The Ca²⁺-activated cation channel TRPM4 is a negative regulator of angiotensin II-induced cardiac hypertrophy

Considering that hypertension and increased plasma catecholamine levels, which can influence the tendency to develop hypertrophy, were recently reported in global TRPM4-deficient mice [25], we have generated a cardiac-specific TRPM4 knockout mouse line (Trpm4 ^cKO), through crossbreeding of Trpm4 ^flox mice with MLC2a-Cre mice [40] (Fig. 1a). This results in cardiac-specific deletion of exon 15 and 16 from the Trpm4 gene, which encode for the first transmembrane domain of the TRPM4 protein [37]. Western blotting confirmed deletion of the TRPM4 protein from the heart of Trpm4 ^cKO mice (Fig. 1b). To compare blood pressure in awake, freely moving WT and Trpm4 ^cKO mice, we implanted telemetric devices. Blood pressure and heart rate as measured over the course of 2 weeks after transmitter implantation were essentially identical in Trpm4 ^cKO mice and MLC2a-Cre-negative Trpm4 ^flox/− mice (Supplementary Figure 1).

In previous studies, we have shown by different approaches that TRPM4 protein is expressed in mouse heart, both in ventricles and atria, but absent in TRPM4 knockout mice [25]. Western blot analysis showed clear expression in hearts of WT and Cre⁻ Trpm4^flox/− mice, but TRPM4 was absent in the heart of Cre⁺ Trpm4 ^flox/− mice (Trpm4 ^cKO , Fig. 1b). To evaluate the effect of TRPM4 on the development of cardiac hypertrophy, we performed chronic angiotensin II (AngII) infusion in WT and Trpm4 ^cKO mice. AngII is a widely used drug to induce cardiac hypertrophy experimentally [27, 38]. We have used mini-osmotic pumps, implanted under the back skin, to deliver 3 mg/kg/day AngII during 2 weeks. As a negative control, we implanted pumps filled with saline in both genotypes. To analyze the hypertrophic effect of AngII, at the end of the experiment, the heart weight:body weight (HW/BW) and heart weight:tibial length (HW/TL) ratios were compared. In the saline-treated groups, no differences were observed between WT and Trpm4 ^cKO mice. However, in the AngII-treated groups, Trpm4 ^cKO mice displayed a significant increase in HW/BW and HW/TL ratios, compared to WT mice (Fig. 2a, b). To test age-related spontaneous development of hypertrophy in Trpm4 ^cKO and WT mice, we measured HW/BW and HW/TL ratios in ~1-year-old mice. We detected no difference in heart weight or in body weight between the two genotypes (Fig. 2c). To test whether the presence of the Cre recombinase in the Trpm4 ^cKO mice has any effect on the AngII-induced hypertrophic response, we performed AngII infusion experiments on Cre-positive and Cre-negative Trpm4 ^+/+ mice. As shown in Supplementary Figure 2, the presence of Cre recombinase had no significant effect on HW/BW and HW/TL ratios after AngII treatment in our experimental conditions. Finally, we tested whether the AngII treatment had any effect on TRPM4 expression on the WT heart. To this end, we performed TRPM4-specific qPCR and Western blot (Fig. 2d, e), which showed no difference in the amount of TRPM4 mRNA or protein level between control and AngII-treated WT mice.

To further evaluate the hypertrophic response, we performed histological analysis of the cardiac muscle form saline- and AngII-treated mice. Representative pictures from each group are shown in Fig. 3a. The level of cardiac fibrosis was determined with Masson’s trichrome staining. Fibrosis was obvious in the hearts after AngII application (Fig. 3b), but we did not observe differences between the genotypes (Supplementary Figure 3). Instead, the cross-sectional area of WT and Trpm4 ^cKO myocytes were 193.2 ± 14 and 271.4 ± 12 µm², respectively, after AngII treatment (n = 150 cells from 3 mice for each genotype, p < 0.05; Fig. 3c).

