated cardiac injury. An unbiased high-throughput sequencing study was performed on both microRNAs and mRNAs. RNA sequencing showed that differentially expressed genes were enriched in hypertrophic cardiomyopathy pathway by Kyoto Encyclopedia of Genes and Genomes annotation in NHE1 transgenic hearts. These genes were classified as contraction defects (e.g., Myl2, Myh6, Mybpc3, and Actb), impaired intracellular Ca²⁺ homeostasis (e.g., SERCA2a, Ryr2, Rcan1, and CaMKII delta), and signaling molecules for hypertrophic cardiomyopathy (e.g., Itga/b, IGF-1, Tgfb2/3, and Prkaa1/2). microRNA sequencing revealed that 15 microRNAs were differentially expressed (2-fold, P < 0.05). Six of them (miR-1, miR-208a-3p, miR-199a-5p, miR-21-5p, miR-146a-5p, and miR-30c-5p) were reported to be related to cardiac pathological functions. The integrative analysis of microRNA and RNA sequencing data identified several crucial microRNAs including miR-30c-5p, miR-199a-5p, miR-21-5p, and miR-34a-5p as well as 10 of their mRNA targets that may affect the heart via NFAT hypertrophy and cardiac hypertrophy signaling. Furthermore, important microRNAs and mRNA targets were validated by quantitative PCR. Our study comprehensively characterizes the expression patterns of microRNAs and mRNAs, establishes functional microRNA-mRNA pairs, elucidates the potential signaling pathways, and provides novel insights on the mechanisms underlying NHE1-medicated cardiac injury.

INTRODUCTION

Despite dramatic medical advances over the past years, cardiovascular diseases (CVD) remain the leading global cause of death. According to the American Heart Association Statistics - 2016 Update (62), CVD accounted for 31% of all global deaths in 2013. Nearly 801,000 people in the US died from heart disease, stroke, and other cardiovascular diseases in 2013. That’s about one of every three deaths in America. Therefore, continued research efforts are required to help improve the survival and quality of life of people suffering from CVD.

The mammalian Na⁺/H⁺ exchangers (NHEs) are a family of plasma membrane transport proteins that mediate an electroneutral 1:1 exchange of intracellular H⁺ for extracellular Na⁺ and, in so doing, regulate intracellular pH (pH_i) and cell volume. To date, 10 NHE isoforms have been identified. NHE isoform 1 (NHE1) is ubiquitously expressed in the body and is the predominant NHE isoform in the sarcolemma of cardiomyocytes (18). Besides being a regulator of pH_i and ionic homeostasis, NHE1 also acts as an anchor for actin filaments and a scaffold for an ensemble of signaling complexes. Thus NHE1 regulates a diverse range of cell behaviors, including migration, proliferation, growth, differentiation, cytoskeleton dynamics, and survival (44).

NHE1 has attracted significant attention in CVD field because 1) NHE1 protein expression and activity are elevated in various forms of cardiac injuries, including in hypertensive, hypertrophied, and diabetic myocardium (6, 15, 30, 33), in hearts subjected to ischemia (21), in isolated cardiomyocytes subjected to chronic acidosis (14), as well as in patients with end-stage heart failure (64), 2) this increase in NHE1 have been shown to mediate pathological changes and dysfunction in the heart (32, 34), 3) a plethora of experimental studies have demonstrated salutary effects of NHE inhibition in protecting the myocardium against ischemic and reperfusion injury as well as attenuating myocardial remodeling and heart failure (1, 9, 16, 18, 34), and 4) cardiac-specific overexpression of activated NHE1 itself is sufficient to induce cardiac hypertrophy, fibrosis, and dysfunction in two independent transgenic mouse lines (47, 63).

The purpose of the current study was to investigate the molecular basis of NHE1-mediated cardiac injury. In the past decade, emerging evidence has indicated that microRNAs (miRNAs) are powerful and dynamic modifiers of CVD (11, 12, 22, 37, 61). miRNAs are small noncoding RNAs that regulate gene expression through inhibiting mRNA translation or promoting mRNA degradation. miRNAs play essential roles in mediating many biological processes as well as the pathogenesis of human diseases such as CVD, cancer, and neurological disorders. Thus, the expression of mRNAs and miRNAs was profiled through an unbiased deep-sequencing study. With the aid of bioinformatics tools, miRNA-mRNA regulatory networks were explored. We identified several critical miRNA-mRNA pairs and elucidated potential signaling pathways. The knowledge obtained help us better understand the roles of miRNA-mRNA in cardiac pathology mediated by NHE1.

MATERIALS AND METHODS

NHE1 transgenic mice.

Transgenic mice that expressed constitutively activated NHE1 in the mouse myocardium have been described earlier (2, 10, 28). The NHE1 construct has a mutation in the cytoplasmic regulatory calmodulin binding domain of the protein (Lys641, Arg643, Arg645, and Arg647 were mutated to Glu residues). The mutation causes activation of NHE1 via an alkaline shift in NHE1 pH dependency (3, 10, 28). The expression of the transgene is under control of the cardiac-specific alpha myosin heavy chain promoter. Female mice at the age of 7 mo old were used for all experiments and age-matched nontransgenic littermates were used as controls. All animal protocols (#S05534) were approved by the Institutional Animal Care and Use Committee at University of California San Diego.

Histology.

Hearts were fixed in 10% buffered formalin overnight, embedded in paraffin, serially sectioned into 5 µm slices, and stained with hematoxylin and eosin (H&E) (63). Cross-sectional area (CSA) was measured by computer-assisted image analysis software (ImageJ, NIH Image) (53). Visual fields were accepted for quantification if the nuclei were visible and cell membranes were intact. Ten fields were randomly selected from two sections per heart. Dimensions of ~500 cells per heart were measured. Five hearts were examined for each group (wild-type control vs. NHE1 transgenic). Interstitial fibrosis (IF) was visualized and determined with picro-sirius red staining. Ten fields were randomly selected from two sections per heart. Three hearts were examined for each group (wild-type control vs. NHE1 transgenic). The maximum fibrosis observed for any section was calculated as the area occupied by red-stained connective tissue divided by the areas occupied by connective tissue plus cardiac myocytes × 100. Intramural vessels, perivascular collagen, endocardium, and trabeculae were excluded (63). All the measurements were done by two blinded investigators. Data are presented as means ± SE. Student's t-test was employed, and P < 0.05 was considered statistically significant.

Small and total RNA isolation.

Small RNA (including miRNA) and total RNA were isolated using mirVana miRNA Isolation Kit (AM1560; Life Technologies, Grand Island, NY) according to manufacturer’s instructions. RNA yield was measured using a NanoDrop 1000 (Thermo Fisher Scientific, Wilmington, DE). The RNA quality was examined by Agilent Bioanalyzer with an RNA 6000 Nano Kit (Agilent Technologies, Santa Clara, CA). RNA samples with an RNA integrity number >8 were used for sequencing.

Library construction for the sequencing.

Total RNA (10 μg) was treated with RiboMinus Transcriptome Isolation Kit (Life Technologies, Grand Island, NY). Using 200 ng of rRNA depleted RNA, we constructed mRNA-Seq libraries with ScriptSeq mRNA-Seq Library Preparation Kit (Epicentre, Madison, WI) according to manufacturer’s instructions. Small RNA libraries were constructed with ScriptMiner Small RNA-Seq library prep kit (Epicentre) according to manufacturer’s instructions. We used 100 ng of small RNA for each library. Libraries were checked on a Bioanalyzer with Agilent DNA High Sensitivity Kit (Agilent Technologies) and quantified with quantitative PCR using a KAPA Library Quantification Kit for Illumina Sequencing (KAPA Biosystems, Wilmington, MA). RNA sequencing (RNA-Seq) was done with 10 pM of total RNA libraries, paired end read at 75 cycles each, and miRNA sequencing was done with 10 pM of small RNA libraries, single read at 36 cycles, both using the Illumina Genome Analyzer IIx (Illumina, San Diego, CA).

