A novel human R25C-phospholamban mutation is associated with super-inhibition of calcium cycling and ventricular arrhythmia

Abstract

Aims

Depressed sarcoplasmic reticulum (SR) Ca²⁺ cycling, a universal characteristic of human and experimental heart failure, may be associated with genetic alterations in key Ca²⁺-handling proteins. In this study, we identified a novel PLN mutation (R25C) in dilated cardiomyopathy (DCM) and investigated its functional significance in cardiomyocyte Ca²⁺-handling and contractility.

Methods and results

Exome sequencing identified a C73T substitution in the coding region of PLN in a family with DCM. The four heterozygous family members had implantable cardiac defibrillators, and three developed prominent ventricular arrhythmias. Overexpression of R25C-PLN in adult rat cardiomyocytes significantly suppressed the Ca²⁺ affinity of SR Ca²⁺-ATPase (SERCA2a), resulting in decreased SR Ca²⁺ content, Ca²⁺ transients, and impaired contractile function, compared with WT-PLN. These inhibitory effects were associated with enhanced interaction of R25C-PLN with SERCA2, which was prevented by PKA phosphorylation. Accordingly, isoproterenol stimulation relieved the depressive effects of R25C-PLN in cardiomyocytes. However, R25C-PLN also elicited increases in the frequency of Ca²⁺ sparks and waves as well as stress-induced aftercontractions. This was accompanied by increased Ca²⁺/calmodulin-dependent protein kinase II activity and hyper-phosphorylation of RyR2 at serine 2814.

Conclusion

The findings demonstrate that human R25C-PLN is associated with super-inhibition of SERCA2a and Ca²⁺ transport as well as increased SR Ca²⁺ leak, promoting arrhythmogenesis under stress conditions. This is the first mechanistic evidence that increased PLN inhibition may impact both SR Ca²⁺ uptake and Ca²⁺ release activities and suggests that the human R25C-PLN may be a prognostic factor for increased ventricular arrhythmia risk in DCM carriers.

1. Introduction

Dilated cardiomyopathy (DCM), a common cause of heart failure, is characterized by dilation and impaired contraction of the left ventricle. Multiple aetiologies may underlie DCM, including rare variants in more than 30 genes, which initiate diverse pathophysiological mechanisms leading to left-ventricular dysfunction or arrhythmia and increased morbidity and mortality.¹

A common clinical hallmark and characteristic of failing cardiomyocytes is the defective Ca²⁺ homeostasis, exhibited by a prolonged decay time of intracellular Ca²⁺ transient and changes in systolic and diastolic Ca²⁺ levels.^2–4 The prolonged Ca²⁺-transient decay time and increased diastolic Ca²⁺ levels may result from impaired ryanodine receptor function, decreased sarcoplasmic reticulum Ca²⁺-ATPase (SERCA2a) levels, and augmented SERCA2a inhibition by phospholamban (PLN).⁵ PLN is a prominent regulator of Ca²⁺ cycling and a primary mediator of the β-adrenergic effects resulting in enhanced cardiac output.^6,7 In the dephosphorylated state, PLN inhibits SERCA2a and shifts its Ca²⁺ activation towards lower apparent Ca²⁺ affinity. However, upon PKA-mediated phosphorylation, the inhibition on SERCA2a by PLN is relieved and its Ca²⁺ affinity is increased.⁷ Thus, PLN plays a key role in the regulation of Ca²⁺ reuptake by SERCA2a to induce relaxation and decrease diastolic Ca²⁺ levels. It is also a prominent mediator of the β-adrenergic stimulatory responses in the heart.⁷

The important role of PLN in cardiac function has prompted searches for human PLN genetic variants, which may be associated with DCM. Indeed, four mutations have been identified in the coding region of PLN. The first mutation was V49G, which resulted in potent inhibition of the Ca²⁺ affinity of SERCA2a. Cardiac overexpression of the V49G mutant PLN in mouse led to super-inhibition of cardiac contractility and remodelling, which progressed to DCM and early mortality.⁸ The second mutation identified had a stop codon for Leucine 39 (L39X-PLN) with onset of DCM and heart failure during the teenage years.⁹ The third human mutation, R9C, had no effects on SERCA2a activity under basal conditions but appeared to trap PKA, which blocked PKA-mediated phosphorylation of even wild-type PLN. The detrimental effects of such chronic inhibition of SERCA2a activity resulted in DCM.¹⁰ The fourth PLN mutation was a deletion of Arg-14 (Arg14Del).¹¹ It was found that heterozygous carriers exhibited inherited DCM with left-ventricular dilation, contractile dysfunction, variable conduction system disease with episodes of malignant ventricular arrhythmias in some carriers and death by middle age.^12,13 The mechanism underlying the detrimental effects of this human mutation involves super-inhibition of SERCA2a activity, likely mediated through a disturbance in the structure of PLN.¹¹ These findings add to accumulating evidence that myocellular calcium dysregulation caused by mutations in human PLN is sufficiently deleterious to cause DCM and initiate heart failure.

