OUP user menu

Alterations in Ca2+ cycling proteins and Gαq signaling after left ventricular assist device support in failing human hearts

Yasuchika Takeishi , Thunder Jalili , Brian D. Hoit , Darryl L. Kirkpatrick , Lynne E. Wagoner , William T. Abraham , Richard A. Walsh
DOI: http://dx.doi.org/10.1016/S0008-6363(99)00415-0 883-888 First published online: 1 March 2000


Objective: Left ventricular assist device support mechanically unloads the failing ventricle with resultant improvement in cardiac geometry and function in patients with end-stage heart failure. Activation of the Gαq signaling pathway, including protein kinase C, appears to be involved in the progression of heart failure. Similarly down-regulation of Ca2+ cycling proteins may contribute to contractile depression in this clinical syndrome. Thus we examined whether protein kinase C activation and decreased Ca2+ cycling protein levels could be reversed by left ventricular assist device support. Methods: Left ventricular myocardial specimens were obtained from seven patients during placement of left ventricular assist device and heart transplantation. We examined changes in protein levels of Gαq, phospholipase C β1, regulators of G protein signaling (RGS), sarcoplasmic reticulum Ca2+ ATPase, phospholamban and translocation of protein kinase C isoforms (α, β1, and β2). Results: The paired pre- and post- left ventricular assist device samples revealed that RGS2, a selective inhibitor of Gαq, was decreased (P<0.01), while the status of Gαq, phospholipase C β1, RGS3 and RGS4 were unchanged after left ventricular assist device implantation. Translocation of protein kinase C isoforms remained unchanged. Left ventricular assist device support increased sarcoplasmic reticulum Ca2+ ATPase protein level (P<0.01), while phospholamban abundance was unchanged. Conclusions: We conclude that altered protein expression and stoichiometry of the major cardiomyocyte Ca2+ cycling proteins rather than reduced phospholipase C β1 activation may contribute to improved mechanical function produced by left ventricular assist device support in human heart failure.

  • G-proteins
  • Heart failure
  • Protein kinases
  • Signal transduction
  • Transplantation

Time for primary review 25 days.

1 Introduction

Activation of the Gαq signaling pathway, including downstream protein kinase C (PKC), plays a critical role in the development of cardiac hypertrophy and heart failure [1–4]. Decreased expression of sarcoplasmic reticulum Ca2+ ATPase (SERCA2a), that may occur in part by a PKC-related process [5], also appears to be involved in mechanical dysfunction of failing myocardium [6–8]. Recently, a novel family of regulators of G protein signaling (RGS) has been found to act as GTPase activating proteins that promote hydrolysis of GTP [9]. Among the RGS family members, RGS2 and RGS4 inhibit selectively Gαq function and resultant phospholipase Cβ activation [10,11]. Thus, alterations in RGS expression in the heart may affect G protein-mediated signal transduction pathways leading to cardiac hypertrophy and failure. Although Zhang et al. have recently reported that RGS3 and RGS4 exist in rodent hearts [12], the status of RGS has not been previously examined in the human heart.

The left ventricular assist device (LVAD) has been used as a ‘bridge’ to cardiac transplantation. LVAD mechanical support, by reducing chamber loading conditions, provides clinical improvement in heart failure along with improved cardiac function and geometry [13,14], regression of hypertrophy [15], and reduction of circulating neurohormones [16]. Accordingly, we examined the relative roles of Gαq-RGS-PKC signaling and calcium cycling protein levels in the favorable responses to LVAD support.

2 Methods

2.1 Patient population and study protocol

The study population consisted of seven consecutive patients (all men, mean age 44±17 years, Table 1) with end-stage heart failure who had undergone LVAD implantation (HeartMate, Thermo Cardiosystems) while waiting cardiac transplantation. Seven normal human hearts not suitable for transplantation for technical reasons were used as a control (five men and two women, mean age 45±17 years). Patients were transplanted 29–97 days after LVAD implantation. Tissue samples from the core of the LV apex removed at device implantation were compared with samples from the explanted heart at the time of transplant. Care was taken to derive myocardial samples from the same region. The investigation conformed with the principles outlined in the Declaration of Helsinki.

