Removal of the Cardiac Troponin I N-terminal Extension Improves Cardiac Function in Aged Mice

Abstract

The cardiac troponin I (cTnI) isoform contains a unique N-terminal extension that functions to modulate activation of cardiac myofilaments. During cardiac remodeling restricted proteolysis of cTnI removes this cardiac specific N-terminal modulatory extension to alter myofilament regulation. We have demonstrated expression of the N-terminal-deleted cTnI (cTnI-ND) in the heart decreased the development of the cardiomyopathy like phenotype in a β-adrenergic-deficient transgenic mouse model. To investigate the potential beneficial effects of cTnI-ND on the development of naturally occurring cardiac dysfunction, we measured the hemodynamic and biochemical effects of cTnI-ND transgenic expression in the aged heart. Echocardiographic measurements demonstrate cTnI-ND transgenic mice exhibit increased systolic and diastolic functions at 16 months of age compared with age-matched controls. This improvement likely results from decreased Ca²⁺ sensitivity and increased cross-bridge kinetics as observed in skinned papillary bundles from young transgenic mice prior to the effects of aging. Hearts of cTnI-ND transgenic mice further exhibited decreased β myosin heavy chain expression compared to age matched non-transgenic mice as well as altered cTnI phosphorylation. Finally, we demonstrated cTnI-ND expressed in the heart is not phosphorylated indicating the cTnI N-terminal is necessary for the higher level phosphorylation of cTnI. Taken together, our data suggest the regulated proteolysis of cTnI during cardiac stress to remove the unique cardiac N-terminal extension functions to improve cardiac contractility at the myofilament level and improve overall cardiac function.

Introduction

Troponin I (TnI),² the inhibitory domain of the troponin (Tn) complex, is an essential protein involved in regulation of the actin-myosin interaction responsible for striated muscle contraction. Troponin I is encoded by three muscle genes expressing fast skeletal, slow skeletal, and cardiac TnI (cTnI) isoforms (1), with the cTnI gene the most recently evolved (2). Unlike the fast and slow skeletal isoforms, the cTnI isoform has evolved to contain a unique 32 amino acid N-terminal extension with the function of modulating cardiac contraction. In response to β-adrenergic-induced stimulatory G-protein (G_sα) activation of protein kinase A (PKA), Ser-23/24 of the cTnI N-terminal extension are bis-phosphorylated (3). Phosphorylation of these residues alters cTnI structure diminishing the cTnI N-terminal interaction with troponin C (TnC) to decrease Ca²⁺-sensitive activation of muscle contraction (4, 5) and increase myosin cross-bridge kinetics (6–8). Functionally, this structural change in the modulatory cTnI N-terminal extension is essential to increase the rate of myocardial relaxation and maintain adequate diastolic filling during the increased heart rate of β-adrenergic stimulation to meet increased cardiac demand. Non-physiological truncations removing the cTnI N-terminal extension result in a cTnI molecule that exhibits similar structural and functional alterations to intact cTnI when phosphorylated at Ser-23/24 (9–11), suggesting the removal of the N-terminal region may also be modulatory to cardiac function. In this study, we investigated the effect of a physiological N-terminal-truncated cTnI core molecule on cardiac muscle function in a transgenic mouse model.

The normal heart of all mammals investigated to date, including humans, contains a low abundance of the cardiac TnI N-terminal-deleted (cTnI-ND) molecule incorporated into the myofilament (12). This cTnI-ND molecule results from the specific proteolytic cleavage of 27 N-terminal amino acid residues from the modulatory cTnI cardiac specific extension. Cardiac remodeling induced by simulated microgravity further up-regulates this specific proteolytic cleavage increasing the cTnI-ND amount (12). Previously we demonstrated the removal of the cTnI N-terminal region does not alter cTnI-ND C-terminal structure nor its binding to cardiac troponin T suggesting the cTnI-ND molecule retains its core functionality (13). Transgenic mouse hearts expressing this cTnI-ND core molecule exhibit a normal life span and cardiac morphology in the presence of decreased myofilament activity at submaximal activation levels of Ca²⁺, increased ex vivo myocardial relaxation and improved ventricular filling suggesting a potentially beneficial effect of the cTnI-ND on heart function (14). An increased cardiac abundance of the cTnI-ND molecule has also been observed in the G_sα-deficient mouse model (G_sα-DF) of heart failure (15), as well as following disruption of PKA localization (16). Recently we demonstrated overexpression of the cTnI-ND molecule in the G_sα-DF Tg mouse model largely prevented the contractile dysfunction in this β-adrenergic-deficient heart failure model (15). These findings indicate that specific proteolytic production of the cTnI-ND molecule during cardiac remodeling by removing the cTnI N-terminal extension produces a functional core cTnI molecule to modulate cardiac function.

It is established that human cardiac function decreases over time with aging. Studies have also demonstrated that aged mice exhibit depressed in vivo cardiac function (17–19). Our objective was to investigate the potential beneficial hemodynamic, mechanical, and biochemical effects of this specific cTnI proteolytic modification on the development of physiological cardiac dysfunction in aged animals. Our results demonstrate young transgenic mice expressing the cTnI-ND molecule (cTnI-ND Tg) exhibit decreased Ca²⁺ sensitivity and increased cross-bridge cycling in detergent extracted cardiac fiber bundles prior to the effects of aging. Furthermore, cTnI-ND expression improved in vivo cardiac relaxation in aged animals as assessed by echocardiography. These data provide novel insights into the in vivo effects of regulated cTnI proteolysis as an adaptive modulator of heart function during cardiac remodeling resulting from physiologically relevant cardiac stress.