In this study, we report a new mutation (R25C) in the coding region of the human PLN gene (PLN), identified in a pedigree with DCM that also showed prominent ventricular arrhythmia and need for implantable cardiac defibrillators (ICDs).^14,15 Exome sequencing of affected family revealed that they had a R25C-PLN mutation, which was associated with super-inhibition of SERCA2a, depressed myocyte contractile and Ca²⁺-kinetic parameters and increased arrhythmias in cardiomyocytes. The mechanisms underlying the detrimental effects of this mutation include enhanced interaction of R25C-PLN mutant with SERCA2a, leading to super-inhibition of SERCA2a activity and increased Cam kinase II (CaMKII) activity associated with hyper-phosphorylation of Serine 2814 in RyR.

2. Methods

For details regarding methods, refer to the Supplementary material online, Methods.

2.1. Identification of a PLN mutation and generation of adenoviruses

For the human study, written informed consent was obtained and methods were conducted in accordance with the Declaration of Helsinki, and the study was approved by the Institutional Review Boards of Oregon Health and Science University and the University of Miami in FL, USA. Genomic DNA was extracted using standard procedures and PLN exomes were sequenced.^15,16 Adenoviruses expressing green fluorescent protein (Ad.GFP), WT-PLN (Ad.WT), and R25C-PLN (Ad.R25C) were generated.⁹

2.2. Myocyte culture and quantitative immunoblotting

Animals were handled according to the Institutional Animal Care and Use Committee at the University of Cincinnati. Myocytes from adult male rats were isolated,¹⁷ cultured for 2 h and infected with adenoviruses. At 24 h post-infection, we determined PLN levels by western blots and initial rates of SR Ca²⁺ uptake.¹⁸ Data were analysed by non-linear regression using the OriginLab 5.1 program.

2.3. Cardiomyocyte contractility and Ca²⁺ transients

Contractions were obtained at 0.5 Hz and Ca²⁺ kinetics were determined in cells loaded with Fura-2 /AM and excited at 340 and 380 nm.

2.4. HEK 293 cells, immunofluorescence, and immunoprecipitation

HEK 293 cells were transfected with GFP-PLN and SERCA2 constructs,¹⁹ fixed, blocked, and incubated with SERCA2 antibody. Samples were counterstained with Alexa Fluor anti-mouse 568 secondary antibody, mounted with Vectashield medium containing DAPI and analysed by confocal microscopy. In parallel, transfected HEK 293 cells were lysed and processed for immunoprecipitation.

2.5. Measurement of Ca²⁺ sparks, waves, and diastolic SR Ca²⁺ leak

Spontaneous Ca²⁺ sparks and waves were obtained in quiescent cells loaded with Rhod-2 AM. Diastolic SR Ca²⁺ leak was measured using the tetracaine method.²⁰

2.6. Aftercontractions in rat cardiomyocytes

Rat ventricular myocytes were paced at 2 Hz in the presence of 1 µmol/L ISO. After 2 or 3 trains of stimulation, pacing was stopped and spontaneous aftercontractions within 2–5 s were recorded.

2.7. CaMKII activity assay

Cell lysates from infected myocytes were processed for CaMKII activity (Cyclex Kit, MBL International, Woburn, MA, USA).

2.8. Statistics

Results are expressed as mean ± SEM. Comparisons were evaluated by one-way ANOVA and a post-hoc Tukey test. Fisher's exact test was used for calcium sparks, waves, and after contractions. Values of P < 0.05 were considered significant.

3. Results

3.1. Identification of a novel PLN mutation in a familial DCM pedigree

Forty-eight variants, meeting our established criteria for exome sequencing analysis,¹⁵ were identified in two sisters (III.1 and III.3) in our family pedigree (Figure 1). Of these 48, three were in previously identified DCM genes, including a novel heterozygous PLN mutation that resulted in change of arginine into cysteine at position 25, and two missense mutations in Titin (TTN) (A3656S and P7178S). Because it is known that even rare (<0.5%) TTN missense mutations are prevalent in control samples,^16,21 these missense variants were considered unlikely to be the cause of DCM. The PLN mutation had high conservation scores (Phastcons 1; GERP 4.31), and based on other heterozygous PLN missense mutations causing DCM,^8–11 it was hypothesized that this mutation also may alter PLN activity. None of the genes harbouring the remaining 45 variants were associated with known cardiovascular disorders. No additional PLN variants were identified in 16 other families.¹⁵

Figure 1.