View this table:
Table 1

Clinical characteristics of subjectsa

No.StatusSexAge (years)DiagnosisE.F. (%)Duration on LVAD (days)DOB/DOPPDEIACEIOthers
1FailingM31DCM2063+++Diuretics, Hydralazine
2FailingM40DCM1660++Ca2+ antagonist
5FailingM39DCM1138+++Ca2+ antagonist
6FailingM62DCM1035+++Digoxin, Amiodarone
7FailingM60ICM1659++Ca2+ antagonist, Hydralazine
10NonfailingF61CVANAClonidine, Probucol
  • a Abbreviations: M, male; F, female; DCM, dilated cardiomyopathy; ICM, ischemic cardiomyopathy; EF, ejection fraction. DOB/DOP, dobutamine/dopamine; PDEI, phosphodiesterase inhibitor; ACEI, angiotensin-converting enzyme inhibitor; CVA, cerebral vascular accident; NA, not available.

2.2 Protein preparation

Myocardial samples were immediately frozen in liquid nitrogen after dissection, and stored at −80°C until use. Fibrotic or adipose tissue, endocardium, epicardium, or great vessels were carefully excised, and the remaining tissue was homogenized with four volumes of ice-cold lysis buffer as reported [2,4,8]. Membrane and cytosolic fractions of detergent-extracted PKC were prepared as previously described [2,4]. Protein concentration was measured by the protein assay (BioRad, Hercules, CA, USA) using bovine serum albumin as a standard.

2.3 Quantitative immunoblotting

Equal amounts of protein extracts for each sample were separated by SDS–PAGE and transferred to nitrocellulose membranes as described [2,4,8]. To ensure equivalent loading and quantitative transfer efficiency of proteins, the nitrocellulose membrane was stained with Ponceau S. After blocking with nonfat dry milk, membranes were incubated with primary antibodies overnight at 4°C in a dilution appropriate to the protein of interest (1:1000 for RGS2, 1:500 for RGS3, and 1:1500 for RGS4). Primary antibodies for Gαq, phospholipase C β1, RGS4, PKCα, β1, and β2 were purchased from Santa Cruz Biotechnology (Santa Cruz, CA, USA). Antibodies for RGS3 and SERCA were gifts from Dr. A. Muslin (Washington University) [12] and Dr. E.G. Kranias (University of Cincinnati) [4,8], respectively. The phospholamban antibody was purchased from Affinity Bioreagents (Golden, CO, USA). The RGS2 antibody was custom made based in the following peptide sequence (EDFKKTKSPQKLSSKARK) by Research Genetics (Huntsville, AL, USA). Membranes were also probed with a primary antibody for calsequestrin (a gift from Dr. R.L. Jones, University of Indiana) as an internal standard. Antibody specificity was verified by attenuation or abolition of immunoreactive bands with corresponding blocking peptide.

Membranes were then incubated with a secondary antibody (horseradish peroxidase-conjugated, KPL laboratories, Gaithesburg, MD, USA) and visualized by enhanced chemiluminescence (Amersham Life Science, Arlington Heights, IL, USA). Antibody labeling was quantified by a computer program (nih) and expressed in relative scan units.

2.4 Statistical analysis

Data are presented as mean±S.D. Protein data among normal, failing pre-LVAD and post-LVAD hearts were compared by analysis of variance followed by the Scheffe test. Hemodynamic data before and after LVAD were analyzed by the Wilcoxon signed-ranks test. Values with P<0.05 were considered as statistically significant.

