Mechanisms underlying dilated cardiomyopathy associated with FKBP12 deficiency

Abstract

Dilated cardiomyopathy (DCM) is a highly prevalent and genetically heterogeneous condition that results in decreased contractility and impaired cardiac function. The FK506-binding protein FKBP12 has been implicated in regulating the ryanodine receptor in skeletal muscle, but its role in cardiac muscle remains unclear. To define the effect of FKBP12 in cardiac function, we generated conditional mouse models of FKBP12 deficiency. We used Cre recombinase driven by either the α-myosin heavy chain, (αMHC) or muscle creatine kinase (MCK) promoter, which are expressed at embryonic day 9 (E9) and E13, respectively. Both conditional models showed an almost total loss of FKBP12 in adult hearts compared with control animals. However, only the early embryonic deletion of FKBP12 (αMHC-Cre) resulted in an early-onset and progressive DCM, increased cardiac oxidative stress, altered expression of proteins associated with cardiac remodeling and disease, and sarcoplasmic reticulum Ca2+ leak. Our findings indicate that FKBP12 deficiency during early development results in cardiac remodeling and altered expression of DCM-associated proteins that lead to progressive DCM in adult hearts, thus suggesting a major role for FKBP12 in embryonic cardiac muscle.

Figure 1.

Cardiac Cre-recombinase and FKBP12 levels. (A) Cre⁺ levels in hearts from 1-wk-old male mice (n = 6 for each genotype). (B) FKBP12 levels in hearts from 1-wk-old male mice (n = 3). (C) Cre⁺ levels in hearts from 3-mo-old male mice (n = 6 for each genotype). (D) FKBP12 levels in hearts from 3-mo-old male mice (n = 14, 7, 4, and 14 for FL, MCK-FKD, MHC-Cre⁺, and MHC-FKD, respectively). (E) Cre⁺ levels and FKBP12 levels in hearts from 6-mo-old-male mice. (n = 4 for FL, MCK-FKD, and MHC-FKD, n = 2 for MHC-Cre⁺ only). (F) Cre⁺ levels and FKBP12 levels in hearts from 3-mo-old female mice (n = 6 for FL, MCK-FKD, and MHC-FKD. (G) Relative mRNA levels for FKBP12 and FKBP12.6 in MHC-Cre⁺ and MHC-FKD mice. (n = 3 for each). (H) MHC-Cre⁺ per total proteins in hearts of hemizygous MHC-Cre⁺ mice from Hamilton lab (H-MHC-Cre⁺) versus hemizygous mice from Jackson Laboratories (B6.FVB-g(Myh6-cre)2182Mds/J, which we designate J-MHC-Cre⁺), n = 3, 3-mo-old. In all panels, the levels of proteins were assessed by western blotting with representative western blots shown at the left in each panel. Proteins of interest were normalized to the total protein level in the same lane, normalized to the average expression level in FL hearts and expressed as %FL. Data were analyzed with one-way ANOVAs and P values indicated. Data are plotted as the mean ± SD. Source data are available for this figure: SourceData F1.

Figure 2.

Effects of FKBP12 deficiency on cardiac function in male and female mice. (A, E, I, and M) Ejection fraction (EF) for 3-mo-old male, 6-mo-old male, 3-mo-old female, and 6-mo-old female mice, respectively. (B, F, J, and N) Fractional shortening (FS) for 3-mo-old male, 6-mo-old male, 3-mo-old female, and 6-mo-old female mice, respectively. (C, G, K, and O) Left ventricular interior dimension in diastole (LVIDd) for 3-mo-old male, 6-mo-old male, 3-mo-old female, and 6-mo-old female mice, respectively. (D, H, L, and P) Left ventricular anterior wall thickness during systole (LVAWs) for 3-mo-old male, 6-mo-old male, 3-mo-old female, and 6-mo-old female mice, respectively. 3-mo-old male mice: n = 13 for FL, n = 7 for MCK-FKD, n = 15 for MHC-Cre⁺, and n = 17 for MHC-FKD. For 6-mo-old male mice: n = 11 for FL, n = 8 for MCK-FKD, n = 8 for MHC-Cre⁺, and n = 7 for MHC-FKD. For 3-mo-old female mice: n = 16 for FL, n = 10 for MCK-FKD, n = 13 for MHC-Cre⁺, and n = 13 for MHC-FKD. For 6-mo-old female mice: n = 10 for FL, n = 5 for MCK-FKD, n = 10 for MHC-Cre⁺, and n = 14 for MHC-FKD. Data were analyzed by one-way ANOVAs and P value indicated. Data are shown as mean ± SD.

Figure 3.