Familial DCM Pedigree with a PLN mutation. Squares represent males and circles represent females. Slash denotes deceased. Darkened symbols indicate idiopathic DCM with implantable cardiac defibrillator (ICD) and grey symbols represent a significant cardiovascular abnormality. Open symbols represent negative cardiovascular history. The presence of the PLN mutation is denoted with (+) and the presence of the LMNA mutation¹⁴ with asterisk.

A pedigree with the PLN mutation status is shown in Figure 1. Sanger sequencing confirmed that the PLN mutation was present in the sisters (III.1 and III.3), the proband (III.2) who also carried a previously identified LMNA mutation,¹⁵ as well as in his mother (II.3), who had DCM but was not found to carry the LMNA mutation.¹⁵

3.2. Clinical characteristics

Clinical characteristics of family members with the PLN mutation are provided (Table 1). The proband (III.2) presented at age 47 with sudden cardiac arrest and was diagnosed with non-ischaemic DCM upon resuscitation. An ICD was placed from which he has received multiple appropriate shocks. Approximately 10 years later, he required heart transplantation due to worsening heart failure. Two of his sisters were also found to harbour the R25C-PLN mutation. His older sister (III.1) was first screened clinically at age 51 and was found to have left-ventricular enlargement but preserved systolic function. Despite already receiving an angiotensin receptor blocker (ARB), 2 years later her ventricular function had deteriorated. Non-sustained ventricular tachycardia (NSVT) was found on Holter monitoring, and an ICD was placed prior to the onset of any heart failure. She was treated with full-dose beta blockade and an ARB. Over the course of the following 10 years, two cardioversions were required due to persistent atrial fibrillation. The proband's younger sister (III.3), who also had NSVT and required an ICD, was diagnosed with DCM at age 45 after presenting with signs of heart failure. Their mother (II.3), who had complained of palpitations since age 47, had premature ventricular contractions at age 57. She was diagnosed with DCM at 71 and had multiple PVCs, couplets, and NSVT for years, for which an ICD was placed at age 84 because of progressive pump dysfunction. She progressed to chronic atrial fibrillation and received multiple cardioversions, and died from heart failure at age 90.

Table 1.

Clinical characteristics of family members with the R25C-PLN mutation

Pedigree position	DCM	Age of diagnosis, years	LVEDD, mm (Z-score)	LVEF, %	ECG/Arrhythmia	Comment
II.3	Yes	71	60 (4.1)	35	AF, PVCs, LBBB, ICD	HF
III.1	Yes	53	58 (3.4)	40	AF, 1AVB, VT, ICD	HF
III.2	Yes	47	72 (4.85)	28	AF, RBBB, PM/ICD VT, SCD	Heart transplant at 58 years
III.3	Yes	45	59.6 (3.99)	17	NSVT, PM/ICD	HF

LVEDD, left-ventricular end-diastolic dimension by echocardiography; Z-score, number of standard deviations of the LVEDD above the population mean; EF, left-ventricular ejection fraction; AF, atrial fibrillation; PVCs, premature ventricular contractions; LBBB, left bundle branch block; ICD, implantable cardiac defibrillator; HF, heart failure; 1AVB, first-degree atrioventricular block; VT, ventricular tachycardia; RBBB, right bundle branch block; PM, pacemaker; SCD, sudden cardiac death; NSVT, non-sustained ventricular tachycardia.

3.3. Expression of mutant PLN in adult cardiomyocytes and Ca²⁺ uptake assay

To determine whether the inhibitory effects of PLN on cardiac contractility are modified by the R25C mutation, recombinant adenoviruses encoding GFP, WT-PLN, or R25C-PLN were transduced into adult rat ventricular myocytes. The protein levels of PLN in infected cardiomyocytes were assessed by SDS–PAGE and western blots. Quantitative immunoblotting of cell lysates revealed a similar ratio of PLN pentamers to monomers in mutant PLN (85.8%) cardiomyocytes, compared with WT-PLN (86.6%) and GFP (84.9%) groups (Figure 2A). Furthermore, upon boiling the cardiac homogenates prior to SDS–PAGE, the mutant could be dissociated into monomers, similar to WT-PLN. There were no significant changes in the SERCA2a protein levels in the infected mutant cardiomyocytes (Figure 2A). When the apparent PLN/SERCA2a ratio was calculated, it was found to be a 1.9-fold in R25C myocytes, compared with the GFP group, and this ratio was similar to that for WT-PLN (1.93) (Figure 2B).

Figure 2.