Transcriptome profiling with RNA-Seq.

Libraries were prepared with three replicates from both control and NHE1 transgenic left ventricle tissue and sequenced using a 75 cycle paired end run on an Illumina platform. The resulting reads were aligned with TopHat2 (36) with default settings for Mus musculus using the mm9 genome index from Illumina iGenome. The aligned reads were processed with Cufflinks (57) to produce a list of differentially expressed genes. The resulting gene was filtered to remove gene where q > 0.05 and the calculated fold change was <1.5×. The data set of sequence analyses can be traced by accession number SRP118869 in the National Center for Biotechnology Information Sequence Read Archive (SRA) (https://www.ncbi.nlm.nih.gov/sra).

miRNA profiling with miRNA-Seq.

Libraries prepared from left ventricle samples of control and NHE1 transgenic mice were sequenced on an Illumina platform using a 36 cycle single end run. Two different methods for determining miRNA expression levels were used on the resulting reads. First, miRNAKey (52) was used to produce a report of both mature and hairpin miRNAs that were differentially expressed in NHE1 transgenic hearts relative to controls. Additionally, miRanalyzer (23) was used to perform the same task to compare the sensitivity of the different methods employed by both pieces of software. The data set of sequence analyses can be traced by accession number SRP118869 in the National Center for Biotechnology Information SRA (https://www.ncbi.nlm.nih.gov/sra).

Functional categorization of differentially expressed mRNAs and miRNAs.

With the aid of The Database for Annotation, Visualization and Integrated Discovery (DAVID) v6.7 program (https://david.abcc.ncifcrf.gov/) (26), Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway annotation (31) was performed on the differentially expressed genes to facilitate biological interpretation of gene function in a network context. Fisher exact test with Benjamini-Hochberg (B-H) correction was used to select significant pathways. The threshold for significance was P < 0.05. Ingenuity Pathways Analysis program (Ingenuity Systems, Redwood City, CA; https://www.ingenuity.com) was chosen to classify differentially expressed miRNAs in terms of cardiac pathologies. The B-H multiple-testing corrected P value was used. The score for each function was shown as –log₁₀ [B-H P value]. The significance threshold was set to a score of 1.3 (i.e., P ≤ 0.05).

qRT-PCR.

Seven differentially expressed miRNAs randomly selected from miRNA-Seq data and seven differentially expressed mRNA targets of microRNA (miR)-30c were validated by qPCR (63). For miRNAs, the standard TaqMan MicroRNA Assays, which employ target-specific stem-loop reverse transcription primers for 3′ extended templates, were carried out according to the manufacturer's instructions (Applied Biosystems, Carlsbad, CA). For mRNAs, the primer sets for specific genes were designed using Primer3 program (https://www.ncbi.nlm.nih.gov/tools/primer-blast/) and synthesized by Invitrogen (Invitrogen, Carlsbad, CA). The fold change in expression level for each specific miRNA and mRNA was calculated by the 2^–ΔΔCt method (39), after normalization to snoRNA202 for miRNA and GAPDH for mRNA (loading control). The final result represents the mean fold change of three individual transgenic samples over controls. Student's t-test was performed and statistical significance was set to P < 0.05.

Integration of miRNA and RNA-Seq data.

Ingenuity Upstream Regulator Analysis (URA) was utilized to identify potential miRNA regulators upstream of significantly changed gene transcripts and to predict whether these miRNA regulators were activated or inhibited. URA is based on expected causal effects between upstream regulators and targets, which are derived from the literature compiled in the Ingenuity Knowledge Base. We examined the known targets of each upstream regulator in our RNA sequencing data set, compared the targets’ actual direction of change to expectations derived from the literature, and then issued a prediction for each upstream regulator. Two scores were used to address two independent aspects of the inference: an “enrichment” score [Fisher’s exact test (FET) P value] that measures overlap of observed and predicted regulated gene sets, and a Z-score that assesses the match of observed and predicted up-/downregulation patterns. An absolute z-score of ≥ 2 was considered significant. An upstream regulator is predicted to be activated if the z-score is ≥2 and inhibited if the z-score ≤−2. When both the z-score (absolute value >2) and the P value (<10^E-3) were significant, the predictions were considered the most reliable. Ingenuity miRNA Target Filter was used to integrate the data sets of differentially expressed miRNAs and gene transcripts of NHE1 transgenic mouse hearts. miRNA-mRNA target relationships were obtained with reciprocal expression patterns and for relevance to cardiovascular signaling pathways.

RESULTS

Cardiac hypertrophy and fibrosis caused by increased NHE1 activation.

NHE1 transgenic hearts have previously been shown to possess approximately threefold increase in NHE1 activity and to have an enlargement in overall size at the age of 2–3 mo old (10, 28, 63). In this study we further characterized the cardiac hypertrophy and fibrosis by examining the cross sections of ventricular cardiomyocytes at the age of 7 mo old. Figure 1, top, represents myocardial cross sections stained with hematoxylin and eosin (H&E) (Fig. 1A) for CSA and with picro-sirius red (Fig. 1B) for IF. Figure 1, bottom, is a quantitative summary of the results, demonstrating that 1) NHE1 activation significantly increased CSA (205.4 ± 11.7%, P < 0.01) as compared with wild-type controls (100.0 ± 10.6%)(Fig. 1A) and 2) no evident fibrosis was detected in the controls, while a significant increase in fibrosis was observed by NHE1 activation (7.6% of area examined, P < 0.01)(Fig. 1B).

Fig. 1.

Elevation of activated Na⁺/H⁺ exchanger isoform 1 (NHE1) induced larger cardiomyocytes and interstitial fibrosis. Histological analysis of the cross sections (×40) of mouse hearts stained with hematoxylin and eosin for cross-sectional area(CSA) (A) and with picro-sirius red for interstitial fibrosis (IF) (B). Top: examples of cross sections from wild-type control and NHE1 transgenic heart. Bottom: summary of CSAs expressed as % of control and fibrosis as % of area examined. **P < 0.01.

Gene expression profile in NHE1 transgenic mouse hearts.

NHE1 transgene-induced gene expression changes were examined at the age of 7 mo, when cardiac pathologies, including cardiac hypertrophy, interstitial fibrosis, and cardiac dysfunction, have been fully developed (10). The fold change of each gene expression was calculated by the ratio of gene expression in NHE1 transgenic mouse hearts to that of the age-matched controls. NHE1 transgenic mice showed a strong transcriptional response (1,331 upregulations and 553 downregulations, P < 0.05) (Supplemental Table S1). (The online version of this article contains supplemental material.) The magnitude of expression alterations was mostly less than threefold (1,205 up and 476 down) (Fig. 2A). However, some genes showed more than fivefold change (32 up and 16 down) (Fig. 2A). Among them, a marked increase (22-fold up) of Slc9a1 was noted in NHE1 transgenic hearts, indicating that NHE1 gene was indeed overexpressed in the transgenic hearts. In addition, two hypertrophic markers, Nppa and Nppb, were increased 44- and 9-fold separately, correlating with a hypertrophy phenotype in NHE1 transgenic mice.

Fig. 2.

Analysis of gene expression profile in mouse hearts with elevated NHE1. A: number of magnitude of gene changes. B: significant Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways that were enriched with differentially expressed genes.

The differentially expressed genes were further subjected to KEGG pathway analysis to determine which significant pathways were enriched with these gene changes. The results revealed that seven KEGG pathways were significantly enriched in the NHE1 transgenic mouse hearts, including hypertrophic cardiomyopathy (HCM) (Fig. 2B). Table 1 lists the genes altered in NHE1 transgenic hearts that are enriched in KEGG HCM pathway. These genes were further classified into three functional categories (Fig. 3): 1) contraction defects, including Myl2, Myh6, Mybpc3, Actb, Des, and Lmna; 2) impaired intracellular Ca²⁺ homeostasis, such as SERCA2a, Ryr2, Rcan1, and CaMKII delta; and 3) signaling molecules that lead to HCM and eventually progress to heart failure, such as Itga, Itgb, ACE1, IGF-1, Tgfb2, Tgfb3, Prkaa1, and Prkab2.