To further investigate the altered hypertrophic response in Trpm4 ^cKO mice, we performed qPCR experiments to evaluate the expression level of known cardiac hypertrophy “marker” genes, i.e., α-actin, myosin heavy chain beta (MYH7), atrial natriuretic peptide (ANP), Rcan1 (a negative regulator of calcineurin of which the transcription is highly dependent on calcineurin–NFAT activation) and the α1H subunit of voltage-gated T-type Ca²⁺ channel (CACNA1H), which is also known to be re-expressed upon cardiac hypertrophy [3]. All these markers showed upregulation after AngII treatment (Supplementary Figure 4), but a specific stronger upregulation of α-actin, ANP and Rcan1 was observed in Trpm4 ^cKO compared to WT (Fig. 3d).

AngII application can have at least two distinct effects on the cardiovascular system, vasoconstriction and subsequent blood pressure elevation, and a direct effect on cardiomyocytes via Ang II receptor type 1 (AT1R). To investigate the direct cellular effect of AngII, we used mouse neonatal ventricular myocytes. Cells were isolated from 1- to 2-day-old pups. The second day after isolation, the culture medium was switched to serum-free medium supplemented with 1 µmol/L AngII. After 2 days incubation with AngII, the cell size was determined. Since neonatal myocytes can have a different gene expression profile compared to adult myocytes, we first tested whether TRPM4 is expressed in WT neonatal cells. Western blotting experiments showed a clear expression of TRPM4 protein in WT and a complete absence in Trpm4 ^−/− neonatal hearts (Fig. 4a). Similar to our in vivo findings, there was no significant difference between WT and TRPM4^-deficient cells in surface area under basal condition. However, after 2 days of AngII treatment, the surface area of the WT cells was 445.6 ± 17 µm² compared to 564.2 ± 14 µm² in case of TRPM4^-deficient cells (50–70 cells from 3 independent preparations for each group, p < 0.01) (Fig. 4b).

To further investigate the cellular effect of AngII, we measured Ca²⁺ transients in intact cardiomyocytes during field stimulation (stimulation frequency 1 Hz). Under basal conditions, Ca²⁺ transients were identical in WT and Trpm4 ^−/− myocytes. Application of 1 µmol/L AngII to WT myocytes had a very moderate effect on the amplitude and the AUC of the Ca²⁺ transients. However, Trpm4 ^−/− cells displayed a more pronounced response upon AngII application (Fig. 5a, b). Analyzing the individual Ca²⁺ transients during AngII application, the increase of the peak values in WT and Trpm4 ^−/− myocytes was 4.8 ± 0.9 and 18.1 ± 5.1 %, respectively (n = 15–17 for each group, p < 0.05). Similar differences were found when analyzing the area under curve (AUC) of transients: the mean AngII-induced increase was 0.9 ± 0.6 % in WT and 13.6 ± 3.8 % in Trpm4 ^−/− myocytes (n = 15–17 for each group, p < 0.05). However, the time constant of the decaying phase was not different between genotypes, suggesting that the Ca²⁺ extrusion mechanisms are unchanged in the Trpm4 ^−/− myocytes (Fig. 5c).

Recent studies demonstrated that store-operated Ca²⁺ entry (SOCE) plays a key role in mediating cardiomyocyte hypertrophy, both in vitro and in vivo [15, 16, 30, 32, 33]. To test whether the different effect of AngII on the global Ca²⁺ transients can be explained by differences in SOCE between WT and Trpm4 ^−/− adult ventricular myocytes, we performed a Ca²⁺ re-addition protocol, using AngII to deplete the IP₃ sensitive stores and thapsigargin to block Ca²⁺ reuptake by SERCA activity. Cells were kept in nominally Ca²⁺-free solution in the presence of 1 µmol/L AngII and 5 µmol/L thapsigargin for 5 min and subsequently perfused with a solution containing 2.5 mmol/L Ca²⁺ to allow SOCE. In these experiments, 30 % of the WT and 65 % of the Trpm4 ^−/− myocytes displayed an increase in [Ca²⁺] _i after Ca²⁺ re-addition. The average fluorescent signals from WT and Trpm4 ^−/− myocytes during Ca²⁺ re-addition is shown in Fig. 6a. Analyzing [Ca²⁺] _i during the Ca²⁺ re-addition, we found a significantly higher increase in the AUC in Trpm4 ^−/− myocytes compared to WT (Fig. 6b). This indicates that AngII-mediated SOCE is increased in Trpm4 ^−/− myocytes.