Quantitative immunoblots from infected cardiomyocytes and Ca²⁺ uptake assays. (A) Representative blots of PLN and SERCA. PLNp, pentameric PLN; PLNm, monomeric PLN; (B) PLN protein levels in GFP, WT and R25C cardiomyocytes expressed as relative ratio of PLN/SERCA2a (n = 4 hearts); (C) Effects of wild-type and mutant R25C-PLN on the apparent Ca²⁺ affinity of SERCA2a. After 24-h infection with adenoviruses, cardiomyocytes were homogenized and the initial rates of oxalate-supported SR Ca²⁺ uptake were measured. Data are expressed as per cent of maximal uptake rates in each group (V_max: 99 ± 4 in GFP, 101 ± 3 in WT, and 96 ± 5 in R25C, nmol/mg/min). Inset: The average EC50 values for the three groups (n = 6 hearts). *P < 0.05, vs. GFP; ^†P < 0.05, vs. WT. Values are mean ± SE.

To examine the functional significance of the R25C substitution in PLN and its impact on SERCA2a regulation, the initial rates of ATP-dependent, oxalate-facilitated SR Ca²⁺ uptake were measured in cell lysates from infected cardiomyocytes. There were no significant differences in the maximal rates of SR Ca²⁺ uptake between GFP, WT, or R25C groups. However, the EC50 value for Ca²⁺ dependence of Ca²⁺ uptake was significantly higher in WT-PLN cells (0.40 ± 0.03 µM), compared with the GFP group (0.22 ± 0.02 µM), consistent with our previous findings.²² Interestingly, R25C-PLN further increased the SERCA2a EC50 value to 0.63 ± 0.07 µM (Figure 2C), Thus, the mutant PLN is a super-inhibitor of the affinity of SERCA2a for Ca²⁺ compared with WT-PLN.

3.4. Functional analysis of the R25C-PLN mutant

Overexpression of WT-PLN has been shown to lead to significant depression of cardiomyocyte function.²² To determine whether the super-inhibitory effects of the R25C mutation on SERCA2a activity translated into alterations at the cellular level, the contractile parameters of infected cardiomyocytes were assessed. The resting cell length in mutant PLN infected myocytes was not altered compared with the WT-PLN or GFP groups. However, the amplitude of basal cell contraction (fractional shortening, FS%) was more depressed in myocytes overexpressing R25C mutant PLN (60%) than in myocytes overexpressing WT-PLN (76%), compared with GFP control (100%) (Figure 3A and B). The maximal velocities of cardiomyocyte shortening and re-lengthening were also more depressed in myocytes with mutant PLN (+dL/dt: 52%; −dL/dt: 46%) than in cells with WT-PLN (+dL/dt: 67%; −dL/dt: 66%), compared with GFP controls (100%) (Figure 3C and D).

Figure 3.

Contractile parameters in Ad.GFP, Ad.WT-PLN, and Ad.R25C-PLN infected cardiomyocytes. (A) Representative cell-shortening traces of Ad.GFP, Ad.WT-PLN, and Ad.R25C-PLN cardiomyocytes (cell-length trace represents the percentage of resting cell length); (B) Fractional shortening (FS%); (C) Maximum rates of contraction (+dL/dt); (D) Maximum rates of relengthening (−dL/dt) (20–25 cells were measured per experiment or each heart; n = 4 hearts for GFP, WT-PLN, and R25C-PLN groups). *P < 0.05, vs. GFP; ^†P < 0.05, vs. WT. Values are mean ± SE.

3.5. Overexpression of R25C-PLN significantly suppresses Ca²⁺ transients, delays the rate of Ca²⁺ removal, and elevates intracellular diastolic Ca²⁺

To determine whether the observed alterations in myocyte mechanics reflected similar alterations in Ca²⁺ handling, intracellular Ca²⁺ transients in infected cardiomyocytes were measured by use of the Fura-2/AM fluorescence indicator (2 µM). Our results demonstrate that the amplitude of Ca²⁺ transients was reduced by 55% in mutant-PLN and by 37% in WT-PLN, compared with GFP cardiomyocytes (Figure 4A and B). The time to 50% decay of the Ca²⁺ signal was prolonged by 65% in cardiomyocytes overexpressing mutant and by 32% in cardiomyocytes overexpressing WT-PLN, relative to GFP group (Figure 4C). In addition, intracellular diastolic Ca²⁺ was also analysed and it was found to be increased by 18% in R25C-PLN compared with WT-PLN or GFP cardiomyocytes (Figure 4D). Thus, the R25C-PLN mutant depressed mechanical and Ca²⁺ kinetic parameters and increased diastolic Ca²⁺ levels, consistent with the enhanced inhibition of SERCA2a.

Figure 4.