Table 1.

Genes altered in NHE1 transgenic hearts was enriched in KEGG HCM pathway

Symbol	Gene_Name
Atp2a2	ATPase, Ca2+ transporting, cardiac muscle, slow twitch 2
Actb	actin, beta
Ace	angiotensin I converting enzyme (peptidyl-dipeptidase A) 1
Des	desmin
Igf1	insulin-like growth factor 1
Itga1	integrin alpha 1
Itga4	integrin alpha 4
Itga6	integrin alpha 6
Itga9	integrin alpha 9
Itgav	integrin alpha V
Itgb1	integrin beta 1 (fibronectin receptor beta)
Itgb5	integrin beta 5
Lmna	lamin A
Mybpc3	myosin binding protein C, cardiac
Myh6	myosin, heavy polypeptide 6, cardiac muscle, alpha
Myl2	myosin, light polypeptide 2, regulatory, cardiac, slow
Prkaa1	protein kinase, AMP-activated, alpha 1 catalytic subunit
Prkab2	protein kinase, AMP-activated, beta 2 noncatalytic subunit
Ryr2	ryanodine receptor 2, cardiac
Tgfb2	transforming growth factor, beta 2
Tgfb3	transforming growth factor, beta 3

HCM, hypertrophic cardiomyopathy; NHE1, Na⁺/H⁺ exchanger isoform 1.

Fig. 3.

Genes altered in NHE1 transgenic hearts were enriched in KEGG hypertrophic cardiomyopathy (HCM) pathway. Stars, differentially expressed genes in NHE1 transgenic mouse hearts. Red, upregulation; blue, downregulation.

miRNA expression profile in NHE1 transgenic mouse hearts.

NHE1 transgene-induced miRNA expression changes at the age of 7 mo were studied in parallel. The fold change of each miRNA expression was calculated by comparing NHE1 transgenic mouse hearts to their age-matched wild-type controls. miRNA sequencing detected that ~60% of known miRNAs were expressed in the left ventricle of mouse hearts (685 miRNAs in NHE1 transgenic mice and 654 miRNAs in controls). Among them, only 15 miRNAs were differentially expressed (2-fold, P < 0.05) (Table 2) when NHE1 transgenic and control mice were compared by miRanalyzer and miRNAkey. Ten miRNAs were upregulated: miR-208b-3p, miR-206-3p, miR-21a-5p, miR-214-5p, miR-125b-1-3p, miR-21a-3p, miR-146b-5p, miR-379-5p, miR-199a-5p, and miR-34c-5p. Five miRNAs were downregulated: miR-30c-5p, let-7d-3p, miR-9-3p, miR-208a-3p, and miR-499-3p.

Table 2.

Differentially expressed miRNAs in the left ventricle of NHE1 transgenic mouse hearts

	Seq					qPCR
miRNA Name	RPKM (Control)	RPKM (NHE1ox)	Fold Change	P Value	Corrected P Value (Bonferroni)	Fold Change	P Value
mmu-miR-208b-3p	2241.66	23190.52	10.35	0.00E +00	0.00E +00	9.24	0.0004	*
mmu-miR-206-3p	9.49	61.22	6.45	0.00E +00	0.00E +00	23.75	0.0007	*
mmu-miR-21a-5p	18566.08	71506.61	3.85	0.00E +00	0.00E +00	2.64	0.0393	*
mmu-miR-214-5p	7.00	22.52	3.21	1.10E-12	1.29E-09
mmu-miR-125b-1-3p	7.68	24.06	3.13	4.52E-13	5.28E-10
mmu-miR-21-a-3p	22.59	68.68	3.04	0.00E +00	0.00E +00
mmu-miR-146b-5p	464.06	1267.98	2.73	0.00E +00	0.00E +00
mmu-miR-379-5p	140.53	371.39	2.64	0.00E +00	0.00E +00
mmu-miR-199a-5p	235.42	599.51	2.55	0.00E +00	0.00E +00	2.25	0.0569
mmu-miR-34c-5p	98.28	247.26	2.52	0.00E +00	0.00E +00
mmu-miR-30c-5p	1578.11	539.98	−2.92	0.00E +00	0.00E +00	−1.43	0.0001	*
mmu-let-7d-3p	64.16	21.95	−2.92	0.00E +00	0.00E +00	−1.67	0.3922
mmu-miR-9-3p	89.47	28.85	−3.1	0.00E +00	0.00E +00
mmu-miR-208a-3p	13986.97	3826.18	−3.66	0.00E +00	0.00E +00	−4.00	0.1649
mmu-miR-499-3p	36.60	9.01	−4.06	0.00E +00	0.00E +00

miRNA (also miR), microRNA; RPKM, reads per kilobase million; qPCR, quantitative real-time PCR.

^*P < 0.05.

The data were further analyzed with Ingenuity Pathways Analysis (IPA) to better understand the biological significance of these miRNA changes induced by elevated NHE1. Since NHE1 transgenic mice exhibited cardiac pathology, our focus centered on the miRNAs involved in cardiac pathological processes. Accordingly, six miRNAs that have been previously demonstrated by the experiments in relation to cardiac pathological function were altered by the NHE1 transgene too (Table 3). The majority of these miRNAs showed changes in the same direction as expected from the literature. These miRNAs contribute to several cardiac pathologies, such as cardiac fibrosis, cardiac infarction, cardiac hypertrophy, etc. (Fig. 4), which correlated well with the phenotypes we observed in NHE1 transgenic mice.

Table 3.

Differentially expressed miRNAs in cardiac pathologies in the left ventricle of NHE1 transgenic mouse hearts

Category	P Value	Molecules
Cardiac fibrosis	2.9E-06–1.42E-02	miR-199a-5p (and other miRNAs w/seed CCAGUGU), miR-30c-5p (and other miRNAs w/seed GUAAACA), miR-208a-3p (and other miRNAs w/seed UAAGACG), miR-21-5p (and other miRNAs w/ seed AGCUUAU)
Cardiac infarction	1.99E-05–1.99E-05	miR-146a-5p (and other miRNAs w/ seed GAGAACU), miR-1-3p (and other miRNAs w/ seed GGAAUGU), miR-208a-3p (and other miRNAs w/ seed UAAGACG)
Cardiac hypertrophy	4.71E-05–2.64E-03	miR-1-3p (and other miRNAs w/ seed GGAAUGU), miR-199a-5p (and other miRNAs w/ seed CCAGUGU), miR-208a-3p (and other miRNAs w/ seed UAAGACG), miR-21-5p (and other miRNAs w/ seed AGCUUAU)
Cardiac inflammation	2.11E-03–2.11E-03	miR-208a-3p (and other miRNAs w/ seed UAAGACG)
Cardiac enlargement	4.74E-03–4.74E-03	miR-199a-5p (and other miRNAs w/ seed CCAGUGU)
Cardiac damage	6.32E-03–6.32E-03	miR-21-5p (and other miRNAs w/ seed AGCUUAU)
Heart failure	1.06E-02–1.06E-02	miR-199a-5p (and other miRNAs w/ seed CCAGUGU), miR-21-5p (and other miRNAs w/seed AGCUUAU)
Cardiac arteriopathy	1.32E-02–1.32E-02	miR-1-3p (and other miRNAs w/ seed GGAAUGU), miR-208a-3p (and other miRNAs w/seed UAAGACG)
Cardiac necrosis/cell death	1.62E-02–1.62E-02	miR-21-5p (and other miRNAs w/ seed AGCUUAU)

Functional categorization of miRNA changes in NHE1 transgenic hearts was created with the Ingenuity Pathway Analysis program. The threshold for significance was P < 0.05.

Fig. 4.

