Mechanisms of adaptive hypertrophic cardiac remodeling in a large animal model of premature ventricular contraction-induced cardiomyopathy

Abstract

Frequent premature ventricular contractions (PVCs) promoted eccentric cardiac hypertrophy and reduced ejection fraction (EF) in a large animal model of PVC-induced cardiomyopathy (PVC-CM), but the molecular mechanisms and markers of this hypertrophic remodeling remain unexplored. Healthy mongrel canines were implanted with pacemakers to deliver bigeminal PVCs (50% burden with 200-220 ms coupling interval). After 12 weeks, left ventricular (LV) free wall samples were studied from PVC-CM and Sham groups. In addition to reduced LV ejection fraction (LVEF), the PVC-CM group showed larger cardiac myocytes without evident ultrastructural alterations compared to the Sham group. Biochemical markers of pathological hypertrophy, such as store-operated Ca²⁺ entry, calcineurin/NFAT pathway, β-myosin heavy chain, and skeletal type α-actin were unaltered in the PVC-CM group. In contrast, pro-hypertrophic and antiapoptotic pathways including ERK1/2 and AKT/mTOR were activated and/or overexpressed in the PVC-CM group, which appeared counterbalanced by an overexpression of protein phosphatase 1 and a borderline elevation of the anti-hypertrophic factor atrial natriuretic peptide. Moreover, the potent angiogenic and pro-hypertrophic factor VEGF-A and its receptor VEGFR2 were significantly elevated in the PVC-CM group. In conclusion, a molecular program is in place to keep this structural remodeling associated with frequent PVCs as an adaptive pathological hypertrophy.

Keywords: angiogenesis; cardiac arrhythmia; compensatory hypertrophy; signal transduction.

Figure 1

Hypertrophy of the left ventricular (LV) tissue in premature ventricular contractions-induced cardiomyopathy (PVC-CM). Representative LV longitudinal- and cross-sections stained with WGA-AF633 for membrane visualization (red) and with DAPI for nuclear staining (blue). Images were acquired using confocal microscopy. In the longitudinal sections, the length and width were measured for each myocyte and in the cross-sections, the area was determined for each myocyte. These measurements were done using the AIVIA software. Several micrographs per animal were quantified and the data were analyzed using hierarchical statistical analysis (nested t-test; the total number of cells analyzed in the longitudinal sections were 1,790 versus 2,293 myocytes and in cross-sections were 15,270 versus 17,424 myocytes in the PVC-CM and Sham groups, respectively). Each point in the graph represents the mean myocyte size value for the micrograph and the data is also displayed as box and whisker plot.

Figure 2

Ultrastructural determinations using transmission electron microscopy (TEM). TEM micrographs of representative fields of LV thin sections for the indicated experimental groups are shown at two magnifications. An example of a myofibrillar splitting initiation site is indicated by the white arrow.

Figure 3

Evaluation of signaling pathways and molecular markers of pathological hypertrophy. (a) Measurement of SOCE using Mn²⁺ quench of Fura2 fluorescence. Ca²⁺ permeation after intracellular Ca²⁺ store depletion was estimated using Mn²⁺ in freshly isolated myocytes. The rate of Mn²⁺ entry was determined as Slope 2 – Slope 1 difference (ΔSlope). Several myocytes per animal were measured and the data were analyzed using hierarchical statistical analysis (nested t-test; 65 and 51 myocytes were analyzed for PVC-CM and Sham group, respectively); each point represents ΔSlope for a myocyte and the data is also displayed as box and whisker plot. (b) Western blots for the indicated proteins (CALN A, calcineurin A; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; P-NFAT, phosphorylated nuclear factor of activated Tcells [phospho Ser168/Ser170]; β-MHC, β-myosin heavy chain; ACTA1, skeletal type α-actin; ANP, atrial natriuretic peptide) and the position of the closest molecular weight marker is indicated. Each band represents an LV sample obtained from a different animal (n = 6 per condition). (c) The first graph represents the calcineurin activity measured in the LV extract for the different animals in each condition; all the other graphs represent densitometric quantification of the western blots depicted in (b); signals of each band were divided by the GAPDH (loading control) magnitude and normalized to the average value in the Sham group; degree of phosphorylation of NFAT was determined by dividing the signal of phosphorylated versus total NFAT and normalizing to the averaged value in the Sham group. Results are represented by mean ± SD and all p values were > .05 (unpaired t-test was used in CALN activity and blot, NFAT, P-NFAT, β-MHC, and ACTA1 analyses). Note that for ANP experiments, p = .064 (Mann–Whitney test).