Ca²⁺ kinetics in Ad.GFP, Ad. WT-PLN, and Ad.R25C-PLN cardiomyocytes. Infected myocytes were incubated with Fura-2/AM for half an hour and Ca²⁺ transients were measured. (A) Representative tracings of Ca²⁺ transients; (B) Ca²⁺ transient amplitude in infected cardiomyocytes; (C) Time to 50% decay (T50) of Ca²⁺ signal; (D) Intracellular diastolic Ca²⁺; (E) Representative tracings of caffeine (10 mM)-induced Ca²⁺ transient; (F) Caffeine-induced Ca²⁺ transient amplitude; (G) Time to 50% decay (T50) of caffeine-induced Ca²⁺ transient peak (20–25 cells were measured per experiment or each heart; n = 4 hearts for GFP, WT-PLN, and R25C-PLN groups). *P < 0.05, vs. GFP; ^†P < 0.05, vs. WT. Values are mean ± SE.

The effects of mutant-PLN on intracellular Ca²⁺ transients prompted further studies on the influence of R25C-PLN on SR Ca²⁺ load. It was observed that the caffeine-induced Ca²⁺ transient peak was reduced by 37% in R25C-PLN cardiomyocytes and by 20% in WT-PLN cardiomyocytes, compared with the GFP group (Figure 4E and F), indicating a super-inhibitory effect of mutant-PLN on SR Ca²⁺ content. However, the time constant for 50% decay of the caffeine-induced Ca²⁺ transient (T₅₀), which mainly reflects Ca²⁺ extrusion by the sodium/calcium exchanger (NCX), was similar between the three groups (Figure 4G).

3.6. Isoproterenol stimulation relieves the inhibitory effects of R25C-PLN

Phospholamban has been postulated to be phosphorylated during β-adrenergic stimulation, resulting in relief of its inhibitory effects on SERCA2a and cardiac function.⁷ To determine whether the super-inhibitory effects of mutant PLN could be also relieved by β-agonists, cardiac myocytes infected with Ad.GFP, Ad.WT-PLN, or Ad.R25C-PLN were stimulated with isoproterenol and their mechanical and Ca²⁺ kinetic parameters were assessed. It is interesting to note that maximal stimulation, obtained at 100 nmol isoproterenol, resulted in complete reversal of the inhibitory effects of wild-type or mutant PLN. Similarly, under maximal isoproterenol stimulation, the inhibition on the amplitude of systolic Ca²⁺ transient and the rate of decay of this signal were fully reversed in cardiomyocytes overexpressing mutant PLN. The maximally stimulated values were similar among the three groups (see Supplementary material online, Figure S1).

3.7. The mutant R25C-PLN exhibits enhanced association to SERCA2 and PKA abrogates this effect

To gain insights into the mechanisms associated with increased inhibition of SERCA2a by R25C-PLN, we co-expressed SERCA2 with either WT or mutant PLN in HEK cells to evaluate their interactions in the absence of endogenous proteins. For this, we generated a GFP-WT-PLN or GFP-R25C-PLN constructs and transiently co-transfected these along with SERCA2 in HEK 293 cells. Initial immunofluorescence studies indicated that, similar to GFP-WT-PLN, the GFP-R25C-PLN mutant exhibited co-localization with SERCA2 in ER (Figure 5A). Subsequent immunoprecipitation studies from co-transfected cells revealed a significant increase in the levels of SERCA2 in the R25C-PLN sample, compared with WT-PLN (Figure 5B and C), indicating enhanced formation of the SERCA2/R25C-PLN protein complex. Interestingly, parallel experiments in cell lysates, that had been previously phosphorylated by PKA, revealed that phosphorylation of R25C-PLN abolishes the increased interaction of this mutant with SERCA2, as similar levels of SERCA2 were observed in the phosphorylated R25C-PLN and WT-PLN protein complexes (Figure 5D and E). These findings are consistent with relief of the mutant PLN super-inhibitory effects on contractility and Ca²⁺-cycling upon Iso-stimulation of cardiomyocytes (see Supplementary material online, Figure S1).

Figure 5.

(A) The R25C-PLN mutant co-localizes with SERCA2 in transfected HEK 293 cells, similar to WT-PLN. Nuclei are stained with DAPI. Scale bar, 5 μm; (B and C) R25C-PLN mutant exhibits enhanced association to SERCA2. Immunoprecipitation assays in HEK 293 cells that co-express GFP-PLN and SERCA2 were performed using GFP antibody. Quantification of SERCA2 levels revealed a significant increase in the SERCA2/R25C-PLN protein complex compared with GFP-WT-PLN (n = 4) values are means ± SE; *P < 0.05, compared with WT-PLN); (D and E) enhanced binding of R25C-PLN with SERCA2a was abolished upon PKA phosphorylation. Immunoprecipitations were performed in lysates from HEK 293 transfected cells that had been previously phosphorylated with PKA. Western blot analysis (D) determined similar levels of SERCA2 in both WT-PLN and R25C-PLN samples and quantitative analysis (E) showed no difference in SERCA2 binding between WT-PLN and R25C-PLN (n = 4).