EXPERIMENTAL PROCEDURES

Animals

Transgenic mice overexpressing the cTnI-ND molecule have been previously described (14). Animal care and use was performed in accordance with the guidelines of the Institutional Animal Care and Use Committee at the University of Illinois at Chicago.

SDS-PAGE and Western Blot Analysis

Expression of cTnI and cTnI-ND in cardiac muscle was determined by SDS-PAGE separation of ventricular Triton X-100 permeabilized myofibril (13) transferred to nitrocellulose or polyvinylidene difluoride membrane and detected by Western blot using the monoclonal antibody C5 (Fitzgerald) as previously described (20). Resultant blots were stripped and re-probed with an anti-actin antibody (AC-40, Sigma) to determine equal loading. Resultant films were scanned and quantified using ImageQuant TL (GE). Intact cTnI Ser-23/24 PKA phosphorylation was determined by Western blot as above with the rabbit phosphospecific troponin I (cardiac) (S23/24) antibody (Cell Signaling) and an anti-rabbit horseradish peroxidase-linked secondary antibody (GE). Following detection the resultant blot was striped and re-probed for loading normalization to total intact cTnI with mouse C5 primary antibody and anti-mouse alkaline phosphatase-linked secondary antibody (Sigma) by nitro blue tetraxolium/5-bromo-4-chloro-3-indolyl-phosphatase development. This combination of antibodies and differential development methods is critical to avoid carryover of the pTnI Ser-23/24 signal into the total TnI C5 Western. Cardiac myosin heavy chain isoform expression was determined by SDS-PAGE separation of whole ventricle homogenates on 18 × 18 cm gels as previously described. Resultant gels were stained with Gel Code (Pierce) and scanned for quantification using ImageQuant TL (GE) or transferred to nitrocellulose membrane for MHC identification by Western blot using the monoclonal antibody FA2 (21).

Echocardiographic Measurements

Transthoracic two-dimensional-targeted M-mode and pulsed Doppler echocardiography were performed with a 15-MHz linear array transducer (Acuson Sequoia C256 system) as previously described (22). Briefly, mice were anesthetized with 0.5–1.5% isoflurane in 100% oxygen, and body temperature was monitored by rectal thermometer and maintained at 36–37 °C with a heating pad. The transducer placed on a layer of acoustic coupling gel applied to the left hemithorax and mice imaged in a shallow left lateral decubitus position. M-mode images of the left ventricle were obtained from the parasternal short axis view at the level of the papillary muscle. Interventricular septal and left ventricular posterior wall thicknesses and left ventricular internal dimensions at the end of diastole and systole were measured by the American Society of Echocardiography leading edge method on the M-mode tracings. Fractional shortening of the left ventricle was calculated from digital images as: left ventricular fractional shortening percentage (LV FS%) = (LVIDd-LVISd)/LVIDd × 100, where LVIDd is the internal diastolic dimension of the left ventricle, and LVISd is the internal systolic dimension of the left ventricle. The mean velocity of circumferential fiber shortening (Vcf) was calculated as Vcf = FS/ET, where ET is the ejection time through the aortic valve.

Diastolic transmitral inflow recordings were acquired from apical four-chamber views using 7 MHz pulsed Doppler echocardiography as previously described (22). Briefly, The probe was positioned substernally at the xyphoid and the Doppler range gate depth set to 4 mm to obtain optimal signals from the left ventricular inflow and outflow tracts. The sample volume was positioned along the long axis in the middle of the mitral ring at the tips of the opened cusps of the mitral valve to determine Left ventricular isovolumic relaxation time (IVRT), the time from the aortic valve closure to the mitral valve opening. Stroke volume (SV) was measured from pulsed Doppler and cardiac output (CO) calculated as SV multiplied by heart rate. M-mode and Doppler tracings were conducted with a paper speed of 200 mm/s.

Simultaneous Measurement of Isometric Tension and ATPase Activity

The simultaneous measurement of steady state isometric tension and ATPase activity in left ventricular detergent extracted papillary bundles from 4-month-old mice over a range of free Ca²⁺ was conducted as previously described (20, 23). Briefly, sarcomere length was set to 2.2 μm by laser diffraction. Papillary bundles were activated over a range of free [Ca²⁺] to measure steady-state isometric tension and ATPase activity. Measurement of ATP hydrolysis was stoichiometrically coupled to NADH consumption measured by UV absorption (340 nm) in a small chamber (about 25 μl). Only bundles that maintained greater than 80% maximal tension throughout the experimental protocol were included for analysis.

Myofilament Enrichment and Sample Preparation

Subcellular fractionation of the cardiac myofilament fraction was conducted as previously described (24) with slight modification. Ventricular tissue was homogenized in relax buffer (containing in mm; 75 KCl, 2 MgCl₂, 2 EDTA, 1 NaN₃, 10 imidazole, pH 7.2) containing 1% Triton X-100, phosphatase inhibitor mixture I (Calbiochem) and protease inhibitor mixture (Sigma-Aldrich). Resultant homogenates were centrifuged at 18,000 × g for 10 min at a temperature of 4 °C and the supernatant fraction removed. The pellet was then washed two times in relax buffer with inhibitors but without Triton X-100. Following washes the pellet was extracted in UTC buffer (8 m urea, 2 m thiourea, 4% CHAPS) with periodic vortexing for 1 h, sonicated with 1 s bursts 3 times at the lowest power and clarified by centrifugation (10 min, 18,000 × g at 4 °C). The resultant soluble fraction was frozen at −80 °C until analysis.