Based on our SOCE experiments, we hypothesized that during AngII application, the release of Ca²⁺ from IP₃-sensitive stores can activate TRPM4, and the depolarizing current via the channel can reduce the driving force for Ca²⁺ entry via SOCE channels, similar to the case in mast cells [37]. This hypothesis also suggests that TRPM4 activation during the store depletion might lead to moderate depolarization of the resting membrane potential. To test this, we performed membrane potential measurement during the above-discussed Ca²⁺ re-addition protocol, using the perforated patch technique (Fig. 6c). In WT myocytes, we observed a significant 7.3 mV membrane depolarization during the store depletion. In contrast, no significant depolarization was measured in Trpm4 ^−/− myocytes (Fig. 6d) (WT: n = 13; Trpm4 ^−/−: n = 14, p < 0.001). In these experiments, we never observed depolarization beyond −60 mV, allowing us to exclude a significant contribution of L-type Ca²⁺ channel activity to the [Ca²⁺] _i increase during Ca²⁺-reapplication. Furthermore, application of the L-type Ca²⁺ channel blocker nifedipine (1 µmol/L) had no effect on the [Ca²⁺] _i increase in Ca²⁺ re-addition experiments (Supplementary Figure 5). Application of 30 mmol/L K⁺ at the end of the experiment depolarized the membrane potential and induced a Ca²⁺ transient, indicating that the myocytes are intact and able to maintain their resting membrane potential.

To test whether the difference in [Ca²⁺] _i increase after Ca²⁺ re-addition between WT and Trpm4 ^−/− myocytes is mediated by a TRPM4-dependent membrane depolarization, we performed the Ca²⁺ re-addition experiment while clamping the membrane potential at −75 mV. Under this condition, we no longer observed a difference in the rise in [Ca²⁺] _i between WT and Trpm4 ^−/− myocytes during the Ca²⁺ re-addition phase (Fig. 6e).

To further investigate the mechanisms behind the altered hypertrophic response in Trpm4 ^cKO mice, calcineurin enzyme activity was measured. Previously, we have seen increased expression of Rcan1, an endogenous reporter of calcineurin–NFAT activity, in AngII-treated Trpm4 ^cKO animals (Fig. 3d), which is most likely the result of the enhanced activation of calcineurin [42]. It has been shown previously that the calcium-activated phosphatase calcineurin is both necessary and sufficient to induce cardiac hypertrophy and that both its protein expression and phosphatase activity are increased in response to hypertrophic stimuli [28, 35]. To test calcineurin activity directly, we performed a calcineurin enzyme activity assay. We observed a significant increase in calcineurin activity in the hearts of AngII-treated animals and this increase was even higher in Trpm4 ^cKO animals (n = 10 for each group, p < 0.05) (Fig. 7a). However in sham-treated animals, total phosphatase activity was not significantly different between WT and Trpm4 ^cKO samples (Supplementary Figure 6). To further investigate calcineurin activity, we performed calcineurin-specific Western blot analysis on the same hearts. In line with the increased phosphatase activity, we observed ~1.6 times more protein in WT AngII-treated and ~2.4 times more in Trpm4 ^cKO AngII-treated animals, whereas there was no significant difference in calcineurin protein level between the genotypes after sham treatment (Fig. 7b). These results suggest that higher SOCE in AngII-treated Trpm4 ^cKO myocytes leads to an enhanced calcineurin signaling and to a stronger hypertrophic response.

To obtain cardiac-specific Trpm4 knockout mice, Trpm4 ^L3F2 mice (see Fig. 1) were crossed with EIIa-Cre mice (obtained from the Jackson Laboratory). From the offspring, Trpm4 ^flox mice were selected for further breeding. Subsequently, Trpm4 ^flox mice were crossed with MLC2a-Cre mice, which resulted in cardiac-specific deletion of exon 15 and 16 of the TRPM4 gene, which encode the first transmembrane domain of the TRPM4 protein. Western blotting confirmed deletion of the protein from the heart (see Fig. 1 in the main text). Mice were routinely genotyped using PCR. For more details on the targeting strategy, see [40]. Experiments were performed on 3- to 6-month-old male C57BL/6 N mice and age- and sex-matched Trpm4 ^−/− or cardiac-specific Trpm4 knockout (Trpm4 ^cKO ) mice. Mice were housed with a 12-h light/12-h dark cycle and allowed water and standard food ad libitum. All animal experiments were performed according to the Guide for the Care and Use of Laboratory Animals (8th edition, National Research Council, USA, 2011) and were approved by the local ethics committee. All mention of Trpm4 ^cKO in this paper refers to cardiac-specific Trpm4 knockout mice.