Functional categorization by cardiac pathology. MicroRNAs (miRNAs) significantly altered by NHE1 transgene were classified into associated cardiac pathological functions (as depicted in x-axis) with Ingenuity Pathway Analysis (IPA) software. Functions are listed from most significant to least. y-Axis: –log₁₀ [Benjamini-Hochberg (B-H) P value]. The significance threshold was set to 1.3 (P ≤ 0.05), as delineated by the horizontal line within the bar graph.

Validation of RNA-Seq and miRNA-Seq results with qPCR.

The seven differentially expressed miRNAs and seven differentially expressed mRNA targets of miR-30c were validated by qPCR. As presented in Fig. 5A, miRNAs, four out of seven tested, i.e., miR-208b-3p, miR-206-3p, miR-21a-5p, and miR-30c-5p, were confirmed by qPCR, and the other three miRNAs, i.e., miR-199a-5p, let-7d-3p, and miR-208a-3p, demonstrated the same trend of changes as in the sequencing study with no statistical significance. For mRNA targets of miR-30c, three out of seven tested were confirmed by qPCR with statistical significance: Igf1, Prkar1a, and Plcb4 (Fig. 5B).

Fig. 5.

Verification of sequencing results with quantitative real-time RT-PCR (qPCR). Seven randomly selected, significantly altered miRNAs (A) and seven differentially expressed mRNA targets of miR-30c (B) from the sequencing studies were tested by real-time RT-PCR. The gene expression level was normalized to snoRNA202 for miRNA and to GAPDH for mRNA. The final result represents the mean fold change of three individual transgenic samples over controls. *P < 0.05, Student's t-test.

Integrated analysis of miRNA and mRNA expression in NHE1 transgenic mouse hearts.

To investigate the contribution of miRNA regulation to differential gene expression in NHE1 transgenic mouse hearts, the gene transcripts that were altered by NHE1 transgene were subjected to Ingenuity URA. We identified 30 mature miRNAs as potential upstream regulators that are responsible for the gene expression changes (Table 4). Seven of these miRNAs were actually changed in NHE1 transgenic hearts (P < 0.05, 2-fold, miRNAkey): miR-30c-5p, miR-21-5p, miR-34a-5p, miR-1-3p, miR-199a-5p, miR-153-3p, and miR-133a-3p. Interestingly, miR-30c-5p showed the same directional change as predicted by URA z-score, strongly indicating that it has a key regulatory role in the observed gene expression alterations.

Table 4.

Mature miRNA regulators upstream of changed genes in NHE1 transgenic hearts

Upstream Regulator	Expr Fold Change	Expr P Value	Predicted Activation State	Bias-corrected z-score	Activation z-score	Flags	P Value of Overlap	Target Molecules in Data Set
miR-16-5p	−1.825	0.00E +00	inhibited		−5.842	bias	1E-08	ACP2, ACTR1A, BCL2, CAPRIN1, CCND1, CFL2, DMTF1, DNAJB4, GOLPH3L, GRN, GTF2H1, H3F3A/H3F3B, HARS, IGF1, IGF2R, KPNA3, LAMC1, LAMTOR3, MAP2K1, MCL1, MRPL20, NPR3, PAFAH1B2, PHLDB2, PISD, PURA, RARS, RTN4, SERPINE2, SKAP2, SLC12A2, SPTLC1, SQSTM1, TMEM109, TPI1, TPPP3, UGDH, UGP2, VTI1B
miR-124-3p	*		inhibited		−5.168	bias	2E-06	ACAA2, ARFIP1, ARPC1B, ATL3, ATP6V0E1, BLOC1S6, CTDSP2, CTGF, CTNND1, F11R, FCHO2, GSN, HADH, HES1, HTATIP2, ITGB1, KLF15, LAMC1, MDFIC, MMP2, MTMR6, MTPN, NECAP2, PGRMC2, POLR3G, PTTG1IP, R BMS1, ROCK1, RYK, SERPINB6, SLC50A1, SSFA2, STAT3, SWAP70, TLN1, TMBIM1, TMEM109, ZNF367
miR-29b-3p	−1.641	0.00E +00	inhibited		−3.429	bias	5E-06	ARPC3, COL15A1, COL3A1, COL4A2, FBN1, GMFB, KLF4, LAMC1, MCL1, PMP22, PPIC, PPM1D, PURA, SPARC, TGFB3, TUBB2A
miR-30c-5p	−2.923	0.00E +00	inhibited		−3.845	bias	1E-05	ATP2A2, BECN1, CTGF, ELMOD2, GALNT1, GNAI2, MAT2A, NPR3, NT5E, PAFAH1B2, PRPF40A, PTGFRN, RBMS1, SEC23A, SEC62, SLC38A2, STRN, STX7, TMED2, TMED7, TMEM59, WDR92
miR-17-5p	−1.300	8.44E-07	inhibited	−1.456	−3.275	bias	5E-05	APP, BCL2, BIRC5, BMPR2, CCND1, CRIM1, PURA, RGS5, RHOA, RHOC, S1PR1, STAT3, TGFBR2, VIM
miR-192-5p	−1.426	0.00E +00		1.453	0.603	bias	9E-05	BIRC5, DHFR, IGF1, TYMS, ZEB1
miR-199a-5p	2.547	0.00E +00	inhibited		−2.376	bias	0.0001	ACTA1, ALOX5AP, BGN, BIRC5, C3, DCN, FN1, HIF1A, ITGA1, KLF4, PIGR, SAMD9L, TGFBR1, TGFBR2, ZBTB16
miR-155-5p	−1.686	8.36E-14	inhibited		−3.418	bias	0.0007	ARFIP1, ATP6V1C1, CCND1, CYR61, DSG2, FAR1, IL13RA1, KRAS, MATR3, MOSPD2, NARS, NT5E, PDLIM5, PICALM, POLE3, RHOA, SDCBP, SERPINE1, TBCA, TCF7L2, TNFRSF10A, TXNRD1, WDFY1
miR-542-3p	1.723	3.73E-08			−1.955	bias	0.0008	ILK, RPL11, RPL22, RPS23
miR-199a-3p			inhibited		−2.433	bias	0.0009	CALU, CD44, FN1, ITGA6, PON2, VCAN
miR-21-5p	3.851	0.00E +00	inhibited		−2.694	bias	0.001	ACTA2, BMPR2, BTG2, C8orf44-SGK3/SGK3, CFL2, MMP2, PECAM1, SERPINB5, STAT3, TCF21, TGFB2, TGFBR2
miR-19b-3p	−1.185	4.89E-03		−0.544	−1.506	bias	0.0016	BIRC5, BMPR2, CCND1, CTGF, THBS1
miR-291a-3p			inhibited		−3.365	bias	0.0016	ADAM9, APP, C3, CCND1, CD44, CFL2, FAM13B, FYCO1, INSIG2, KLHL12, NUP58, PCGF5, STX11, UBXN1, ZHX1
miR-1-3p	6.451	0.00E +00	inhibited		−4.083	bias	0.0019	ANXA2, AP3B1, AXL, BCL2, CAP1, CNN3, CORO1C, FBLN2, FSTL1, GNPDA2, H3F3A/H3F3B, IGF1, IP6K2, KCNJ2, LRP1, PICALM, PPIB, SDC4, SERPINB5, SLC44A1, SNX6, TAGLN2, THBS1, Tmsb4x (includes others), WDFY1, YWHAQ
miR-293-5p					−1.664	bias	0.0031	AKAP6, GOLPH3, KLF15, MGST1, NR4A1, SSR3
miR-34a-5p	2.578	0.00E +00			−1.895	bias	0.0038	BCL2, BIRC5, CCND1, CDK1, DHFR, ICAM1, IKBIP, KLF4, MAP2K1, STIM1
miR-451a	−1.593	0.00E +00			−1.998	bias	0.0041	BCL2, CCND1, MIF, MMP2
miR-1195	*						0.0053	IGF1, STAT3
miR-129-5p					−1.969	bias	0.0061	BMPR2, CAMTA1, GALNT1, PDS5A
let-7a-5p	1.324	0.00E +00	inhibited		−3.219	bias	0.0062	ACTA2, BIRC5, CCND1, CHMP2A, COL3A1, CSDE1, ITGA1, ITGA4, KRAS, MTPN, PGRMC1, SLC25A32, SMOX, SPCS3, STAT3, TGFBR1, THBS1, TLR4, TYMS, VIM
miR-320b							0.0108	IGF1, MCL1, VIM
miR-138-5p	*						0.0108	RHOC, ROCK2, VCAN
miR-153-3p	−2.773	4.73E-07					0.0151	BCL2, MCL1
miR-193a-3p	*						0.0163	CCND1, MCL1, PTK2
miR-30a-3p	−1.132	0.00E +00	inhibited		−2.236	bias	0.0226	CYR61, RAB8B, RSU1, THBS1, TUBA1A
miR-149-5p	−1.622	0.00E +00					0.0288	RAP1A, RAP1B
miR-125b-5p	1.230	0.00E +00	inhibited		−2.224	bias	0.0297	ACSS1, ADAMTS1, GSS, H3F3A/H3F3B, ID2, IGFBP3, LIPA, MAN1A1, MAP2K1, PIGR, SGPL1, VPS4B
miR-133a-3p	−2.692	0.00E +00			−1.875	bias	0.0299	BTBD3, Cdc42, CORO1C, CTGF, KLF15, MCL1, MSN, RHOA
miR-200b-3p	*			−0.39	−1.444	bias	0.0317	ERBIN, ERRFI1, VIM, WDR37, ZEB1
miR-135a-5p	−1.168	3.25E-01					0.0408	ALOX5AP, APC, JAK2