Two-dimensional Differential in Gel Electrophoresis Analysis

Muscle tissue from NTg and cTnI-ND hearts were minimally labeled with fluorescent cyanine dyes (GE Healthcare) similar to that previously described (25, 26) with all procedures protected from exposure to light. Ventricular myofilament-enriched fractions were labeled with either Cy2, Cy3, or Cy5 at 10 pmol/50 μg protein and incubated on ice for 3 h when the reactions were quenched with 10 nmol/liter lysine. In separate experiments labels were reversed to ensure that analysis was not affected by dye affinity or sensitivity. Analysis of total ventricular myofilament proteins was carried out essentially as previously described (25). Briefly, 40 μg of 3 differentially labeled Ntg and cTnI-ND samples were mixed in UTC buffer containing 0.5% (v/v) Immobilized pH gradient (IPG) 3–11 NL buffer (GE Healthcare), 1% (v/v) Destreak (GE Healthcare), and 1% (w/v) dithiothreitol and actively rehydrated for 12 h at 50 mv and 20 °C on 24 cm nonlinear 3–11 IPG strips (GE Healthcare). Following rehydration proteins were focused by a linear method (250 V for 15 min, 10,000 V for 3 h, and 10,000 V for 40,000 V/h). At this time strips were removed and cut into 3 equal sections. The acidic and basic sections were then processed for the second dimension whereas the middle section containing MyBPC was returned to a 7.5-cm focusing tray and further focused at 6,000 V for 45 min. Following focusing strips were reduced by incubation in 6 m urea, 30% (v/v) glycerol, 7% (w/v) SDS, 1% (w/v) dithiothreitol for 10 min and then blocked in 6 m urea, 30% v/v glycerol, 7% (w/v) SDS, 2.5% (w/v) iodoacetamide for 10 min prior to secondary separation by SDS-PAGE electrophoresis on Laemmli 12% (200:1) gels. Analysis of cTnI was conducted by similar methods except 40 μg of each Cy labeled ventricular myofilament fraction was mixed with UTC buffer containing 0.5% (v/v) immobilized pH gradient (IPG) 7–11 buffer (GE Healthcare), 1% (v/v) Destreak (GE Healthcare), 4 mm tributylphosphine (Sigma), and 1.5% (w/v) dithiothreitol and focused on 18 cm 7–11 IPG strips (GE Healthcare). Immediately following focusing a paper wick saturated with 5% (w/v) dithiothreitol was placed under the negative end of the strip on the electrode and the strip further focused at 10,000 V for 5 min. The strip was then equilibrated and subjected to second dimension separation on Laemmli 12% (29:1) gels. Resultant gels were then imaged using a Typhoon 9410 (GE Healthcare) with Cy2 monitored with a 530-nm BP 40 filter excited with a 488-nm laser; Cy3 with a 580-nm BP 30 filter and 532-nm laser excitation and Cy5 with a 670-nm BP 30 filter and 630-nm laser excitation at 100-μm resolution and the photomultiplier tube set to obtain pixel intensity near 90,000. In a second experiment, NTg and Tg ventricular myofilament fractions labeled with either Cy2 or Cy5 were fractionated for cTnI analysis as above and either stained with ProQ Diamond according to the manufacturer's instructions (Molecular Probes) to identify phosphorylated protein spots or imaged, transferred to nitrocellulose, imaged again, and subjected to Western blot using a monoclonal anti-cTnI antibody (C5, Fitzgerald) to identify cTnI protein spots. Two-dimensional DIGE data were analyzed with DeCyder software (GE Healthcare). Protein spot changes (in percent) were expressed as means ± S.E. and changes with p values of <0.05 considered statistically significant.

RESULTS

cTnI-ND Expression Preserves Cardiac Contractile Function in Aged Animals

Based upon our recent finding that expression of the cTnI-ND molecule in the heart compensates for the G_sα-DF induced heart dysfunction (15), we sought to investigate the in vivo functional effects of cTnI-ND expression in the heart to affect physiological aging induced cardiac depression. Representative Western blots using the cTnI specific mAb C5 on ventricular myofilament preparations from 4- and 16-month-old cTnI-ND Tg mice demonstrates cTnI-ND Tg mice express similar levels of cTnI-ND protein at both ages (Fig. 1C). At 16 months of age neither body weight nor heart rate were significantly different between cTnI-ND and NTg mice (Table 1). Doppler echocardiographic analysis demonstrates that 16-month-old cTnI-ND Tg mice exhibit improved systolic function (increased left ventricular fractional shortening and circumferential fiber shortening), trended toward increased cardiac output (p = 0.06) and demonstrated improved diastolic function (decreased isovolumetric relaxation time) compared to similar aged NTg mice (Table 1).

FIGURE 1.

Transgenic mouse hearts express a high level of cTnI-ND. Structural maps of fast skeletal muscle TnI, slow skeletal muscle TnI, cTnI, and cTnI-ND are aligned with regions for the binding of TnC, TnT, and actin demonstrating the protein region affected by cTnI-ND (A). Representative gels of ventricular myofibrils from NTg and cTnI-ND Tg samples that were separated by SDS-PAGE and stained with Coomassie (B) or transferred to polyvinylidene difluoride for Western blot with the cTnI-specific monoclonal antibody C5 (C). Gels demonstrate that both 4- and 16-month-old cTnI-ND Tg mice express similar high amounts of the cTnI-ND at similar loading as determined by subsequent Western with the anti-actin monoclonal antibody AC-40.

TABLE 1.

Cardiac function of aged NTg and cTnI-ND Tg mice

Echocardiography measurement demonstrates 16-month-old cTnI-ND transgenic mice exhibit improved systolic and diastolic function compared to NTg controls. n, number of animals; LV, left ventricle; IDd, internal diastolic dimension; ISd, internal systolic dimension; PWD, posterior wall in diastole; FS, fractional shortening; Vcf, circumferential fiber shortening; IVRT, isovolumic relaxation time. Values are average ± S.D.