3 Results

Changes in hemodynamic and echocardiographic data before and after LVAD implantation are summarized in Table 2. Functional parameters showed significant improvement after LVAD support. Fig. 1 shows representative immunoblots of RGS2, calsequestrin, RGS3, SERCA2a, and phospholamban. Group data obtained from seven normal and seven pre/post-LVAD failing hearts are summarized in Table 3. Increased protein abundance of RGS2 in the pre-LVAD failing heart was reduced significantly after LVAD implantation (P<0.01). Calsequestrin levels were unchanged among all groups. SERCA2a protein expression was depressed in the failing heart (P<0.01) as previously reported [6–8]. Following mechanical unloading with LVAD support, SERCA2a levels increased in the failing left ventricle (P<0.01). RGS3 protein expression was higher in the failing heart than in the normal heart (P<0.01), but unchanged after LVAD implantation. Phospholamban protein levels were decreased in the failing heart compared to the normal heart (P<0.01), but remained unchanged after LVAD implantation. The protein levels of Gαq, phospholipase C β1 and RGS4 were unchanged among normal, pre-LVAD, and post-LVAD samples.

Fig. 1

Immunoblot analysis was performed with antibodies specific to RGS2, RGS3, SERCA2a and phospholamban with and without incubation with the corresponding blocking peptide (+Pep). The upper half of RGS2 membrane was probed with anti- calsequestrin antibody. Positions of the molecular weight marker (kD) are indicated on the left. N, Normal heart; Pre, Pre-LVAD failing heart; Post, Post-LVAD failing heart.

View this table:
Table 2

Hemodynamic and echocardiographic dataa

RAP (mmHg)17±412±3*
PCWP (mmHg)27±58±4*
Systolic PAP (mmHg)58±1331±8*
Diastolic PAP (mmHg)28±315±6*
LVEDD (mm)77±650±10*
Cardiac index (l/min/m2)2.9±0.53.6±0.7*
  • a Abbreviations: RAP, right atrial pressure; PCWP, pulmonary capillary wedge pressure; PAP, pulmonary artery pressure; LVEDD, left ventricular end-diastolic dimension.

  • * P<0.05 compared to pre-LVAD.

View this table:
Table 3

Changes in protein abundance after LVAD implantationa

Normal hearts (n=7)Failing hearts (n=7)
Phospholipase C β1166±16154±15162±4
Monomeric form215±12145±31*172± 21*
Pentameric form154±27123±14*111±12*
  • a Data represent scan unit. *P<0.05 and *P<0.01 versus normal hearts; *P<0.01 compared to pre-LVAD.

These alterations in protein expression after LVAD did not correlate with time on LVAD. In addition, the relations between changes in protein abundance and functional parameters such as pulmonary capillary wedge pressure, left ventricular end-diastolic dimension, and cardiac index were not significant.

We have recently reported that PKCα, β1, and β2 isoforms showed translocation in human failing hearts [3]. Thus in the present study, we examined alterations in subcellular distribution of these PKC isoforms after LVAD implantation.Representative immunoblots of PKC isoforms are shown in Fig. 2. As reported previously, immunoreactivities of three PKC isoforms in membrane fractions were increased in failing hearts compared to normal hearts (Table 4). However, mechanical unloading with LVAD support did not alter subcellular distribution of any PKC isoforms.

Fig. 2

Representative immunoblots of PKC isoforms. Immunoreactivity of the membrane fraction was increased in failing hearts compared to normal hearts, but their subcellular distribution was unchanged between pre- and post-LVAD implantation; abbreviations as in Fig. 1.

View this table:
Table 4

Changes in PKC isoforms translocation after LVAD implantation

Normal hearts (n=7)Failing hearts (n=7)
Membrane/cytosol ratio0.78±0.120.86±0.120.95±0.21*
Membrane/cytosol ratio1.09±0.231.04±0.131.05±0.19
Membrane/cytosol ratio1.48±0.251.77±0.19*1.88±0.33*
  • * P<0.05 and *P<0.01 compared to normal hearts.

4 Discussion

The major findings of the present study are: (1) protein levels of Gαq and phospholipase C β1 were similar among normal, failing pre-LVAD, and post-LVAD hearts, (2) translocation of PKCα, β1, and β2 isoforms was unchanged after LVAD implantation, (3) RGS2, RGS3 and RGS4 were expressed in the human heart, and expression levels of RGS2 and RGS3 were increased in failing hearts compared to normal hearts, (4) RGS2 was decreased after LVAD implantation and (5) SERCA2a levels increased after LVAD implantation.