3.8. Increased Ca²⁺ sparks, waves, diastolic SR Ca²⁺ leak, and stress-induced aftercontractions in R25C-PLN cardiomyocytes

As described earlier, all R25C affected members in the family pedigree developed cardiac arrhythmias (Figure 1 and Table 1). The molecular trigger for arrhythmia is enhanced SR Ca²⁺ leak, evidenced by increases in Ca²⁺ sparks or waves at the cardiomyocyte level.²³ To determine the effect of R25C on SR Ca²⁺ release, we examined Ca²⁺ spark properties in intact quiescent cells. The line-scan and three-dimensional images for Ca²⁺ sparks are presented in Figure 6A. It was observed that spark frequency was two-fold in the mutant PLN cardiomyocytes, compared with WT-PLN or GFP cardiomyocytes (Figure 6B). Next, the frequency of spontaneous Ca²⁺ waves was examined in GFP, WT-PLN, and R25C-PLN myocytes. It was found that Ca²⁺ waves were developed in 40% of R25C cardiomyocytes, compared with 4% of WT-PLN cells and 6% of GFP cells (Figure 6C and D). Thus, R25C-PLN increases the frequency of Ca²⁺ sparks and Ca²⁺ waves in cardiomyocytes.

Figure 6.

Ca²⁺ sparks, waves, diastolic SR Ca²⁺ leak, and stress-induced aftercontractions (Acs) in GFP, WT-PLN, and R25C-PLN cardiomyocytes. (A) Representative line-scan and three-dimensional (3D) images of Ca²⁺ sparks acquired in infected cardiomyocytes; (B) Cumulative data on Ca²⁺ spark frequency; (C) Representative line-scan and 3D images of Ca²⁺ waves acquired in R25C-PLN cardiomyocytes; (D) Percentage of cells showing Ca²⁺ waves (10–15 cells were measured per experiment or each heart; n = 6 hearts for GFP, WT-PLN, and R25C-PLN groups); (E) Representative traces of SR Ca²⁺ leak were obtained from the three groups. Ca²⁺ leak was determined as the tertacaine sensitive drop in diastolic −2 ratio; (F) Comparison of average diastolic SR Ca²⁺ leak; (G) Quantification of leak/SR load relationships in GFP, WT, and R25C myocytes (ratio of twitch Ca²⁺ transient/caffeine-induced Ca²⁺ transient (10–12 cells were measured per experiment or each heart; n = 4 hearts for GFP, WT-PLN and R25C-PLN groups); *P < 0.05, vs. WT and GFP. Values are mean ± SE. (H) Representative traces of Acs; (I) Percentage of the infected cardiomyocytes that developed Acs (10–12 cells were measured per experiment or each heart; n = 6 hearts for GFP, WT-PLN, and R25C-PLN groups). *P < 0.05, vs. GFP and WT. Values are mean ± SE.

Recent studies have suggested that RyR2-mediated Ca²⁺ leak occurs in part as Ca²⁺ sparks, although there is RyR-mediated and Ca²⁺ spark-independent leak.²⁴ Therefore, total diastolic SR Ca²⁺ leak was also measured using the tertacaine protocol²⁵ (Figure 6E and F). The ratio of SR Ca²⁺ leak to SR Ca²⁺ load was significantly larger in R25C cardiomyocytes, compared with GFP and WT cardiomyocytes (Figure 6G), suggesting that R25C-PLN increases the SR Ca²⁺ leak.

Next, we determined the role of R25C-PLN under stress conditions by measuring the frequency of aftercontractions in GFP, WT-PLN, and R25C cardiomyocytes at 2-Hz field stimulation in the presence of 1 µM Isoproterenol. Spontaneous aftercontractions occurred in 74% of R25C cells within 5 s after pacing was stopped, compared with 17% of WT-PLN and 16% of GFP cells (Figure 6H and I). Taken together, these findings suggest that overexpression of R25C-PLN enhances the propensity for spontaneous Ca²⁺ release from the SR, resulting in increased susceptibility to arrhythmia.

3.9. Increased SR Ca²⁺ leak and arrhythmias in R25C myocytes are associated with augmented phosphorylation of RyR2 at serine-2814

