Maternal obesity impairs fetal cardiomyocyte contractile function in sheep

Abstract

Obesity is a major public health problem worldwide. In the United States, one-third of women of reproductive age are obese. Human studies show that maternal obesity (MO) predisposes offspring to cardiovascular disease. However, the underlying mechanisms remain unclear. Given the similarities between pregnancy in sheep and humans, we studied sheep to examine the impact of MO on fetal cardiomyocyte contractility at term. We observed that MO impaired cardiomyocyte contractility by reducing peak shortening and shortening/relengthening velocity, prolonging time to relengthening. MO disrupted Ca²⁺ homeostasis in fetal cardiomyocytes, increasing intracellular Ca²⁺ and inducing cellular Ca²⁺ insensitivity. The Ca²⁺-release channel was impaired, but Ca²⁺ uptake was unaffected by MO. The upstream kinases that phosphorylate the Ca²⁺-release channel—ryanodine receptor-2, PKA, and calmodulin-dependent protein kinase II—were activated in MO fetuses. Contractile dysfunction was associated with an increased ratio of myosin heavy chain (MHC)-β to MHC-α and upregulated cardiac troponin (cTn)-T and tropomyosin, as well as cTn-I phosphorylation. In summary, this is the first characterization of the effects of MO on fetal cardiomyocyte contractility. Our findings indicate that MO impairs fetal cardiomyocyte contractility through altered intracellular Ca²⁺ handling, overloading fetal cardiomyocyte intracellular Ca²⁺ and aberrant myofilament protein composition. These mechanisms may contribute to developmental programming by MO of offspring cardiac function and predisposition to later life cardiovascular disease in the offspring.—Wang, Q., Zhu, C., Sun, M., Maimaiti, R., Ford, S. P., Nathanielsz, P. W., Ren, J., Guo, W. Maternal obesity impairs fetal cardiomyocyte contractile function in sheep.

Obesity is an exponentially increasing public health epidemic and economic burden worldwide (1–3). Currently, ∼18–35% of pregnant women in the United States are obese (4, 5). Epidemiologic studies suggest that maternal obesity (MO) during pregnancy exhibits intergenerational effects by programming offspring (F1) to increased risk of obesity and cardiometabolic problems (6–9), including insulin resistance, heart disease, hypertension, and vascular dysfunction (10–14). Both maternal under- and overnutrition play important roles in programming fetal heart development and function (15–18). Several studies have shown that MO increases the risk of congenital heart defects and impairs F1 fetal diastolic function (19–23). Furthermore, impaired myocardial function, increased septal thickness, and lower left ventricle (LV) Doppler velocity have been reported in fetuses of human mothers with MO (24). In addition, MO has been shown to program F1 cardiac structure (25–27).

Multiple studies of fetal cardiovascular programming have been performed in rodent models (18, 25, 28–33). In addition to their many strengths, rodent models of fetal programming have some limitations related to differences between precocial and altricial species in the timing of developmental mechanisms that result in programming. A good example is the timing of the peripartal increase in the circulating glucocorticoids that are responsible for terminal differentiation of multiple organs (34). With reference to the cardiovascular system, rodent heart rates are much higher than those in sheep and primates, including humans. Therefore, studies in precocial species are necessary to enable translation to programming in human development. Sheep share many similarities with human pregnancy (i.e., singleton gestation being the most common litter size, which is important in terms of the nutritional burden placed on the mother; comparable maternal size and adiposity; maternal:fetal weight ratio; length of gestation, important in the duration of nutritional challenges; birth weight; similar organogenesis for major body systems; equivalent rates of prenatal protein accretion and fat deposition; and relative maturity at birth) (35–43). Thus, a sheep MO model has unique potential for the study of the effects of MO programming on the fetal heart.

We have developed and characterized a diet-based sheep model of MO. Nulliparous ewes were fed an obesogenic diet at 150% National Research Council (NRC) recommendations (44) or a control diet at 100% NRC recommendations from 60 d before conception throughout gestation. To the best of our knowledge, this is the only sheep model of overnutrition programming by obesity that begins before pregnancy and thus parallels obesity, which is almost always present in women before pregnancy. This model has been well characterized. During the entire period of pregnancy, MO ewes gain a mean of 65–70% of their initial body weight, whereas control ewes gain only 15–20% (16, 45–47). At midgestation, fetuses of obese mothers exhibit a weight gain 30% greater than controls, altered organ growth and development, and increased adiposity in combination with elevated plasma glucose and insulin levels (47, 48). From mid to late gestation, ventricular weight, LV and right ventricle free wall weights, and LV wall thickness of fetuses of obese ewes increase compared with fetuses of control ewes (15, 49, 50). By late gestation, fibrosis and increased fetal heart connective tissue accumulation, associated with an upregulated TGF-β/p38 signaling pathway, are obvious in the MO fetal sheep myocardium (50). Furthermore, MO induces irregular myofiber orientation, increases interstitial space, and increases lipid droplet accumulation in fetal ventricular tissue, with upregulated levels of proinflammatory mediators (51). Despite the thicker wall and morphologic changes, MO fetal heart function is compromised, with impaired cardiac reserve when afterload is increased (16).