Figure 4

Evaluation of mitogen-activated protein kinases (MAPKs) and phosphatases in left ventricular (LV) samples. (a) Western blot for the indicated proteins (P-ERK1/2, phosphorylated extracellular signal-regulated protein kinases 1 and 2 (phospho Thr202/Tyr204); P-JNK1/2/3, phosphorylated c-Jun NH 2-terminal kinases 1, 2, and 3 (phospho Thr183/Thr221); P-p38, phosphorylated p38 kinase (phospho Thr180/Tyr182); GAPDH, glyceraldehyde-3-phosphate dehydrogenase; PP1, protein phosphatase 1; PP2, protein phosphatase, and the position of the closest molecular weight marker is indicated. Each band represents an LV sample obtained from a different animal (n = 6 per condition). (b) Densitometric quantification of the western blots. Intensity of each band was divided by GAPDH (loading control) magnitude and normalized to the average value in the Sham group; degree of phosphorylation was determined by dividing the signal of phosphorylated protein versus total protein and normalizing to the averaged value in the Sham group. Results are represented by mean ± SD and * p < .05 (unpaired t-test in both graphs), **p < .0038 (unpaired t-test), and ***p < .0003 (unpaired t-test).

Figure 5

Evaluation of growth factors associated with hypertrophic and antiapoptotic signalling in left ventricular (LV) tissue. (a) Western blot for the indicated proteins (P-AKT, phosphorylated protein kinase B (phospho Ser473); P-mTOR, phosphorylated mammalian target of rapamycin (phosphor Ser2448); P-S6, phosphorylated ribosomal protein S6 (phospho Ser235/Ser236); P-4EBP1, phosphorylated Eukaryotic translation initiation factor 4E-binding protein 1 (phospho Thr37/Thr46); VEGF-A, Vascular endothelial growth factor A; VEGFR2, VEGF receptor 2; GAPDH, glyceraldehyde-3-phosphate dehydrogenase), and the position of the closest molecular weight marker is indicated. Each band represents an LV sample obtained from a different animal (n = 6 per condition). (b) Signals of the test band were divided by GAPDH (loading control) magnitude and normalized to the average value in the Sham group; degree of phosphorylation was determined by dividing the signal of phosphorylated protein versus total protein and normalizing to the averaged value in the Sham group. Results are represented by mean ± SD and. *p < .05 (unpaired t-test for mTOR and 4EBP1, and Mann–Whitney for VEGFR2), **p < .01 (unpaired t-test used in all graphs), and ***p = .0004 (unpaired t-test).

Figure 6

Schematic representation of the effect of frequent premature ventricular contractions (PVCs) on hypertrophic/antiapoptotic pathways. Frequent PVCs resulted in overexpression of the angiogenic and hypertrophic growth factor VEGF-A. Prohypertrophic and antiapoptotic signaling pathways downstream growth factors ERK, JNK, and Akt/mTOR are hyperphosphorylated and/or overexpressed in LV samples from animals showing PVC-induced cardiomyopathy (PVC-CM). S6 is overexpressed favoring protein synthesis, but 4EBP1 is less phosphorylated and overexpressed, repressing eIF4E opposing protein synthesis. Counter-balancing signals may be at play to keep this hypertrophic response adaptive (see Discussion). Data suggest that frequent PVCs induce an adaptive pathological hypertrophy.