General parameters	NTg (n = 3)	cTnI-ND (n = 4)
Body weight (g)	32.9 ± 3.1	28.3 ± 2.5
Heart rate (bpm)	484 ± 25	532 ± 47
IDd (mm)	3.61 ± 0.25	3.43 ± 0.04
ISd (mm)	2.49 ± 0.43	1.76 ± 0.13a
FS (%)	31.3 ± 7.5	48.8 ± 4.0a
Vcf (circ/sec)	7.51 ± 2.16	11.97 ± 1.04a
Stroke volume (ml)	0.020 ± 0.002	0.036 ± 0.013
Cardiac output (ml/min)	9.87 ± 1.43	19.49 ± 6.73b
IVRT (msec)	23.0 ± 3.4	13.5 ± 3.8a

^a p < 0.05 vs. NTg.

^b p = 0.06 vs. NTg.

cTnI-ND Tg Mouse Hearts Exhibit Altered Sarcomeric Mechanics

To establish the causal effect of the cTnI-ND structural change to modulate contractile function prior to the compensatory effects of aging we investigated the effect of cTnI-ND expression on muscle mechanics from young mice. Calcium-dependent development of force and ATPase activity we simultaneously measured in detergent-extracted papillary fiber bundles from cTnI-ND Tg and NTg mouse hearts at 4 months of age. As illustrated in Fig. 2, fiber bundles isolated from cTnI-ND Tg hearts exhibit a significantly decreased Ca²⁺ sensitivity of force development compared with fibers isolated from NTg hearts (pCa₅₀: cTnI-ND = 4.95 ± 0.02; NTg = 5.68 ± 0.07, p < 0.01) in the absence of altered maximal force (Fig. 2 (inset) F_max: cTnI-ND = 24.19 ± 2.7; NTg = 25.97 ± 3.0, p > 0.05). Furthermore, tension cost, a parameter reflecting the rate of cross-bridge detachment (27), was significantly increased in fibers from cTnI-ND Tg mice (Fig. 3; tension cost, cTnI-ND = 11.39 ± 0.49; NTg = 9.61 ± 0.41, p < 0.05) without a change in the number of cycling myosin cross-bridges as stiffness was not altered (stiffness: cTnI-ND 1.90 ± 0.11; NTg = 1.85 ± 0.08, p > 0.05). The average parameters are summarized in Table 2. These findings demonstrate the structural effects of the cTnI-ND to modulate myofilament function in the young animal prior to the cardiac stress effects of aging.

FIGURE 2.

Detergent-extracted cardiac bundles from cTnI-ND Tg mouse hearts exhibit decreased Ca²⁺ sensitivity. Cardiac fiber bundles were isolated from cTnI-ND and NTg mouse hearts, detergent extracted with Triton X-100 and the Ca²⁺-activated development of steady-state tension and ATPase activity simultaneously measured. Tension versus Ca²⁺ plots demonstrate cTnI-ND Tg fibers exhibit significantly decreased Ca²⁺ sensitivity compared with NTg fibers without an effect on maximal stress development (inset). NTg, solid black line and black bar, n = 5; cTnI-ND, dashed gray line and hatched bar, n = 5.

FIGURE 3.

Detergent-extracted cardiac bundles from cTnI-ND Tg mouse hearts exhibit an altered rate of cross-bridge cycling. Simultaneous tension and ATPase activity were measured in detergent extracted cardiac fiber bundles isolated from cTnI-ND and NTg mouse hearts. The relation between ATPase activity and tension demonstrates that compared with controls, cTnI-ND Tg fibers exhibit increased ATP hydrolysis and tension cost (inset). NTg, solid black line, and bar, n = 5; cTnI-ND, gray dashed line and hatched bar, n = 5. *, p < 0.05 versus NTg.

TABLE 2.

Mechanical characteristics of cTnI-ND Tg mice

The Ca²⁺-activated tension and myofibrillar ATPase measurements of skinned fiber bundles from NTg and cTnI-ND Tg hearts. Values are mean ± S.E. pCa₅₀; inverse log of the [Ca²⁺] in μmol/liter required to develop 50% maximal tension; Fmax, maximally developed tension in mN/mm²; Hill, the slope of the tension-Ca²⁺ plot; Stiffness in arbitrary units, high frequency muscle stiffness (500 Hz); parameter is proportional to the number of attached cross-bridges; TC, tension-cost, slope of the ATPase vs. tension plot; n, the number of fibers (one fiber per animal) in each group.

	pCa50	Fmax	Hill	Stiffness	TC	n
NTg	5.68 ± 0.07	26.0 ± 3.0	2.82 ± 0.24	1.85 ± 0.08	9.88 ± 0.45	5
cTnI-ND	5.42 ± 0.02a	21.2 ± 2.7	3.88 ± 0.36	1.90 ± 0.11	11.39 ± 0.49a	5

^a p < 0.05 vs. NTg.

cTnI-ND Tg Mouse Hearts Exhibit Altered Myofilament Protein Modification

Previously we demonstrated cardiac specific Tg expression of the slow skeletal isoform of cTnI, also lacking the unique cardiac N-terminal extension, decreased the age-dependent increase of β myosin heavy chain (β-MHC) expression in rodents. To determine if cTnI-ND Tg hearts also exhibit altered β-MHC we separated α-MHC and β-MHC isoforms by SDS-PAGE and quantified MHC expression following Coomassie staining. The results in Fig. 4 demonstrate cardiac myofilament homogenates from 4-month-old NTg and cTnI-ND Tg mice exhibit minimal expression of the β-MHC isoform (Fig. 4A). At 16 months of age, NTg mice exhibit increased expression of β-MHC, however β-MHC expression in the cTnI-ND Tg heart was similar to that expressed at 4 months of age and significantly decreased compared with NTg (Fig. 4B). These results demonstrate expression of the cTnI-ND protein in the heart prior to aging diminishes the age-dependent expression of β-MHC observed in the NTg mouse.