8- to 12-week-old mice were heparinized and killed by intraperitoneal injection of pentobarbital (Nembutal, CEVA). The hearts were removed and perfused retrograde for 2 min through the aorta with a Langendorff apparatus at 37 °C with oxygen-saturated nominally calcium-free solution containing (in mmol/L): 117 NaCl, 4 KCl, 1 KH₂PO₄, 4 NaHCO₃, 1.7 MgCl₂, 10 HEPES, and 10 glucose (pH 7.4). This step was followed by 4–6 min of perfusion with digestion solution [as above, but supplemented with 1 mg/ml collagenase B (Worthington)]. Subsequently, the heart was removed from the setup, atria were discarded and ventricular muscle was minced in collagenase solution (see above). The suspension was filtered through a nylon filter, and cells were centrifuged at 500×g for 2 min. The supernatant was discarded and the myocytes were resuspended in the first solution, but supplemented with 1 mmol/L CaCl₂. Calcium-tolerant, rod-shaped myocytes with clear striations were used for measurements on the day of the isolation.

For imaging of intracellular Ca²⁺ dynamics, isolated ventricular myocytes were loaded with the Ca²⁺ indicator Fluo-4-AM (1 μmol/L) for 15 min at room temperature. Cells were seeded on glass coverslips in extracellular solution containing (in mmol/L): 118 NaCl, 4.7 KCl, 2.52 CaCl₂, 1.64 MgSO₄, 24.88 NaHCO3, 1.18 KH₂PO₄, 6 glucose, 2 Na pyruvate. This solution was continuously bubbled with carbogen (5 % CO₂, 95 % O₂). Electrical stimulation was achieved using field stimulation. Fluorescence was detected with a photo-multiplier tube and acquired by the Patchmaster software. Excitation light (480 nm) was provided by a Polychrome V monochromator (Till Photonics, Germany).

For store-operated calcium entry measurement, myocytes were handled as above. Myocytes were treated with 1 μmol/L AngII and 5 μmol/L thapsigargin for 5–6 min in nominally Ca²⁺-free buffer to deplete SR stores. Myocytes were then perfused with high Ca²⁺ (2.5 mmol/L), in the continuous presence of thapsigargin and AngII, and store-operated entry was measured as an increase in Fluo-4 fluorescence. For the measurement of the resting membrane potential during SOCE experiment, perforated patch technique was used in current clamp mode. In these experiments the pipette solution contained (in mmol/L) 10 KCl, 10 NaCl, 70 K₂SO₄, 1 MgCl₂ 10 Hepes and 240 µg/ml amphotericin B, pH 7.3.

8- to 12-week-old mice were anesthetized with isoflurane. Mini-osmotic pumps (Alzet, model: 1002) filled with 3 mg/kg/day Ang II (Calbiochem) or with saline were implanted subcutaneously. Two weeks after the operation, animals were killed by cervical dislocation, hearts were removed, cleaned from fat tissue and big vessels, washed in PBS and weighed. Subsequently, the left tibia was dissected and the length of the bone was measured.

Neonatal mice (1- to 2-day-old) were used for isolation. Hearts were removed aseptically; atria and big vessels were discarded. The ventricles were cut into small pieces and digested in 0.25 % trypsin for 15 min at 37 °C. Subsequently, supernatants were removed and mixed with an equal volume of HBSS with Ca²⁺ and Mg²⁺. This step was repeated four to five times until the hearts were completely digested. The cells were then cultured in DMEM supplemented with 20 % FBS for 2–3 h, allowing the attachment of non-myocytes. Later, the cells in the supernatant were used and cultured in DMEM supplemented with 20 % FBS. After 1 day, the medium was changed to DMEM without FBS and supplemented with 1 µM AngII for 48 h.

