JPH-2 interacts with Cai-handling proteins and ion channels in dyads: Contribution to premature ventricular contraction-induced cardiomyopathy

Abstract

Background: In a canine model of premature ventricular contraction-induced cardiomyopathy (PVC-CM), Cav1.2 is downregulated and misplaced from transverse tubules (T tubules). Junctophilin-2 (JPH-2) is also downregulated.

Objectives: The objectives of this study were to understand the role of JPH-2 in PVC-CM and to probe changes in other proteins involved in dyad structure and function.

Methods: We quantify T-tubule contents (di-8-ANEPPS fluorescence in live myocytes), examine myocyte ultrastructures (electron microscopy), probe JPH-2-interacting proteins (co-immunoprecipitation), quantify dyad and nondyad protein levels (immunoblotting), and examine subcellular distributions of dyad proteins (immunofluorescence/confocal microscopy). We also test direct JPH-2 modulation of channel function (vs indirect modulation through dyad formation) using heterologous expression.

Results: PVC myocytes have reduced T-tubule contents but otherwise normal ultrastructures. Among 19 proteins examined, only JPH-2, bridging integrator-1 (BIN-1), and Cav1.2 are highly downregulated in PVC hearts. However, statistical analysis indicates a general reduction in dyad protein levels when JPH-2 is downregulated. Furthermore, several dyad proteins, including Na/Ca exchanger, are missing or shifted from dyads to the peripheral surface in PVC myocytes. JPH-2 directly or indirectly interacts with Cai-handling proteins, Cav1.2 and KCNQ1, although not BIN-1 or other scaffolding proteins tested. Expression in mammalian cells that do not have dyads confirms direct JPH-2 modulation of the L-type Ca channel current (Cav1.2/voltage-gated Ca channel β subunit 2) and slow delayed rectifier current (KCNQ1/KCNE1).

Conclusion: JPH-2 is more than a "dyad glue": it can modulate Cai handling and ion channel function in the dyad region. Downregulation of JPH-2, BIN-1, and Cav1.2 plays a deterministic role in PVC-CM. Dissecting the hierarchical relationship among the three is necessary for the design of therapeutic interventions to prevent the progression of PVC-CM.

Keywords: Cardiomyopathy; Dyad; Excitation-contraction coupling; Premature ventricular contractions; Sarcoplasmic reticulum; T tubules.

Fig. 1. PVC myocytes have reduced t-tubule contents, lowered JPH-2 protein level but otherwise normal ultra-structures

(A) Orthogonal views of di-8-ANEPPS fluorescence signals from myocytes of SHAM and PVC LVs. (B) Quantification of t-tubule contents based on % of di-8-ANEPPS signals within cell borders. Small symbols represent data points from individual myocytes (2 animals each); large symbols are mean values with standard deviation bars pooled from all myocytes; numbers of myocytes examined in parentheses. (C) Top and middle: transmission electron microscope images of SHAM and PVC myocytes. Bottom: Summary of sarcomere lengths, same format as Fig. 1B. (D) Top: Immunoblot (IB) image of JPH-2 in whole tissue lysates from CON, SHAM and PVC LVs. Bottom: Coomassie blue stain confirming even protein loading among lanes.

Fig. 2. Probe JPH-2 ‘interactome’ in canine LV myocytes by co-immunoprecipitation (co-IP)

JPH-2 goat Ab immunoprecipitates (IP) from canine LVs are probed with Abs targeting proteins of three functional groups (listed on top). (A) Representative immunoblot images of direct input ‘DI’, negative control ‘(−)’, supernatant ‘Super’, and IP, with protein targets and size marker positions noted on the left. The ratio of band intensities in IP vs Super lanes (IP:Super) reflects degree of protein binding to JPH-2. Lower right ‘Self’: IB with JPH-2 rabbit Ab to confirm high efficiency of immunoprecipitation. (B) Summary of ‘IP:Super’ values of proteins detected in JPH-2 IPs. (C) List of proteins not detected in JPH-2 IPs. (n): number of samples tested.

Fig. 3. Confirm direct JPH-2 modulation of I_CaL and I_Ks channels using mammalian cell heterologous expression