Effects of FKBP12 deficiency on Ca ²⁺ handling. (A) Representative Ca²⁺ transients in FL (top, black) and MHC-FKD (bottom, blue) cardiomyocytes. 1 Hz stimulation (E.S.) and spontaneous Ca²⁺ waves (SCW) are indicated with arrows. (B) Average amplitude with 1 Hz stimulation. (C) Time to the peak of the Ca²⁺ transient. (D) Rate of decay of the Ca²⁺ transient. (E) Percent of fibers per mouse with spontaneous Ca²⁺ transients or waves (SCW) occurring after electrical stimulation. (F) Velocity of the fast spontaneous Ca²⁺ transients. (G) Velocity of slow Ca²⁺ waves. (H) Caffeine-induced transients. FL: n = 4 mice, 13 cardiomyocytes, MHC-FKD: 5 mice, 11 cardiomyocytes, MHC-FKD: n = 4 mice, 14 cardiomyocytes. (I) Cytosolic Ca²⁺ levels were assessed with Fura 2. FL: n = 4 mice, 20 cardiomyocytes, MCK-FKD: n = 4 mice, 15 myocytes, MHC-FKD, n = 3 mice, 21 cardiomyocytes. (I and J) Ca²⁺ spark frequency. (K) Ca²⁺ spark amplitude. (L) Time to peak. (M) Decay rate (Tau). (N) Full-width half maximum (FWHM). (O) Full-duration half maximum (FDHM). (P) Calculated spark mass. Ca²⁺ transient and spark data: FL, n = 6 mice, 42 cardiomyocytes. MCK-FKD, n = 5 mice, 42 cardiomyocytes. MHC-Cre⁺, n = 5, 36 cardiomyocytes. MHC-FKD, n = 6 mice, 47 cardiomyocytes. Data were analyzed by a nested one-way ANOVA. Data were obtained from cardiomyocytes from male mice.

Figure S1.

SCW without or before stimulation and during stimulation. (A and B) The percentage of fibers per mouse displaying SCW without or before stimulation (A) and during stimulation (between transients) (B). (C) Representative spontaneous Ca²⁺ transient after the stimulation. (D) Representative slow Ca²⁺ wave after stimulation. Representative long duration Ca²⁺ sparks after 1 Hz pacing. (E) Representative Ca²⁺ sparks were detected in the 20-s period after 1 Hz pacing of cardiomyocytes. (F) Representative long duration Ca²⁺ sparks after 1 Hz pacing. Scale bars are 1 s.

Figure 4.

Effects of FKBP12 deficiency on t-tubule structure. (A) FM-4-64-stained t-tubules in cardiomyocytes from hearts of FL, MCK-FKD, and MHC-FKD mice. (B–E) (B) Sarcomere spacing, (C) T-tubule regularity, (D) LE density, and (E) T-tubule integrity were analyzed using AutoTT (Guo and Song, 2014). Three male mice of each genotype were used for this study and a total of 23, 26, and 18 cardiomyocytes of the FL, MCK-FKD, and MHC-FKD mice. Data were analyzed with nested one-way ANOVAs with P values provided. Data are shown as mean ± SD.

Figure 5.

Expression of ECC proteins. (A) 12-wk-old male hearts from FL, MCK-FKD, MHC-FKD and MHC-Cre mice were examined for (A) Ca_V1.2 expression. FL, n = 12, MCK-FKD, n = 12, MHC-Cre⁺, n = 8, MHC-FKD, n = 7. (B) Total RYR2 expression, FL, n = 16, MCK-FKD, n = 11, MHC-Cre⁺, n = 8, MHC-FKD, n = 13. (C) Phosphorylated RYR2 at S2807 FL, n = 8, MCK-FKD, n = 8, MHC-Cre⁺, n = 8, MHC-FKD, n = 6. (D) CASQ2 expression. FL, n = 16, MCK-FKD, n = 11, MHC-Cre⁺, n = 8, MHC-FKD, n = 14. (E) SERCA2a FL, n = 6, MCK-FKD, n = 6, MHC-Cre⁺, n = 6, MHC-FKD, n = 6. (F) Full-length JPH2 and its fragment expression FL, n = 21, MCK-FKD, n = 16, MHC-Cre⁺, n = 13, MHC-FKD, n = 18. (G) SPEG expression, FL, n = 17, MCK-FKD, n = 22, MHC-Cre⁺, n = 17, MHC-FKD, n = 20. (H) Oxyblot on cardiac homogenates from MHC-Cre⁺, FL, MCK-FKD, and MHC-FKD mice. (n = 6 for each group). Proteins of interest were normalized to the total protein level in the same lane, normalized to the average expression level in FL hearts, and expressed as %FL. Each band was analyzed. Data were analyzed by one-way ANOVAs. P values are indicated. Data are shown as mean ± SD. Source data are available for this figure: SourceData F5.

Figure 6.