FIGURE 4.

Aged cTnI-ND Tg mouse hearts exhibit decreased β myosin heavy chain isoform expression. Whole ventricular homogenates were isolated from NTg and cTnI-ND Tg hearts at 4 and 16 months of age and subjected to SDS-PAGE to determine MHC isoform expression. Representative gel demonstrating that at 4 months of age NTg and cTnI-ND hearts expressed only α myosin heavy chain (α-MHC) (A); however at 16 months of age NTg hearts exhibited increased expression of the β myosin heavy chain (β-MHC) isoform while the β-MHC isoform remained minimal in cTnI-ND Tg hearts (B). NTg, solid bars; cTnI-ND, hatched bars. *, p < 0.05 versus NTg.

To further investigate the molecular changes responsible for altered contractile function in Tg mice expressing the cTnI-ND, we quantified myofilament protein post-translational modification status by two-dimensional isoelectric focusing (two-dimensional IEF) SDS-PAGE. Ventricular myofilament post-translational status was evaluated by two-dimensional difference in gel electrophoresis (two-dimensional DIGE). Myofilament preparations from NTg and cTnI-ND Tg hearts were differentially minimally labeled with Cydye, the independent samples pooled, separated by two-dimensional IEF SDS-PAGE over a non-linear pH range of 3–11 and the myofilament proteins imaged for each Cydye (NTg or cTnI-ND Tg) independently. Fig. 5 demonstrates representative two-dimensional DIGE images of 4- and 16-month-old NTg and cTnI-ND Tg myofilament preparations following merged overlay of Cy5-labeled NTg (red) and Cy2-labeled cTnI-ND Tg (green) samples. As result of the two-dimensional DIGE merge, spots containing a higher amount of protein from Cy5 NTg are viewed as red, those containing more Cy2 cTnI-ND Tg are green, and protein spots of similar amount are viewed as yellow. Spot quantification of the individual grayscale Cydye myofilament two-dimensional DIGE scans at both 4 and 16 months of age demonstrated no significant difference in the spot amounts of actin, troponin T, myosin light chain 1, myosin light chain 2, or desmin between cTnI-ND Tg and NTg hearts. Tropomyosin and cTnI were not well resolved in this protocol. Unlike these proteins, the spot profile of myosin-binding protein C (MyBP-C) distribution was significantly altered in TnI-ND Tg compared with NTg at 16 but not 4 months of age; however total MyBP-C phosphorylation as assessed by one-dimensional SDS-PAGE gels stained with ProQ Diamond was not significantly different (data not shown). These findings demonstrate cTnI-ND expression did not cause significant changes in the majority of the major myofilament protein modification status.

FIGURE 5.

Myofilament post-translational modification in cTnI-ND Tg mouse hearts. Ventricular myofilament preparations from NTg and cTnI-ND Tg hearts were differentially labeled with Cydye, the independent samples pooled, separated by two-dimensional isoelectric focusing differential in gel electrophoresis (2D DIGE) on non-linear 3–11 IPG (immobilized pH gradient) strips, cut for molecular weight separation on mini gels, the myofilament proteins independently imaged for each Cydye (NTg or cTnI-ND Tg) and the resultant images quantified. Representative merged two-dimensional DIGE images of 4- and 16-month-old NTg (Cy2; red) and cTnI-ND Tg (Cy5; green) myofilament preparations demonstrate no difference in actin, troponin T (TnT), myosin light chain 1 (MLC-1), myosin light chain 2 (MLC-2), or desmin, whereas the myosin-binding protein C (MyBP-C) and cTnI spot profile was dramatically different between NTg and cTnI-ND Tg hearts. As result of the two-dimensional DIGE merge, protein spots containing a higher amount of protein from Cy5 NTg are viewed as red, those containing more Cy2 cTnI-ND Tg are green, and protein spots of similar amount are viewed as yellow. NTg, red, 4-month-old n = 3, 16-month-old n = 3; cTnI-ND, green, 4-month-old n = 3, 16-month-old, n = 4.

Our Tg model specifically expressed a modified cTnI such that the N terminus containing the cTnI PKA phosphorylation sites Ser-23/24 was removed. Because two-dimensional isoelectric focusing in the 3–11 pH range did not adequately resolve the cTnI species, we employed two-dimensional DIGE over a non-linear pH range of 7–11 to investigate the effect of cTnI-ND expression on cTnI post-translational modification. Fig. 6A shows a representative two-dimensional DIGE image with merged overlay of Cy5-labeled NTg (green) and Cy2-labeled cTnI-ND Tg (red) samples and a representative close-up view of the cTnI area (Fig. 6B). Following imaging, the same gel was transferred to nitrocellulose and probed with a specific monoclonal antibody against a C-terminal cTnI epitope that recognizes both intact and cTnI-ND with similar affinity. Cardiac TnI was resolved into 8 spots of two molecular weights (Fig. 6C); 7 intact cTnI spots migrating at about 24 kDa (28) and a single lower molecular mass cTnI spot corresponding to the cTnI-ND molecule as it was only identified in sufficient quantity in the cTnI-ND Tg heart. ProQ Diamond phosphoprotein staining identified the 6 most acidic intact cTnI spots as containing phosphorylated amino acid residues (P1–P6) whereas the most basic spot was not phosphorylated (U) (Fig. 6D). Importantly, Cydye visualization (Fig. 6B) and Western blot (Fig. 6C) only identified a single lower molecular weight cTnI species corresponding to the cTnI-ND that was not detected by ProQ Diamond phosphoprotein staining (Fig. 6D). These findings demonstrate the cTnI-ND protein expressed in the heart of the Tg mice is not phosphorylated at any of the remaining potential cTnI phosphorylation sites.