^*Low abundance.

To find out the most biologically relevant mRNA targets of differentially expressed miRNAs, Ingenuity miRNA Target Filter analysis was conducted to explore all the possible miRNA-mRNA interactions that were either experimentally validated from TarBase and miRecords or predicted from TargetScan and Ingenuity Knowledgebase. Among 15 differentially expressed miRNAs, only 10 miRNAs have targeting information available. These 10 miRNAs have 5,385 potential mRNA targets. The expression pairing of our miRNA-mRNA data revealed that eight miRNAs targeting 530 mRNAs did change in NHE1 transgenic mouse hearts. Filtered by inverse expression pattern, it showed eight miRNAs targeting 220 mRNAs. These miRNA-mRNA target relationships were further prioritized either by 1) cardiovascular signaling pathways, six miRNAs targeting 18 mRNAs met the criteria (Table 5) or by 2) cardiovascular disease, the results were narrowed to eight miRNAs targeting 70 mRNAs (Table 6). miR-199a and its mRNA target ITGA4 as well as miR-30c and its nine mRNA targets, CAMK2D, GNAI2, GNAO1, IGF1, KRAS, PPP1R14C, PPKAR1A, RASA1, and TNIP1, have been shown to be involved in cardiovascular signaling and also to be relevant to cardiovascular diseases.

Table 5.

miRNA-mRNA target relationships with opposite expression and related to cardiovascular signaling

ID	Fold Change	Confidence	ID	Symbol	Fold Change
mmu-miR-146b-5p	2.732	experimentally observed	215257	IL36G	−2.024
mmu-miR-199a-5p	2.547	high (predicted)	16401	ITGA4	−1.756
mmu-miR-199a-5p	2.547	high (predicted)	242083	PPM1L	−1.551
mmu-miR-208b-3p	10.345	moderate (predicted)	236915	ARHGEF9	−1.518
mmu-miR-21a-5p	3.851	moderate (predicted)	140491	PPP1R3A	−1.885
mmu-miR-30c-5p	−2.923	high (predicted)	108058	CAMK2D	1.711
mmu-miR-30c-5p	−2.923	high (predicted)	71745	CUL2	1.756
mmu-miR-30c-5p	−2.923	experimentally observed, high (predicted)	14678	GNAI2	2.210
mmu-miR-30c-5p	−2.923	high (predicted)	14681	GNAO1	2.769
mmu-miR-30c-5p	−2.923	high (predicted)	16000	IGF1	2.952
mmu-miR-30c-5p	−2.923	high (predicted)	16653	KRAS	1.558
mmu-miR-30c-5p	−2.923	high (predicted)	108083	PIP4K2B	2.123
mmu-miR-30c-5p	−2.923	high (predicted)	18798	PLCB4	2.080
mmu-miR-30c-5p	−2.923	high (predicted)	76142	PPP1R14C	1.879
mmu-miR-30c-5p	−2.923	high (predicted)	19084	PRKAR1A	2.492
mmu-miR-30c-5p	−2.923	high (predicted)	218397	RASA1	1.582
mmu-miR-30c-5p	−2.923	high (predicted)	57783	TNIP1	2.258
mmu-miR-30c-5p	−2.923	high (predicted)	56550	UBE2D2	1.698
mmu-miR-34c-5p	2.516	moderate (predicted)	236915	ARHGEF9	−1.518

Table 6.

miRNA-mRNA target relationships with opposite expression and related to cardiovascular diseases