We then investigated whether the increased SR Ca²⁺ leak in R25C cardiomyocytes was associated with increase in the phosphorylation of RyR2. Previous studies have shown that enhanced phosphorylation of Ser2808 (PKA site) and Ser2814 (CaMKII site) in RyR2 can regulate RyR2 function.^26,27 Thus, we examined the phosphorylation state of RyR2 in R25C cardiomyocytes, by performing western blots with phospho-specific antibodies against the RyR2 Ser2808 and Ser2814 sites. It was observed that R25C cells had significantly increased phosphorylation levels of RyR at Ser2814, but not at Ser2808 (Figure 7A–C). This prompted us to determine the activity of CaMKII in cell lysates from GFP, WT-PLN, and R25C cardiomyocytes, using a non-radiographic CaMKII ELISA. We found that the activity of CaMKII in R25C myocytes was increased to two-fold compared with WT-PLN or GFP groups (Figure 7D), consistent with the increase in diastolic Ca²⁺ in these cells (Figure 4D). To further confirm this finding, phosphorylation of CaMKII at Thr286 residue, which represents permanent activation of the CaMKII enzyme, was determined in cell lysates from GFP, WT-PLN, and R25C cardiomyocytes using western blots. It was observed that the level of phosphorylation of CaMKII at Thr286 was considerably higher in R25C cells, compared with the WT-PLN or GFP groups (Figure 7E and F). Interestingly, the increased CaM kinase activity did not reflect altered phosphorylation of PLN (see Supplementary material online, Figure S2) in the mutant cells, suggesting that mutant PLN either altered CaMKII binding or increased PLN-phosphatase activity. To further examine the contribution of CaMKII mediated phosphorylation of RyR2 in the R25C-induced increases of Ca²⁺ sparks and waves, KN93, the selective inhibitor of CaMKII was utilized in parallel studies with KN92 as a control. We found that KN93 completely abrogated the increased Ca²⁺ spark frequency (Figure 7G), indicating that the aberrant SR Ca²⁺ release was contributed by the increased CaMKII phosphorylation of RyR. Furthermore, KN93 fully abolished the Ca²⁺ waves, elicited by R25C-PLN (Figure 7H). These data suggest that the aberrant SR Ca²⁺ leak and resultant arrhythmia were associated with increased CaMKII activity and hyper-phosphorylation of RyR at Ser2814.

Figure 7.

Phosphorylation of RyR2 and CaMKII activity. (A) Representative blots of phosphorylation and total levels of RyR2; (B and C) Percentage of phosphorylated Ser2808 (pSer2808) and Ser2814 (pSer2814) RyR2 in infected cardiomyocytes (n = 6 hearts); (D) CaMKII activity in GFP, WT, and R25C cardiomyocytes (n = 7 hearts); (E) Representative blots of phosphorylation and total levels of CaMKII; (F) Percentage of phosphorylated Thr286 (pT286) CaMKII in infected cardiomyocytes (n = 4 hearts); (G and H) Ca²⁺ spark frequency and percentage of Ca²⁺ waves recorded in GFP, WT, and R25C cardiomyocytes in the absence or presence of CaMKII inhibitor KN93 (1 µmol/L), with KN92 (1 µmol/L) used as a control (15–20 cells were measured per experiment or each heart; n = 4 hearts for GFP, WT-PLN, and R25C-PLN groups). *P < 0.05, vs. WT-Basal, ^†P < 0.05, vs. R25C-Basal. Values are mean ± SE.

4. Discussion

In the current study, we identified a novel R25C-PLN mutation in familial DCM, which acts as a super-inhibitor of SERCA2a and results in decreased SR Ca²⁺ content, Ca²⁺ transients, and impaired contractile function. The depressed SR Ca²⁺ resequestration is associated with increased CaMKII activity and hyper-phosphorylation of RyR2 at Ser2814, leading to aberrant Ca²⁺-leak and after-contractions under stress conditions. This is the first evidence that an experiment by nature to alter PLN function results in increased SR Ca²⁺-leak and propensity to arrhythmia, a significant component of the cardiovascular phenotype observed in the carriers of the PLN mutation.

Our findings provide valuable new insights into the established paradigm of DCM gene/phenotype relationships. We^1,28 and others have suggested that the vast majority of DCM cases and families, shown to have genetic cause, appear monophenotypic-a classic ‘pure’ DCM with no atypical cardiovascular or syndromic features. For these families, ventricular or supraventricular arrhythmias have been attributed to a generic progression towards advanced heart failure with its manifold cellular abnormalities, including increased susceptibility to arrhythmia rather than a gene-specific abnormality. Dissecting between the DCM-elicited arrhythmias and those resulting from a specific molecular defect remains challenging, although the exception has been the category of ‘DCM with prominent arrhythmia’,²⁸ assigned predominantly to variants in LMNA, but also observed with variants in SCN5A²⁹ or DES.³⁰ Our current work augments this established gene/phenotype paradigm, as previously reported rare variants in PLN have been associated with predominantly monophenotypic DCM, while the R25C PLN variant leads to promiment arrhythmias, which we suggest may be due in part to novel, specific mechanisms. These PLN-mutation carriers presented with arrhythmia in middle-age and in the setting of DCM, characterized by an ‘adult-onset’, as commonly observed in most of genetic cardiomyopathy. Another recent report with the R14del-PLN variant, a founder mutation in a large number of Dutch individuals, also suggested a propensity to arrhythmia, although a key predictor of arrhythmia was reduced systolic function (<45%).¹³