(A) a: JPH-2 decreases % Cav1.2 on cell surface without changing whole cell Cav1.2 protein level. Left: Representative IB images of whole-cell lysate (W) and biotinylated fraction (b) from COS-7 expressing cDNAs listed on top, probed with antibodies targeting Cav1.2 (top) and JPH-2 (bottom). Lack of JPH-2 band in ‘b’ lane confirms no cytoplasmic protein contamination. Right: summary from four experiments. Small symbols represent data from individual experiments, with data of ‘− JPH-2’ and ‘+ JPH-2’ linked by lines. Large symbols represent data averaged from experiments with Cav_β2 coexpression. b-d JPH-2 increases I_CaL current density (pA/pF) without altering I_CaL gating kinetics. ‘− JPH-2’ and ‘+ JPH-2’ data are color-coded gray and black. b: I_CaL traces recorded from HEK293 cells transfected with cDNAs marked on top; voltage clamp protocol diagrammed in inset. c: average peak current density-voltage relationships. d: maximal current density (I_max, based on Boltzmann fit, see below) normalized by mean value of ‘− JPH-2’ cells. (B) a: JPH-2 increases % Q1 on cell surface without changing whole cell Q1 protein level. Format and data analysis the same as ‘Aa’. Lack of α-actin band in ‘b’ lanes confirms no cytoplasmic protein contamination. b-e JPH-2 reduces I_Ks current density and alters its gating kinetics. ‘− JPH-2’ and ‘+ JPH-2’ data are color-coded gray and black. ‘Q1 alone’ and ‘Q1/E1’ data are presented by open and closed symbols or histogram bars, respectively. b: current traces recorded by perforated patch clamp of COS-7 cells transfected with cDNAs listed on top; voltage clamp protocol diagrammed in inset. c: average activation curves. d: I_max based on Boltzmann fit and normalized by mean value of ‘− JPH-2’ cells. e: summary of half-maximum activation voltage (V_0.5). To estimate I_max and V_0.5 for each cell, the relationship between peak amplitudes of tail currents (I_tail) and test pulse voltages (V_t) is fit with a Boltzmann function: I_tail = I_max/[1+exp(V_0.5−V_t)/k], where k is a slope factor. The ratio, I_tail/I_max, is expressed as ‘fraction activated’ in ‘c’. In Ad and Bd, numbers in histogram bars are those of cells patch clamped from 3 independent expression experiments. t-test ‘+ JPH-2’ vs ‘− JPH-2’, * p < 0.05, ** p < 0.01.

Fig. 4. JPH-2 does not modulate I_to or I_Kr channels in COS-7 cells despite co-immunoprecipitation of JPH-2 with Kv4.3 and hERG

(A) a: Probing JPH-2 immunoprecipitates from canine ventricular myocytes (left) or COS-7 cells (right, expressing Kv4.3/KChIP2 plus JPH-2) with Kv4.3 mouse mAb (top). The membrane is reprobed with JPH-2 rabbit Ab to confirm effective JPH-2 immunoprecipitation in both cases (bottom). DI, (−) and IP have the same meaning as in Fig. 2. b-d JPH-2 does not affect peak current density or voltage-dependence of inactivation of Kv4.3/KChIP2. (B) a: Probing JPH-2 immunoprecipitates for ERG1 (canine LV myocytes) or hERG (COS-7 cells) (top). The same membrane is reprobed with JPH-2 rabbit Ab (bottom). b-d JPH-2 does not affect peak tail current density or voltage-dependence of activation of hERG. In both panels, ‘−JPH-2’ and ‘+ JPH-2’ data are color-coded gray and black, and voltage clamp protocols are diagrammed in inset of ‘d’.

Fig. 5. JPH-2, BIN-1, and Cav1.2 are uniquely downregulated in PVC-CM

(A) Representative immunoblot images. The original immunoblots have 11 -14 sample lanes, 5 of which are CON sample lanes (for data normalization, see below). For clarity, only two each of CON, SHAM, and PVC lanes are shown, and white lines denote where extra lanes have been removed. Protein targets and size marker positions are noted on the left. Dots on the right mark the bands used for densitometry quantification. Proteins are divided into three functional groups marked on top. (B) Summary of densitometry quantification. (n): numbers of LV samples examined. For every protein examined, each sample is represented by one data point mostly averaged from 2 – 6 independent IB experiments. To allow data averaging from separate IB experiments, for each immunoblot the band intensities are normalized by the mean value of band intensities in CON lanes. Statistics: for each protein, one-way ANOVA is used to evaluate whether there is any significant difference among the CON, SHAM, and PVC groups. If there is, Tukey's test is used for all-pairwise comparisons. * JPH-2, BIN-1 and Cav1.2 protein levels in PVC samples are significantly lower than respective levels in CON and SHAM samples.

Fig. 6. Positive correlation between JPH-2 and other dyad proteins

(A) Selected examples of linear regression between the proteins noted in the panels (normalized protein levels, as described for Fig. 5, plotted along Y-axis) and JPH-2 (X-axis). The probability values (of occurrence by chance) based on linear regression statistic are noted in the panels. (B) Pearson correlation coefficients are plotted against corresponding linear regression probability values for 17 proteins. KCNE2 is excluded. Cav1.2 and BIN-1 are uniquely highly correlated with JPH-2. The other 15 proteins fall along the dotted line with a negative slope.

Fig. 7. Dyad proteins distribution patterns in CON and PVC myocytes

Each panel shows confocal images of fluorescence signals from the same CON (top) or PVC (bottom) myocyte, representing immunofluorescence signals detected by Abs raised in different hosts (Table S1) targeting proteins noted in the panels, or Alexa647-conjugated wheat germ agglutinin (WGA). Horizontal lines point to transverse striations of dyads or t-tubules. Asterisks mark lateral surfaces. Open arrows point to JPH2⁺ vesicles.

Fig. 8

Proposed mechanisms for the deterministic roles of JPH-1, BIN-1 and Cav1.2 in the progression of PVC- CM.