ID	Fold Change	Confidence	ID	Symbol	Fold Change
mmu-miR-206-3p	6.451	experimentally observed, high (predicted)	16518	KCNJ2	−1.738
mmu-miR-206-3p	6.451	high (predicted)	243780	KIAA1147	−1.615
mmu-miR-206-3p	6.451	experimentally observed	20724	SERPINB5	−2.811
mmu-miR-146b-5p	2.732	high (predicted)	223455	MARCH6	−1.517
mmu-miR-146b-5p	2.732	experimentally observed	19204	PTAFR	−1.995
mmu-miR-199a-5p	2.547	high (predicted)	16401	ITGA4	−1.756
mmu-miR-199a-5p	2.547	moderate (predicted)	18072	NHLH2	−2.017
mmu-miR-199a-5p	2.547	high (predicted)	19017	PPARGC1A	−2.497
mmu-miR-208b-3p	10.345	high (predicted)	13511	DSG2	−1.540
mmu-miR-21a-5p	3.851	moderate (predicted)	26888	CLEC4A	−1.842
mmu-miR-21a-5p	3.851	moderate (predicted)	15488	HSD17B4	−1.693
mmu-miR-21a-5p	3.851	moderate (predicted)	223455	MARCH6	−1.517
mmu-miR-21a-5p	3.851	experimentally observed, moderate (predicted)	20724	SERPINB5	−2.811
mmu-miR-30c-5p	−2.923	high (predicted)	54375	AZIN1	3.088
mmu-miR-30c-5p	−2.923	high (predicted)	12043	BCL2	3.264
mmu-miR-30c-5p	−2.923	experimentally observed, high (predicted)	56208	BECN1	1.717
mmu-miR-30c-5p	−2.923	high (predicted)	12177	BNIP3L	2.628
mmu-miR-30c-5p	−2.923	high (predicted)	12321	CALU	1.950
mmu-miR-30c-5p	−2.923	high (predicted)	108058	CAMK2D	1.711
mmu-miR-30c-5p	−2.923	high (predicted)	100072	CAMTA1	2.314
mmu-miR-30c-5p	−2.923	high (predicted)	12554	CDH13	2.600
mmu-miR-30c-5p	−2.923	experimentally observed	14219	CTGF	9.390
mmu-miR-30c-5p	−2.923	high (predicted)	13371	DIO2	9.348
mmu-miR-30c-5p	−2.923	high (predicted)	67731	FBXO32	1.552
mmu-miR-30c-5p	−2.923	high (predicted)	108148	GALNT2	1.621
mmu-miR-30c-5p	−2.923	experimentally observed, high (predicted)	14678	GNAI2	2.210
mmu-miR-30c-5p	−2.923	high (predicted)	14681	GNAO1	2.769
mmu-miR-30c-5p	−2.923	high (predicted)	15976	IFNAR2	1.921
mmu-miR-30c-5p	−2.923	high (predicted)	16000	IGF1	2.952
mmu-miR-30c-5p	−2.923	high (predicted)	16004	IGF2R	1.507
mmu-miR-30c-5p	−2.923	high (predicted)	72999	INSIG2	1.964
mmu-miR-30c-5p	−2.923	high (predicted)	16653	KRAS	1.558
mmu-miR-30c-5p	−2.923	high (predicted)	84035	KREMEN1	1.611
mmu-miR-30c-5p	−2.923	high (predicted)	18826	LCP1	2.665
mmu-miR-30c-5p	−2.923	high (predicted)	74511	LRRC17	4.488
mmu-miR-30c-5p	−2.923	high (predicted)	17156	MAN1A2	1.772
mmu-miR-30c-5p	−2.923	experimentally observed, high (predicted)	232087	MAT2A	1.520
mmu-miR-30c-5p	−2.923	high (predicted)	225164	MIB1	2.080
mmu-miR-30c-5p	−2.923	high (predicted)	218613	MIER3	1.614
mmu-miR-30c-5p	−2.923	experimentally observed	18162	NPR3	2.899
mmu-miR-30c-5p	−2.923	high (predicted)	14815	NR3C1	1.857
mmu-miR-30c-5p	−2.923	experimentally observed, high (predicted)	23959	NT5E	2.982
mmu-miR-30c-5p	−2.923	high (predicted)	56426	PDCD10	2.002
mmu-miR-30c-5p	−2.923	high (predicted)	18596	PDGFRB	1.724
mmu-miR-30c-5p	−2.923	high (predicted)	233489	PICALM	1.849
mmu-miR-30c-5p	−2.923	high (predicted)	330260	PON2	2.059
mmu-miR-30c-5p	−2.923	high (predicted)	67738	PPID	1.524
mmu-miR-30c-5p	−2.923	high (predicted)	76142	PPP1R14C	1.879
mmu-miR-30c-5p	−2.923	high (predicted)	19084	PRKAR1A	2.492
mmu-miR-30c-5p	−2.923	high (predicted)	19255	PTPN2	1.875
mmu-miR-30c-5p	−2.923	moderate (predicted)	59021	RAB2A	1.540
mmu-miR-30c-5p	−2.923	high (predicted)	215449	RAP1B	2.562
mmu-miR-30c-5p	−2.923	high (predicted)	218397	RASA1	1.582
mmu-miR-30c-5p	−2.923	experimentally observed	56878	RBMS1	1.683
mmu-miR-30c-5p	−2.923	high (predicted)	56437	RRAD	1.570
mmu-miR-30c-5p	−2.923	high (predicted)	18787	SERPINE1	2.606
mmu-miR-30c-5p	−2.923	high (predicted)	21366	SLC6A6	1.570
mmu-miR-30c-5p	−2.923	high (predicted)	20843	STAG2	1.552
mmu-miR-30c-5p	−2.923	experimentally observed	53331	STX7	1.800
mmu-miR-30c-5p	−2.923	high (predicted)	320165	TACC1	1.630
mmu-miR-30c-5p	−2.923	high (predicted)	21826	THBS2	1.561
mmu-miR-30c-5p	−2.923	high (predicted)	21858	TIMP2	2.602
mmu-miR-30c-5p	−2.923	high (predicted)	57783	TNIP1	2.258
mmu-miR-30c-5p	−2.923	high (predicted)	21951	TNKS	1.680
mmu-miR-30c-5p	−2.923	high (predicted)	70892	TTLL7	3.382
mmu-miR-30c-5p	−2.923	high (predicted)	14479	USP15	1.623
mmu-miR-30c-5p	−2.923	moderate (predicted)	109815	VIMP	1.613
mmu-miR-30c-5p	−2.923	high (predicted)	97064	WWTR1	2.326
mmu-miR-34c-5p	2.516	high (predicted)	14081	ACSL1	−2.295
mmu-miR-34c-5p	2.516	high (predicted)	17161	MAOA	−1.886
mmu-miR-34c-5p	2.516	moderate (predicted)	17318	MID1	−2.236
mmu-miR-34c-5p	2.516	high (predicted)	66052	SDHC	−1.718
mmu-miR-379-5p	2.643	moderate (predicted)	17318	MID1	−2.236

To study the cardiovascular signaling mediated by these miRNA-mRNA targets that lead to cardiac pathology observed in NHE1 transgenic mouse hearts, IPA core analysis was performed. The role of nuclear factor of activated T cells (NFAT) in cardiac hypertrophy [ratio, 10/179 (0.056); z-score, 2.530; B-H P value, 2.48E-05] and cardiac hypertrophy signaling [ratio, 10/223 (0.045); z-score, 2.333; B-H P value, 1.04E-04] appeared to be top ranked canonical pathways; both predicted activation (Fig. 6). Among 179 genes in NFAT hypertrophy pathway, 31 genes were actually changed in NHE1 transgenic mouse hearts (Table 7, Fig. 7). Ten of 31 genes were targets of four differentially expressed miRNAs, miR-30c-5p, miR-21-5p, miR-199a-5p, and miR-34a-5p (Table 8, Fig. 7). Similarly, among 223 genes in cardiac hypertrophy signaling, 35 genes were altered in NHE1 transgenic mouse hearts (Table 9, Fig. 8). Ten of 35 genes were targets of four differentially expressed miRNAs, the same as those in NFAT hypertrophy (Table 8, Fig. 8). The data suggest that these differentially expressed miRNAs may work via regulation of their mRNA targets expression and stimulate NFAT hypertrophy and cardiac hypertrophy signaling, which in turn result in cardiac hypertrophy.

Fig. 6.

Cardiovascular-related canonical pathways involved by miRNA-mRNA pairs. x-Axis, significant pathways and listed from most significant to least; y-axis, –log₁₀ (B-H P value). The significance threshold was set to 1.3 (P ≤ 0.05), as delineated by the horizontal line within the bar graph. Bar colors: orange, predicted pathway activation; blue, predicted inhibition; white, z-score at or very close to 0; gray, no prediction can currently be made. The orange points connected by a thin line represent the ratio (# of genes in a given pathway that meet cut-off criteria, divided by the total # of genes that make up that pathway and that are in the reference gene set).

Table 7.

Differentially expressed mRNAs in NFAT cardiac hypertrophy pathway from left ventricle of NHE1 transgenic mouse hearts

Gene Symbol	Entrez Gene ID	Fold Change	Expected Change	Location	Type(s)
ADCY5	224129	2.388	up	plasma membrane	enzyme
AKAP5	238276	2.822	down	plasma membrane	other
AKT3	23797	1.592	up	cytoplasm	kinase
CAMK1	52163	1.816	up	cytoplasm	kinase
CAMK2D	108058	1.711	up	cytoplasm	kinase
GNAI2	14678	2.210	down	plasma membrane	other
GNAS	14683	1.546	up	plasma membrane	enzyme
GNB1	14688	1.944	up	plasma membrane	enzyme
GNB5	14697	1.769	plasma membrane	enzyme
GNB2L1	14694	1.527		cytoplasm	enzyme
GNG11	66066	1.641	up	plasma membrane	enzyme
IGF1	16000	2.952	up	extracellular space	growth factor
IL11	16156	−1.678	up	extracellular space	cytokine
IL6ST	16195	1.752	up	plasma membrane	transmembrane receptor
KRAS	16653	1.558	up	cytoplasm	enzyme
MAP2K1	26395	2.255	up	cytoplasm	kinase
MAP2K2	26396	1.616	up	cytoplasm	kinase
MAP2K6	26399	−1.948	up	cytoplasm	kinase
PDIA3	14827	1.924	up	cytoplasm	peptidase
PIK3C3	225326	1.668	up	cytoplasm	kinase
PLCB4	18798	2.080	up	cytoplasm	enzyme
PRKAR1A	19084	2.492	up	cytoplasm	kinase
PRKAR2A	19087	1.621	up	cytoplasm	kinase
PRKCD	18753	1.999	up	cytoplasm	kinase
RCAN1	54720	5.226	down	nucleus	transcription regulator
RRAS2	66922	2.233	up	plasma membrane	enzyme
SHC1	20416	1.505	up	cytoplasm	kinase
TGFB2	21808	7.666	up	extracellular space	growth factor
TGFB3	21809	3.013	up	extracellular space	growth factor
TGFBR1	21812	1.746	up	plasma membrane	kinase
TGFBR2	21813	1.787	up	plasma membrane	kinase

Ratio, 31/179 (0.173); z-score, 3.528; P value, 3.26E-05. NFAT, nuclear factor of activated T cells.