Recently regulators of G protein signaling (RGS) were identified as intracellular GTPase activating proteins (GAP) that accelerate GTP hydrolysis, thus limiting the G protein activation [9–11,17]. It has been reported that RGS1, RGS2, RGS4 and RGS16 have GAP activity toward the α subunit of heterotrimeric G proteins of the Gi and/or Gq. Differences in receptor sensitivity of the RGS proteins have also been demonstrated [18]. The present study is the first to report the expression of RGS2, RGS3 and RGS4 proteins in the human heart. Recently Tamirisa et al. found that phenylephrine- and endothelin-1-mediated induction of the atrial natriuretic factor gene was inhibited in neonatal rat cardiomyocytes that were transfected with RGS4 [19]. It has been also demonstrated that transgenic mice with cardiac specific overexpression of RGS4 showed reduced left ventricular hypertrophy in response to pressure overload [20]. RGS2 and RGS4 are selective inhibitors of Gαq [10,11], and RGS2 is 10-fold more potent than RGS4 in blocking Gαq-directed activation of phospholipase C β1 [11]. Thus increased RGS2 protein levels in failing human hearts observed in the present study may act to inhibit enhanced phospholipase C β1 activation. The increased RGS2 protein expression was reversed after LVAD support in the present study. Despite the reduction in RGS2 protein levels, there was no change in the degree of PKC isoform translocation after LVAD support. Since increased expression of RGS protein may be a counter-regulatory mechanism to inhibit G protein signaling, we might expect that an isolated decrease in the protein abundance of RGS2 would effect an increase in PKC translocation. Since this was not observed, other factor(s) are involved in the steady state translocation of PKC isoforms before and after LVAD placement.

We did not measure changes in PKC activity after LVAD in the present study. Ping et al. [21] reported that ischemic preconditioning causes selective translocation of PKC isoforms without changes in PKC activity. In isolated guinea pig hearts, we have observed that oxidative stress-induced PKC isoforms translocation from the cytosol to the membrane fraction is not accompanied with changes in PKC activity (unpublished data). These data suggest that measurements of total PKC activity are not sufficiently sensitive to detect the involvement of PKC in ischemic preconditioning and oxidative stress-induced intracellular signaling.

Although still somewhat controversial, decreased abundance of cardiomyocyte Ca2+ cycling proteins may also play an important role in the cardiac dysfunction of heart failure [6–8]. We previously reported in pressure-overload induced failing hearts that angiotensin-converting enzyme inhibition prevented down-regulation of SERCA2a levels and improved intracellular Ca2+ handling and resultant isolated heart function [4]. In the present study we demonstrated for the first time that decreased SERCA2a protein expression in the failing human hearts was improved after LVAD support with no concomitant change in phosholamban levels. Phospholamban in its dephosphorylated state inhibits SERCA2a, and the stoichiometry between SERCA2a and phospholamban appears to be one of the major determinants of cardiac contractility [22]. Thus an increase in SERCA2a protein level results in a more favorable stoichiometry and may contribute, at least in part, to improved intracellular Ca2+ handling [23] and resultant mechanical function of the left ventricle [14] following LVAD implantation.


We thank Ms. Nancy Bowling (Indianapolis) for generously providing normal human heart tissue. This study was supported in part by a SCOR in Heart Failure grant (P50 HL52318) from the National Institutes of Health (R.A.W.) and a grant from the Japanese Heart Foundation (Y.T.).


  1. [1]
  2. [2]
  3. [3]
  4. [4]
  5. [5]
  6. [6]
  7. [7]
  8. [8]
  9. [9]
  10. [10]
  11. [11]
  12. [12]
  13. [13]
  14. [14]
  15. [15]
  16. [16]
  17. [17]
  18. [18]
  19. [19]
  20. [20]
  21. [21]
  22. [22]
  23. [23]
View Abstract