FIGURE 6.

cTnI-ND expressed in the Tg heart is not phosphorylated. Two-dimensional DIGE analysis was conducted on myofilament preparations from 4-month-old NTg and cTnI-ND Tg hearts to evaluate cTnI post-translational modification. A, representative two-dimensional DIGE-merged images of Cy2-labeled NTg (green) and Cy5-labeled cTnI-ND Tg (red) myofilaments separated on non-linear 7–11 IPG strips. B, merged two-dimensional DIGE magnified view of NTg and cTnI-ND Tg cTnI. C, following transfer to nitrocellulose and Western blot with a monoclonal antibody against the C terminus of cTnI 7 intact cTnI spots of about 24 kDa and 1 cTnI-ND spot were identified. D, ProQ phosphoprotein staining of an identical gel demonstrates the 6 most acidic intact cTnI spots are phosphorylated (P1 to P7) whereas the most basic spot was not phosphorylated (U). Interestingly, the lower molecular weight cTnI-ND species migrated as a single spot and was not phosphorylated. NTg, green; cTnI-ND, red.

As the cTnI-ND protein is not phosphorylated we sought to determine if expression of the cTnI-ND in the Tg heart altered phosphorylation of the remaining intact cTnI. Intact total cTnI phosphorylation (pTnI) by one-dimensional ProQ Diamond phosphoprotein staining demonstrates cTnI-ND Tg hearts contain a significantly decreased amount of pTnI at both 4 and 16 months of age compared with NTg (Fig. 7). (Intact cTnI phosphorylation in A.U., NTg 4 months old = 483 ± 72.7; TnI-ND Tg 4 months old = 142 ± 34.5; NTG 16 months old = 321 ± 124.3; TnI-ND Tg = 91 ± 63.2. NTg versus TnI-ND Tg p < 0.05, n = 6.) Next, we sought to determine if the cTnI-ND Tg decrease in total pTnI resulted from altered cTnI Ser-23/24 phosphorylation by Western blot using a cTnI Ser-23/24 phosphospecific antibody. The representative Western blot in Fig. 8 shows pTnI at Ser-23/24 was also significantly decreased in cTnI-ND hearts at both 4 and 16 months of age compared with NTg at a similar intact cTnI loading (cTnI-ND 4 months old = 31.3% ± 23.3% of NTg; cTnI-ND 16 months old = 41.9% ± 30.6% of NTg, p < 0.05, n = 4). These findings suggest Ser-23/24 phosphorylation of cTnI largely accounts for the difference in total pTnI observed by ProQ Diamond staining. To further investigate the non-phosphorylated cTnI we quantified non-phosphorylated cTnI in NTg and Tg heart fractionated by two-dimensional DIGE over a non-linear pH range of 7–11. Quantification of the cTnI non-phosphorylated species (spot U) amount as a percent of the total cTnI (spots U and P1-P6) in 4-month-old hearts confirms the above findings that the non-phosphorylated cTnI in cTnI-ND Tg was significantly increased compared with cTnI-ND hearts (Fig. 9A, WT = 8.4 ± 1.7%; TnI-ND Tg = 19.2 ± 4.6%. p < 0.05, n = 3). Interestingly, at 16 months of age the percent of cTnI that was not phosphorylated in the NTg was increased such that it was no longer different from the cTnI-ND Tg (Fig. 9B, WT = 13.2 ± 4.6%; TnI-ND Tg = 18.3 ± 2.9%. p > 0.05, n = 3). This finding of increased non-phosphorylated cTnI in the 16-month-old Ntg mice is consistent with the trend of decreased total cTnI phosphorylation as detected by ProQ Diamond staining in NTg, although it was not significant. These data demonstrate the remaining intact cTnI in the cTnI-ND Tg heart is clearly decreased in total as well as Ser-23/24 phosphorylation, however the role of other cTnI phosphorylation sites are unknown.

FIGURE 7.

Total intact cTnI is significantly decreased in the heart of cTnI-ND Tg mice. Ventricular myofilament homogenates were separated by one-dimensional SDS-PAGE and total intact cTnI phosphorylation determined by ProQ Diamond staining. Representative gels demonstrating total cTnI phosphorylation was significant decreased at both 4 (A) and 16 (B) months of age. Total intact cTnI phosphorylation by ProQ Diamond is normalized to intact cTnI loading as determined by Coomassie staining of the same gel. NTg, solid dark bars; cTnI-ND, hatched bars. *, p < 0.05 versus NTg, n = 6.

FIGURE 8.

cTnI-ND Tg exhibit decreased intact cTnI Ser-23/24 phosphorylation. Representative Western blots of ventricular tissue with the rabbit anti-phospho cTnI Ser-23/24 antibody that specifically recognizes cTnI only when phosphorylated at Ser-23/24 in conjunction with the anti-rabbit horseradish peroxidase linked secondary antibody demonstrates cTnI-ND Tg hearts exhibit significantly decreased cTnI PKA phosphorylation at both 4 (A) and 16 (B) months of age. After stripping, the blot was re-probed for total intact cTnI with the mouse cTnI primary antibody C5 and anti-mouse alkaline phosphatase-linked secondary antibody followed by nitro blue tetraxolium/5-bromo-4-chloro-3-indolyl-phosphatase development to normalize for total intact cTnI. This combination of antibodies and differential development methods is critical to avoid carryover of the pTnI Ser-23/24 signal into the total TnI blot. *, p < 0.05 versus NTg, n = 4.