Fig. 7.

Differentially expressed miRNA-mRNA pairs in nuclear factor of activated T cells (NFAT) cardiac hypertrophy pathway from NHE1 transgenic mouse hearts. Red, upregulation; green, downregulation.

Table 8.

Differentially expressed miRNAs and their mRNA targets in NFAT hypertrophy pathway and cardiac hypertrophy signaling

Gene Symbol	Fold Change	Expected Change	Target of miRs	Location	Type(s)
CAMK2D*	1.711	up	mmu-miR-30c-5p	cytoplasm	kinase
GNAI2	2.210	down	mmu-miR-30c-5p	plasma membrane	other
IGF1	2.952	up	mmu-miR-30c-5p	extracellular space	growth factor
KRAS	1.558	up	mmu-miR-30c-5p	cytoplasm	enzyme
MAP2K1	2.255	up	mmu-miR-34a-5p	cytoplasm	kinase
PLCB4	2.080	up	mmu-miR-30c-5p	cytoplasm	enzyme
PRKAR1A	2.492	up	mmu-miR-30c-5p	cytoplasm	kinase
TGFB2	7.666	up	mmu-miR-21–5p	extracellular space	growth factor
TGFBR1	1.746	up	mmu-miR-199a-5p	plasma membrane	kinase
TGFBR2	1.787	up	mmu-miR-21-5p, mmu-miR-199a-5p	plasma membrane	kinase
GNAO1†	2.769		mmu-miR-30c-5p	plasma membrane	enzyme

^*In NFAT hypertrophy pathway only;

^†in cardiac hypertrophy signaling only.

Table 9.

Differentially expressed mRNAs in cardiac hypertrophy signaling from the left ventricle of NHE1 transgenic mouse hearts

Gene Symbol	Entrez Gene ID	Fold Change	Expected Change	Location	Type(s)
ADCY5	224129	2.388	up	plasma membrane	enzyme
EIF2B5	224045	1.504	down	cytoplasm	translation regulator
GNAI2	14678	2.210	down	plasma membrane	other
GNAO1	14681	2.769		plasma membrane	enzyme
GNAS	14683	1.546	up	plasma membrane	enzyme
GNB1	14688	1.944	up	plasma membrane	enzyme
GNB5	14697	1.769		plasma membrane	enzyme
GNB2L1	14694	1.527		cytoplasm	enzyme
GNG11	66066	1.641	up	plasma membrane	enzyme
HSPB1	15507	1.836	up	cytoplasm	other
IGF1	16000	2.952	up	extracellular space	growth factor
KRAS	16653	1.558	up	cytoplasm	enzyme
MAP2K1	26395	2.255	up	cytoplasm	kinase
MAP2K2	26396	1.616	up	cytoplasm	kinase
MAP2K6	26399	−1.948	up	cytoplasm	kinase
MYL1	17901	4.363	up	cytoplasm	other
MYL2	17906	−1.755	up	cytoplasm	other
MYL6	17904	2.373	up	cytoplasm	other
MYL7	17898	−2.259	up	cytoplasm	enzyme
MYL12B	67268	1.776	up	cytoplasm	other
PDIA3	14827	1.924	up	cytoplasm	peptidase
PIK3C3	225326	1.668	up	cytoplasm	kinase
PLCB4	18798	2.080	up	cytoplasm	enzyme
PRKAR1A	19084	2.492	up	cytoplasm	kinase
PRKAR2A	19087	1.621	up	cytoplasm	kinase
RHOA	11848	1.888	up	cytoplasm	enzyme
RHOC	11853	2.482	up	plasma membrane	enzyme
RHOQ	104215	1.609	up	plasma membrane	enzyme
ROCK1	19877	1.851	up	cytoplasm	kinase
ROCK2	19878	1.902	up	cytoplasm	kinase
RRAS2	66922	2.233	up	plasma membrane	enzyme
TGFB2	21808	7.666	up	extracellular space	growth factor
TGFB3	21809	3.013	up	extracellular space	growth factor
TGFBR1	21812	1.746	up	plasma membrane	kinase
TGFBR2	21813	1.787	up	plasma membrane	kinase

Ratio, 35/223 (0.157); z-score, 3.889; P value: 8.69E-05.

Fig. 8.

Differentially expressed miRNA-mRNA pairs in cardiac hypertrophy signaling from NHE1 transgenic mouse hearts. Red, upregulation; green, downregulation.

DISCUSSION

NHE1 protein expression and activity are elevated in various cardiac diseases, including ischemia-reperfusion (I/R) injury, cardiac hypertrophy, and heart failure. Two transgenic mouse lines with cardiac-specific overexpression of activated NHE1 develop cardiac hypertrophy, fibrosis and cardiac dysfunction as demonstrated by our group (63) and Nakamura et al. (47). Conversely, NHE1 inhibition has been proven to prevent or induce regression of hypertrophy in different models of cardiac hypertrophy (5, 6, 15, 16, 33, 38, 42). Substantial evidence has accumulated to indicate a key role of NHE1 in the promotion of cardiomyocyte growth. A number of signaling pathways have been proposed to mediate NHE1-induced hypertrophic response (17, 47). However, which mRNAs and miRNAs get involved in this process and how they contribute to cardiac hypertrophy have not been thoroughly investigated. The questions we aimed to answer by the present study were 1) what are the mRNAs and miRNAs that are altered by activated NHE1? 2) what are the functional pairs of miRNAs-mRNAs among these NHE1-induced changes? and 3) what are the signaling pathways that miRNA-mRNA pairs work through to result in cardiac hypertrophy?

Diverse stimuli, such as hormones (endothelin-1, angiotensin II, and α1-adrenergic agonists), growth factors, stretch, I/R injury, and sustained acidosis, activate NHE1 through ERK1/2 (extracellular signal-regulated kinase 1 and 2), p90^RSK (p90 ribosomal S6 kinase), and CaMKII (Ca²⁺/calmodulin-dependent protein kinase) (8, 19, 24, 35, 41, 43, 45, 55, 59, 60). NHE1 activation results in elevation of intracellular Na⁺ concentration ([Na⁺]_i) and subsequent increase in [Ca²⁺]_i via reverse mode of Na⁺/Ca²⁺ exchanger (NCX). Elevated [Ca²⁺]_i in turn activates two important Ca²⁺-dependent prohypertrophic signaling molecules, calcineurin and CaMKII, and causes nuclear translocation of NFAT and nuclear exclusion of histone deacetylase (HDAC)4, thus promoting hypertrophy-associated gene expression (59). For the first time, our study comprehensively profiled expression changes of mRNAs and miRNAs induced by activated NHE1. Moreover, we classified hypertrophic genes into distinct functional categories, which help us better understand their impacts in cardiac hypertrophy and heart failure.

First, our current results showed that activated NHE1 downregulated myosin light polypeptide (Myl2) and heavy polypeptide (Myh6) gene expression in 7 mo old NHE1 transgenic hearts, contrasting with an increased expression at postnatal 2 wk (63). Myosin is a main component of sarcomeres, which is a basic contractile unit of striated muscle tissue. At the younger age of 2 wk, activated NHE1 promotes the synthesis of new contractile proteins and assembly of sarcomeres that ultimately increases contractile force and preserves heart function. But with the aging of the NHE1 transgenic mice, reduced expression of myosin suggests that the hearts lose the ability to maintain contractility and might evolve from cardiac hypertrophy to heart failure.