Immunoprecipitation of JPH2 and RYR2 from hearts of FL, MCK-FKD, MHC-Cre⁺ and MHC-FKD. JPH2 and RYR2 were immunoprecipitated from hearts of FL, MCK-FKD and MHC-FKD mice using procedures we have previously described (Lee et al., 2023). Proteins in the IP were identified by mass spectrometry. (A) Volcano plot of the JPH2 binding proteins from the hearts of FL mice. (B) A volcano plot of the JPH2 binding proteins from the hearts of MCK-FKD mice. (C) A volcano plot of the JPH2 binding proteins from the hearts of MHC-FKD mice. (D) Volcano plot of the RYR2 binding proteins from the hearts of FL mice. (E) Volcano plot of the RYR2 binding proteins from the hearts of MCK-FKD mice. (F) Volcano plot of the RYR2 binding proteins from the hearts of MHC-FKD mice. (G) Normalized FKBP12.6 in RYR2 IPs. (H) Phosphorylation of RYR2 from PRM at the SPEG phosphorylation site. In these studies, we could not distinguish between phosphorylation at S2367 and S2368. (I) Phosphorylation of RYR2 by PRM at the PKA phosphorylation site (S2807). No differences were detected among FL, MCK-FKD, and MHC-FKD in either the JPH2 or RYR2 IPs. JPH2 IPs: n = 11 for each genotype. RYR2 IPs, FL, n = 7, MCK-FKD, n = 10, MHC-FKD, n = 10. Data in G–I are shown as mean ± SD.

Figure 7.

Differential expression and pathway enrichment analysis of RNA-seq on 12-wk-old mouse whole-heart samples by genotype. (A–C) Volcano plots of the log₂ fold changes in gene expression and Benjamini-Hochberg adjusted P values for (A) MHC-FKD, (B) MCK-FKD, and (C) MHC-Cre⁺ versus FL hearts. Black: P_adj > 0.05 & abs(logFC) < 1; Orange: P_adj < 0.05 & abs(logFC) < 1; Blue: P_adj > 0.05 & abs(logFC) > 1; Red: P_adj < 0.05 and abs(logFC) > 1. Genes with adjusted P value <0.05 are labeled. (D) Venn diagrams of the overlap in significantly downregulated and upregulated genes (adjusted P value <0.05) in MHC-FKD, MCK-FKD, and MHC-Cre⁺ versus FL hearts. (E) Plot of the most significantly enriched IPA cardiotoxicity functions as determined by the Benjamini-Hochberg adjusted P values for the Fisher’s exact tests on the overlap between the set of 348 analysis-ready genes with unadjusted P value <0.05 (203 up and 145 down) and the set of genes in each toxicological function. The set of differentially expressed analysis-ready genes associated with each cardiotoxicity function is displayed on the corresponding bar. N = 6 mice per genotype.

Figure S2.

Differentially expressed genes in total RNA from MHC-FKD versus MHC-Cre ⁺ hearts. (A) Volcano plot of the log₂ fold changes in gene expression and Benjamini-Hochberg adjusted P values for MHC-FKD versus MHC-Cre⁺ hearts. (B) Volcano plot of the log₂ fold changes in gene expression and Benjamini-Hochberg adjusted P values for MHC-FKD versus MCK-FKD hearts. Black: P_adj > 0.05 and abs(logFC) < 1; Orange: P_adj < 0.05 and abs(logFC) < 1; Blue: P_adj >0.05 and abs(logFC) > 1; Red: P_adj < 0.05 and abs(logFC) > 1.

Figure 8.

Evaluation of proteins identified in RNA-seq. (A–F) Protein expression of several RNA-seq detected transcripts was validated using western blot of hearts from 3-mo-old mice. Proteins of interest were normalized to the total protein level in the same lane, normalized to the average expression level in FL hearts, and expressed as %FL. Data were analyzed by one-way ANOVAs with P value indicated, n = 6 for all genotypes. Data are shown as mean ± SD. Source data are available for this figure: SourceData F8.

Figure 9.

Effects of FKBP12 deficiency on 1-wk-old mice. (A) Protein expression of several RNA-seq detected transcripts was validated using western blot of hearts from (A) 1-wk-old mice. Proteins of interest were normalized to the total protein level in the same lane, normalized to the average expression level in FL hearts, and expressed as %FL. Data were analyzed by one-way ANOVAs with P value indicated. n = 6 for all genotypes. Data are shown as mean ± SD. (B) Oxyblot and analysis of oxidized proteins in hearts of 7-day-old mice of each genotype. (C) Iodine contrast microCT for cardiac phenotyping of P0 FL and MHC-FKD mice. Micro CT 3D imaging was performed on iodine contrasted FL (n = 5) and MHC-FKD (n = 8) neonates at birth (postnatal day 0, P0) to examine the gross morphology and phenotyping. 3D whole volume rendering and digitally sectioned at sagittal, coronal, and transverse axes and were a mixture of males and females (seven females, eight males). Sex differences were not detected. (D) Digital image showing cardiac structure of the right atrium (RA), left atrium (LA), right ventricular wall (RW), right ventricle (RV), ventricular septum (VS), left ventricle (LV), and left ventricular wall (LW) were examined. The thickness of the left/right ventricular walls and septum and the width of the left/right ventricles were measured along the dashed red line for each sample. (E and F) Quantitative analysis shows significant increases the (E) cross section of ventricles (P = 0.0045) and (F) LV Chamber width (P = 0.0478) in MHC-FKD neonates. (G and H) RV chamber (G), LV wall (H). (I and J) (I) RV wall and (J) ventricular septum (all datapoints provided) were not significantly different. Data were analyzed by one-way ANOVAs with P value indicated. Data are shown are mean ± SD. Source data are available for this figure: SourceData F9.