FIGURE 9.

The age-dependent increase in non-phosphorylated intact cTnI is dependent upon cTnI-ND expression. To determine the amount of non-phosphorylated intact cTnI in NTG and cTnI-ND Tg hearts, two-dimensional DIGE analysis was conducted on myofilament preparations from 4 month (A) and 16 month (B) old animals. Representative two-dimensional DIGE images of NTg (green) and cTnI-ND Tg (red) cTnI separated on non-linear 7–11 IPG strips. Single channel grayscale images of NTg and cTnI-ND Tg cTnI were quantified by PDQuest and the amount of non-phosphorylated cTnI calculated as a percentage of the total cTnI spots. Results demonstrate 4-month-old NTg exhibit significantly decreased percentage of non-phosphorylated total cTnI compared with that of the cTnI-ND Tg while the amount of NTg non-phosphorylated cTnI was increased at 16 months of age such that it was no longer different from that of the cTnI-ND Tg heart. NTg, Cy2 labeled in green, 4-month-old n = 3, 16-month-old n = 3; Tg = cTnI-ND, Cy 5 labeled in red, 4-month-old n = 3, 16-month-old n = 4. *, p < 0.05 versus NTg.

DISCUSSION

Our data demonstrate the significance of the cTnI N-terminal extension truncation as a regulatory mechanism to modulate cardiac contraction and maintain cardiac function during stress. Many proteolysis degradation products of proteins are nonspecific, render the protein non-functional and are destructive to the function of the protein. Previously we reported the specific proteolysis of cTnI to selectively remove the N-terminal extension as a regulated modification (12, 16) to improve cardiac function in both a normal (14) and a disease mouse model (15). These findings led to the theory that the regulated proteolytic removal of the cTnI N terminus is not detrimental to the function of cTnI but rather serves as a regulated post-translational modification to modulate cardiac contraction in physiologic response to stress. Consistent with this theory we demonstrate cTnI-ND expression in the heart improves cardiac function in animals exposed to the chronic cardiac stress of aging (Table 1).

Myofilament Effects of cTnI-ND Expression

The structural change in cTnI following N-terminal removal to modulate contraction directly at the myofilament level is central to the cTnI-ND improvement of cardiac function following the chronic stress of aging. Exchange of exogenous cTnI-ND containing Tn into skinned cardiac trabeculae induces decreases Ca²⁺ sensitivity and increases cross-bridge cycling in the absence of the Ca²⁺-handling system (29). Identical to these findings we demonstrate skinned fibers from the cTnI-ND mouse heart exhibit similarly decreased Ca²⁺ sensitivity and increased tension cost, reflecting an increased myosin cross-bridge cycling rate (Figs. 2 and 3) (27). At the muscle level, together these changes increase the rate of contractile velocity and muscle relaxation (6). Such an increased rate of relaxation (decreased IVRT; Table 1) and increased cardiac contractility (percent fractional shortening; Table 1) are exactly what we observed for cardiac function of the aged cTnI-ND mouse in vivo. The presence of these myofilament alterations in the young adult mouse heart, prior to the effects of aging induced cardiac stress, supports the effects of the cTnI-ND modification as at least partially causal to the improved cardiac function in aged mice.

During increased cardiac demand the heart must increase cardiac output by increasing the stroke volume and heart rate. In the presence of the increased rate, the cardiac cycle must also shorten or ventricular filling will become impeded and stroke volume decreased. Transient shortening of the systolic and diastolic portions of the cardiac cycle result from activation of the β-adrenergic pathway. During chronically increased cardiac demand the β-adrenergic pathway is desensitization (30, 31) reducing the ability of the heart to maintain cardiac output, which is detrimental to cardiac function (32, 33). We propose the combination of the cTnI-ND induced decreased Ca²⁺ sensitivity and increased cross-bridge cycling result in an increased relaxation (Tables 1 and 2) to expand the contractile reserve of the heart allowing for an improved response to increased cardiac demand separate from β-adrenergic stimulation. Although the cTnI-ND increased cycling will increase the energy consumed per time, at the same stroke volume the increased energy usage is offset by the concurrent decrease in Ca²⁺ sensitivity to shorten the systolic portion of the cardiac cycle and maintain overall contractile energy usage. Unlike other genetic cardiomyopathy models that exhibit decreased energetic efficiency resulting from increases in both Ca²⁺ sensitivity and tension cost (34, 35), cTnI-ND Tg mice do not exhibit altered cardiac morphology or decreased contractile function further supporting the cTnI-ND modification as non-detrimental. We cannot rule out a beneficial effect of cTnI-ND expression as partially resulting from compensatory alterations in Ca²⁺ handling, MyBP-C phosphorylation or the decreased β-MHC expression (Fig. 4) observed in the aged cTnI-ND mouse. However, the presence of the cTnI-ND effects prior to stress strengthens the role of cTnI-ND expression as a beneficial mechanism to maintain function during periods of chronic cardiac stress such as that occurring during natural aging or disease development.

cTnI Post-translational Modification

The cTnI-ND Tg heart demonstrates a significant decrease in the amount of intact cTnI consistent with maintenance of the total cTnI amount in the presence of the cTnI-ND expression (Fig. 1C). Similar to previous publications from our laboratory using one-dimensional non-equilibrium isoelectric focusing (NEIEF), two-dimensional DIGE demonstrates intact cTnI from the NTg heart exists as 7 differently charged species consisting of one non-phosphorylated and 6 phosphate-containing species (Fig. 6 and Ref. 28). The majority of the cTnI-containing phosphate in the NTg was comprised of the P-2 and P-4 species, consistent with the reported high levels of Ser-23/24 pTnI in mice (28). Unlike the intact cTnI, the cTnI-ND expressed in the heart migrated as a single, non-phosphorylated spot (Figs. 6 and 9). Importantly the remaining intact cTnI expressed in the cTnI-ND mice was phosphorylated demonstrating the signaling pathways to pTnI in cTnI-ND Tg mice remained functional (Figs. 6 and 7). The lack of phosphorylation in the cTnI-ND molecule demonstrates either the N-terminal structure is necessary or Ser-23/24 must be phosphorylated prior to phosphorylation of the remaining cTnI sites. This effect may be important for the restricted proteolysis of cTnI to regulate cardiac function.