Second, reduced gene expression of SERCA2a and Ryr2 were detected in the 7 mo old NHE1 transgenic hearts. SERCA2a and ryanodine receptor (RyR)2 are two Ca²⁺ handling molecules located in the sarcoplasmic reticulum (SR). In cardiac myocytes, depolarization elicits a small Ca²⁺ influx through L-type Ca²⁺ channels localized in the t-tubules and subsequently triggers an additional release of Ca²⁺ via RyRs in SR. The amplitude and kinetics of Ca²⁺ release determine cardiac contractile force. Impairments in this release process cause systolic dysfunction. Following contraction, relaxation occurs when Ca²⁺ is recycled into the SR via SR Ca²⁺ ATPase (SERCA) and extruded from the cell via NCX and the plasma membrane Ca²⁺ ATPase. Disrupted Ca²⁺ removal leads to diastolic dysfunction (40, 51). Decreased expression of SERCA2a and Ryr2 correlated well with systolic and diastolic dysfunction observed in NHE1 transgenic mice at the age of 7 mo (failing heart). Our previous study on 2 wk old NHE1 transgenic mice showed no change in expression of SERCA2a and Ryr2 (28, 63), which was in agreement with a study by Nakamura et al. (47) in their NHE1 transgenic mice at the age of 40 days. The difference could be related to age. At an earlier age, NHE1 transgenic hearts are still under compensated hypertrophic phase.

In addition to stimulation of myocyte contraction, Ca²⁺ also plays an important role in regulating the expression of hypertrophy-associated genes. Calcineurin and CaMKII delta appear to be two principal mediators (17, 47). Our data revealed that expression of Rcan1 (regulator of calcineurin 1) and CaMKII delta were upregulated in NHE1 transgenic mice. It is possible therefore that Ca²⁺-dependent hypertrophic signaling is modulated not only at a posttranscriptional level but also at a transcriptional one. A recent study found that CAMTA2 (CaM-binding transcription activator 2) stimulates cardiac growth by opposing class II histone deacetylases (54). Our data showed that CAMTA1 was increased by activated NHE1. Whether CAMTA1 plays a similar hypertrophic role as CAMTA2 needs further investigation.

Third, a myriad of signaling pathways have been implicated in the development of cardiac hypertrophy, including calcineurin-NFAT, MAPKs (ERK, p38, and JNK), RSK, and PI3K/Akt/GSK-3 (20). Our findings are the first to demonstrate that hypertrophic signaling molecules, such as CaMKII, Ras, MEK1/2, PI3K, Akt, PLC, PKA, and PKC, were transcriptionally induced by activated NHE1. A previous study on the same NHE1 transgenic hearts found no differences in phosphorylated ERK, p38, JNK, and RSK at postnatal 2 and 12 wk old as compared with wild-type controls (46). In contrast, Nakamura et al. (47) showed that p38 and ERK were significantly activated in their NHE1 transgenic hearts. The discrepancy is probably caused by a large deletion of NHE1 COOH terminus (amino acid 637–656) in Nakamura’s model, which leads to a more extreme activation of NHE1 protein and an earlier onset of cardiac pathology. Our present data support this concept because these hypertrophic kinases were increased in our model of transgenic hearts at the age of 7 mo old.

Expression patterns of miRNAs were investigated in parallel in NHE1 transgenic hearts in our current study. Upregulation of miR-21, miR-125b, miR-199a, miR-214, miR-208, and miR-146, as well as downregulation of miR-30c, was found not only by our data but also reported in other cardiac hypertrophy models (7, 56, 58), including thoracic aortic-banded hearts, the calcineurin-overexpressed transgenic mice, and phenylephrine-treated neonatal cardiomyocytes, suggesting that they may share a similarity in hypertrophic mechanisms controlled by miRNAs. In particular, the role of certain miRNAs (such as miR-208 and miR-1) has been demonstrated in cardiac hypertrophy. miR-208a and miR-208b are encoded within an intron of alpha- and beta- cardiac muscle myosin heavy chain genes (Myh6, α-MHC and Myh7, β-MHC) respectively. These miRNAs are cotranscribed with their host MHC genes. During normal development, the switch from fetal isoform β-MHC to the adult isoform α-MHC in the mouse occurs shortly after birth (4). Re-induction of β-MHC gene is a well-documented marker of pathological cardiac hypertrophy and normal aging in many experimental models (48). Reduced expression of miR-208a and myh6 and elevated expression of miR-208b and myh7 were noted in NHE1 transgenic hearts. miR-1 is considered as an antihypertrophic miRNA because it negatively regulates CaM, Mef2 (muscle enhancer factor 2), and Gata 4 (27), all of which are components of Ca²⁺-dependent hypertrophic signaling. Our NHE1 transgenic mice showed an increase of miR-1 expression. Whether this miR-1 elevation is a fine modulator of hypertrophy or it is a later consequence of heart failure is not known at present. Future studies will need to examine these possibilities.

miRNAs have profound impact on physiological functions and pathogenesis of diseases through regulating functionally related mRNA targets. We paired expression data of mRNAs and miRNAs in this study to identify important miRNA-mRNA targets. Our integrative analysis uncovered the possibility that miR-30c and its mRNA targets may be key players in NHE1-induced cardiac hypertrophy. Reduced miR-30c in NHE1 transgenic hearts led to increased expression of their mRNA targets. These mRNA targets act in NFAT hypertrophic and cardiac hypertrophic pathways and promote cardiac hypertrophy. miR-30c is abundantly expressed in both cardiac myocytes and cardiac fibroblasts. Reduced expression of miRNA-30c was observed in pathological conditions, such as hypertrophic hearts, diabetic cardiomyocytes, and cancers (25, 29, 49, 50). Downregulation of miR-30c contributes to cardiac hypertrophy, apoptosis, and extracellular matrix remodeling via upregulation of its mRNA targets p53 and p21, Cdc42 and Pak1, as well as CTGF, respectively (13, 29, 49, 50). Our study is the first to link miR-30c with NFAT hypertrophic and cardiac hypertrophic pathways. Future studies are needed to evaluate the therapeutic value of miR-30c in CVD by either gain- and loss-of-function strategy or miR mimic and anti-miR approach.

In summary, our study is the first to perform a global comparative study of mRNA and miRNA expression induced by activated NHE1. Moreover, we categorized the differentially expressed genes in terms of contractility, Ca²⁺ handling/signaling, as well as hypertrophic signaling and found the expression pattern was determined by mouse age examined and severity of cardiac pathology. The current results suggest that NHE1-mediated hypertrophic signaling can be modulated not only at the activity and protein levels but also at the mRNA level. NHE1 activates distinct signaling cascades in response to various stimuli, and hypertrophic gene transcription works as a coincidence detector of many signaling inputs and controls specific sets of genes. The expression of these genes in turn acts along the pathways to fine-tune the hypertrophic response. miR-30c and their mRNA targets were highlighted by integrative analysis of present mRNA and miRNA data. mRNA targets of miR-30c are mainly involved in NFAT and cardiac hypertrophy signaling. We hypothesize that miR-30c may be a promising therapeutic target for cardiac hypertrophy.

GRANTS

This study was supported by National Institutes of Health Grants PO1HD-32573 and RO1NS-037756 to G. G. Haddad.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.X., D.Z., and G.G.H. conceived and designed research; J.X., D.Z., O.P., T.I., and E.X.X. performed experiments; J.X., D.Z., O.P., I.H., T.I., and E.X.X. analyzed data; J.X. and D.Z. interpreted results of experiments; J.X., D.Z., and O.P. prepared figures; J.X. and D.Z. drafted manuscript; J.X., D.Z., O.P., I.H., and G.G.H. edited and revised manuscript; J.X., D.Z., O.P., I.H., T.I., E.X.X., and G.G.H. approved final version of manuscript.