Tissue samples from freshly isolated hearts of WT and Trpm4 ^cKO mice were lysed in calcineurin assay buffer (Enzo Life Science) and centrifuged at 2000×g. The supernatant was then snap frozen at −70 °C. The phosphatase activity assay (BML-AK816) was performed based on the manufacturer’s instruction (Enzo Life Science). Calcineurin phosphatase activity was measured spectrophotometrically by detecting free phosphate released from the calcineurin-specific RII phosphopeptide.

Proteins from freshly isolated heart of wild-type and Trpm4 ^cKO mice were lysed in 1 ml ice-cold lysis buffer (100 mmol/L Tris–HCl, 1 mmol/L MgCl₂, 0.1 mmol/L phenylmethylsulfonyl fluoride [PMSF] [pH 8] and a protease inhibitors cocktail (Proteoguard™ Clontech) using a Polytron homogenizer. Subsequently, 4 ml saccharose buffer (250 mmol/L saccharose, 10 mmol/L Tris–HCl [pH 7.4], 0.1 mmol/L PMSF and protease inhibitor cocktail) was added to the lysate. The obtained homogenates were centrifuged at 3000×g for 15 min to remove any remaining large cellular fragments. The supernatants were ultracentrifuged at 200,000×g for 30 min. Pellets containing total membrane fractions were solubilized in a cold saccharose buffer. Protein concentrations were determined by the bicinchoninic acid assay method, using bovine serum albumin (BSA) as a standard. Samples (80 μg) were subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE) and subsequently transferred to a polyvinylidene fluoride (PVDF) membrane (Bio-Rad, USA). The respective proteins were detected with purified rabbit polyclonal antibody (Ak578) directed against the amino-terminal end of mouse TRPM4 (27), isoform-specific mouse monoclonal calcineurin-Aß (sc365612; Santa Cruz Biotechnology) and monoclonal mouse anti-Na⁺/K⁺ ATPase (1: 5000 dilution) (Abcam, UK) antibodies. Immunoreactive complexes were visualized by chemiluminescence, using anti-rabbit IgG (Sigma, USA) or anti-mouse IgG (GE Healthcare) antibodies conjugated to horseradish peroxidase.

Four mice of each genotype were perfused with saline. Then the hearts were fixed with zinc-based fixative solution (BD Pharmingen, USA). Subsequently, hearts were processed and embedded in paraffin wax. Serial sections were cut to a thickness of 5–7 μm and documented using a Zeiss microscope (Axiovert 40). For Masson’s trichrome staining, a commercially available kit (Sigma-Aldrich) was used based on the supplier’s instructions. For the analysis of the obtained pictures, ImageJ (NIH, USA) software was used.

After euthanasia by cervical dislocation, hearts were removed, washed in ice-cold PBS, immediately snap frozen in liquid nitrogen and kept at −80 °C until final processing. Total RNA was extracted using the RNeasy Mini Kit (Qiagen), following the manufacturer’s protocol. RNA concentration and quality were assessed using the Experion RNA StdSens Analysis Kit (Bio-Rad). For cDNA synthesis, Ready-To-Go You-Prime First-Strand Beads (GE Healthcare) were used based on the supplier’s protocol. Generated cDNAs were stored at −20 °C. qPCR reactions, composed of cDNA template, Universal TaqMan MasterMix (2× concentrated, Life Technologies), TaqMan assay (20× concentrated, Life Technologies) and H₂O, were performed with the 7500 Fast Real-Time PCR System (Life Technologies). mPGK1 and mTBP, selected using the geNorm application, were used as endogenous controls. TaqMan assays used: CACNA1H-Mm00445382_m1, TRPM4-Mm00613173_m1, ACTA1-Mm00808218_g1, ANP-Mm01255748_g1, RCAN1-Mm01213407_m1, MYH7-Mm00600555_m1.

All statistical analyses were performed using Prism (Graphpad) or Origin (Microcal, USA) software. Data are expressed as mean ± SEM, unless otherwise specified. Differences between groups were compared by two-sample t test and two-way repeated measurement ANOVA followed by a Tukey post hoc test, unless mentioned otherwise. When assumptions were not valid, non-parametric tests were performed (Mann–Whitney test, Friedman ANOVA). A p value <0.05 was considered to be statistically significant.