In addition to the complete lack of cTnI-ND phosphorylation, both total and Ser-23/24 phosphorylation of the remaining intact pTnI in the cTnI-ND heart was significantly decreased compared with NTg (Figs. 7 and 8). Altered β-adrenergic signaling was not likely responsible for the differences between NTg and cTnI-ND Tg mice as total MyBP-C phosphorylation was not different between these animals (data not shown). This decreased pTnI likely results from the functional similarity of the cTnI-ND to that of cTnI Ser-23/24 phosphorylation serving as a negative feedback mechanism to depress pTnI at these sites. Consistent with these findings, the amount non-phosphorylated cTnI in the NTg was decreased prior to the stress of aging to 44% of the cTnI-ND Tg at 4 months of age (Fig. 9A). Following adaptation to cardiac stress at 16 months of age the non-phosphorylated cTnI expressed in the NTg increased such that it was no longer different from that observed in the cTnI-ND (Fig. 9B). The difference between the lack of a significant change in total ProQ phosphorylation and non-phosphorylated cTnI at 16 months of age likely results from the variable response in NTg to aging stress as, although not significant, total phosphorylation trended toward being decreased in the NTg at 16 months (Fig. 7B). In addition, the finding that cTnI23/24 phosphorylation between the NTg and cTnI-ND was not significantly altered at 16 compared to 4 months of age (Fig. 8B) strongly suggests a change in phosphorylation of other cTnI Sites as responsible for this difference. Whatever cTnI site-specific phosphorylation changes are responsible for the NTg increase of non-phosphorylated cTnI, this finding clearly represents an adaptation in the NTg heart as a response to aging that did not occur to the same degree in the cTnI-ND Tg, presumably as the result of the cTnI-ND improved Tg cardiac function.

Mechanistic Implications of cTnI-ND Expression

Insight into the molecular mechanism of how removing the cTnI N-terminal extension directly affects myofilament activation can be gained from our understanding of cTnI PKA phosphorylation at Ser-23/24. The role of β-adrenergic stimulation to phosphorylate Ser-23/24 in the N-terminal extension of cTnI causing decreased Ca²⁺ sensitivity (7, 8) and increased cross-bridge cycling (20, 36) is well established. In the presence of Ca²⁺, the cTnI N-terminal extension interacts with the cTnC N-regulatory lobe to stabilize the TnC open conformation and promote the actin-myosin interaction (37, 38). Upon incorporation of negatively charged phosphates at Ser-23/24 of cTnI, the N-terminal structure is altered causing it to rotate away from cTnC (39), weakening its interaction with the cTnC N-lobe (37, 38) and shifting the cTnC regulatory domain open/closed conformation distribution toward the closed conformation. The closed cTnC conformation promotes decreased Ca²⁺ binding affinity to cTnC, binding of the cTnI inhibitory subunit to actin and decreased Ca²⁺ sensitivity of force production.

Previously we demonstrated β-adrenergic stimulation of NTg hearts increased the maximal rate of relaxation and decreased Ca²⁺ sensitivity such that these animals no longer differed from cTnI-ND mice (14). Here we demonstrate for the first time that expression of the cTnI-ND decreases Ca²⁺ sensitivity (Fig. 2), improves in vivo cardiac contractility and lusitropy (Table 1) similarly to cTnI PKA phosphorylation in vivo at the level of the whole heart similar to that observed in Tg mice containing cTnI pseudo-phosphorylated at Ser-23/24 (40). Together with previous biochemical data (9–11), these data suggest removal of the cTnI N-terminal extension behaves similarly to the N-terminal repulsion of cTnI by Ser-23/24 PKA phosphorylation to remove cTnI stabilization of the cTnC regulatory domain shifting its conformation toward the closed state. It is worth noting that this mechanism functions without an overall chronic activation of β-adrenergic pathways, that has proven detrimental to cardiac function (see Ref. 41 for review). Thus the cTnI N-terminal extension serves as a mechanism to modulate cardiac Ca²⁺ sensitivity and cross-bridge cycling by β-adrenergic induced phosphorylation at Ser-23/24 or through its removal, both resulting from an alteration of its interaction with TnC to favor the closed conformation.

Summary

The significance of the findings presented here demonstrate restricted proteolytic removal of the modulatory cTnI N-terminal extension functions as a mechanism for the heart to improve cardiac contractility during stress such as aging. Importantly the structure of this proteolytic truncation does not impair the function of cTnI within the thin filament, but returns the cTnI structure to that of the less adapted skeletal cTnI. Unlike direct activation of the β-adrenergic pathway, that also increases Ca²⁺ cycling, generation of the cTnI-ND functions to maintain cardiac contractility and alleviate the compensatory changes that occur in the presence of chronic stress. This cTnI-ND induced increase in cardiac contractility and lusitrophy in the absence of increased Ca²⁺ cycling supports this modification as a potential long-term treatment for cardiac dysfunction in humans.