Previous studies regarding the role of super-inhibitor PLN mutants in mouse models suggested that the depressed SERCA2 activity and contractility were associated with hypertrophic remodelling and ventricular failure over the long term.^31,32 Similar to those studies, our functional data further support the role of PLN mutations in DCM. The mechanisms underlying the super-inhibitory activity of R25C-PLN appear to include increased association or interaction of this mutant with SERCA2a, resulting in increased inhibition of SR Ca²⁺-transport. Indeed, depressed SR Ca²⁺ cycling is a major characteristic of human and experimental heart failure. Accordingly, cardiac overexpression³³ or gene transfer³⁴ of SERCA2a in rat models of heart failure not only significantly improved cardiac contractile function, but increased survival and cardiac metabolism. Similarly, inhibition of PLN activity resulted in enhanced SR Ca²⁺ cycling and cardiac contractility^34,35 as well as attenuated heart-failure progression.³⁶ In addition, RyR inhibition by calstabin improved cardiac function in failing hearts.²³ These studies suggest that disturbed SR Ca²⁺ cycling may serve as a root cause of heart failure and this hypothesis has been well supported by recent studies in human patients.^37,38

Interestingly, the R25C-PLN was also associated with increased Ca²⁺ sparks, waves, and stress-induced after-contractions, which may have contributed to the clinical features of cardiac arrhythmia in R25C carriers at the cellular level. The underlying pathways appear to involve increased inhibition of SERCA2a by mutant PLN, resulting in elevated diastolic Ca²⁺ and activation of CaMKII³⁹ (Figure 4D). Actually, several studies in neuron cells and cardiomyocytes have shown that sustained intracellular Ca²⁺ elevation enhances CaMKII auto-phosphorylation and increases Ca²⁺/calmodulin-independent kinase activity.^40–42 The activation of CaMKII reflected enhanced phosphorylation of RyR2 at Ser2814, leading to increases in SR Ca²⁺ leak and arrhythmia, consistent with previous studies.^42–44 Indeed, inhibition of CaMKII activity by KN93 abrogated the increases in Ca²⁺-sparks and Ca²⁺-waves, elicited by R25C-PLN. Notably, overexpression of WT-PLN in cardiomyocytes, which resulted in a smaller degree of SERCA2a inhibition than R25C-PLN, did not significantly alter diastolic Ca²⁺ levels, CaMKII activity or phosphorylation of RyR2 at Ser2814. These findings indicate that inhibition of SERCA2a beyond a threshold point has detrimental effects through increased diastolic Ca²⁺ levels and CaMKII activation.

The mechanisms associated with the super-inhibitory effects of R25C-PLN are likely mediated through a disturbed conformational structure of PLN. Arg25 is highly conserved across species and this residue supplies a positive charge, which is important for maintaining the hinge region of PLN in the hydrophilic cytosolic environment. Thus, the Arg25 replacement with cysteine, a non-polar amino acid, could destabilize the hinge angle, leading to conformational changes that enhance the association between PLN and SERCA2a.⁴⁵ Interestingly, the super-inhibitory effects of mutant PLN were completely relieved upon isoproterenol stimulation, which also reduced the binding of R25C-PLN with SERCA2 to similar levels as WT-PLN, indicating that this mutation did not alter the ability of PLN to get phosphorylated. Future studies on the three-dimensional structure of mutant PLB and SERCA2a will provide insights into the mechanisms by which R25C-PLN exerts its inhibitory effects and may contribute to design of appropriate molecules to relieve PLB inhibition on SERCA2a and improve function in DCM carriers.

In summary, the current study demonstrates that a newly identified human R25C-PLN mutant is associated with DCM and increased propensity to arrhythmias. Our findings support the view that suppression of Ca²⁺-resequestration by the SR may result in depressed contractility as well as activation of other detrimental pathways, which further exacerbate impaired Ca²⁺-handling, resulting in arrhythmias and deteriorative remodelling. Thus, the mutation reported in this study, together with the previously reported human PLN mutations, point to the paramount importance of genetic and cardiac screening for these rare exome variants that disrupt normal homeostatic mechanisms for calcium cycling in the human heart.

5. Limitations

Other factors, including environmental or other genetic perturbations, may have also contributed to the R25C-PLN human phenotype. However, the clinical, pedigree, and functional data are compelling and congruent with prior reports of PLN mutations causing DCM.

Supplementary material

Supplementary Material is available at Cardiovascular Research online.

Funding

This work was supported by the National Institutes of Health (HL26057 and HL64018 to E.G.K., HL58626 to R.E.H.); and the European Community's Seventh Framework Program (FP7/2007-2013 under grant agreement no. HEALTH-F2-2009-241526, EUTrigTreat to E.G.K.)