Despite the extensive characterization of this model, the molecular and cellular mechanisms causing compromised heart function in fetuses of MO mothers remain unclear. To rectify this deficiency, we studied the impact of MO on fetal cardiomyocyte contractile function and the underlying molecular mechanisms. We assessed contractility and intracellular Ca²⁺ handling in isolated LV cardiomyocytes from fetuses of normal-weight ewes (control group) and MO ewes (obese group). Cellular and molecular mechanisms that potentially govern contractile function were examined with a focus on intracellular Ca²⁺ handling and contractile and regulatory proteins in the sarcomere.

MATERIALS AND METHODS

Experimental animals

Nulliparous Rambouillet/Columbia crossed ewes obtained from the University of Wyoming Animal Science flock were fed either a highly palatable diet at 100% of NRC (44) recommendations (control) or 150% of NRC recommendations (MO) from 60 d before and throughout pregnancy. Water was available at all times. At gestational d135 (0.9 gestation; term 145 d), fetuses were delivered by C-section from ewes under general anesthesia initiated with ketamine (10 mg/kg) followed by isoflurane inhalation general anesthesia (2.5%) and euthanized by exsanguination, while the ewes and fetuses were still under general anesthesia. Ewes were then euthanized with an overdose of pentobarbital sodium (Abbott Laboratories, Abbott Park, IL, , USA) (45). Fetal hearts were perfused for cardiomyocyte isolation, and heart tissues were collected, snap frozen, and stored at −80°C. The fetuses studied were obtained from 11 control ewes and 14 MO ewes. Among those, 5 control and 7 MO fetuses were used for cardiomyocyte isolation, and 6 fetuses from control ewes and 7 fetuses from MO ewes were used for Western blot analysis. All animal procedures were approved by the Animal Use and Care Committee at the University of Wyoming (Laramie, WY, USA), and the sheep were housed in facilities accredited by the Association for Assessment and Accreditation of Laboratory Animal Care International.

Isolation of fetal sheep cardiomyocytes

The protocol for isolating cardiomyocytes was modified from previous reports (52–55). The fetal hearts were dissected and perfused with heparin (20,000 IU in 10 ml saline) through the coronary vessels via the ascending aorta followed by 10 ml of saturated KCl (34%) at room temperature. The ascending aorta was cannulated, and the fetal heart was mounted in a temperature-controlled (37°C) perfusion apparatus. After perfusion with oxygenated Ca²⁺-free Tyrode buffer (in mM) [NaCl 125, KCl 4.5, MgSO₄ 1.2, NaH₂PO₄ 2, 4-(2-hydroxyethyl)-1-piperazineethanesulfonic acid (HEPES) 25, pyruvate 5, and glucose 10 (pH 7.4)] and gassed with 5% CO₂ and 95% O₂ for 5 min, the fetal heart was digested for 15 min with an enzyme solution (Ca²⁺-free Tyrode buffer with 200 U/ml collagenase and 0.1 mg/ml protease). The fetal heart was then perfused with Ca²⁺-free Kraft-Brühe (KB) solution (in mM) [glutamic acid 74, KCl 30, KH₂PO₄ 30, taurine 20, MgSO₄ 3, EGTA 0.5, HEPES 10, and glucose 10 (pH 7.37)] and gassed with 5% CO₂ and 95% O₂ for 10 min (∼300 ml) to eliminate any remaining enzyme. After the digestion process, the fetal heart was removed from the cannula, and the LV free wall was dissected and washed with KB solution at 37°C. Individual cardiomyocytes were dispersed by cutting the LV free wall into small pieces in KB solution and filtering them through a cell strainer with a pore size of 200 μm (Thermo Fisher Scientific, Waltham, MA, USA).

Extracellular Ca²⁺ recovery

Isolated cardiomyocytes were pelleted by centrifugation at 500 rpm for 5 min. The supernatant was removed and the pellet resuspended in Ca²⁺-free Tyrode solution. The extracellular Ca²⁺ level was gradually raised to 3, 9, 21, 45, 93, 189, 381, and 616 μM to a final concentration of 1 mM with a 10-min interval between each concentration. Cardiomyocytes were then left at 1 mM at room temperature for a half-hour, after which they were ready for mechanical and intracellular Ca²⁺ assessment.

Fetal cardiomyocyte shortening/relengthening

Rod-shaped fetal sheep cardiomyocytes with clear edges were selected for measurement of mechanical properties with a SoftEdge Myocam system (IonOptix, Milton, MA, USA). IonWizard software was used to capture changes in cardiomyocyte length during shortening and relengthening by using the SoftEdge and SarcLen acquisition modules to record cell and sarcomere length. Cardiomyocytes were placed in a C-Stim Cell MicroControls superfusion chamber system (IonOptix) on the stage of an inverted microscope (Olympus, Tokyo, Japan) and were superfused with the contractile buffer containing (in mM) NaCl 131, KCl 4, MgCl₂ 1, glucose 10, HEPES 10, and CaCl₂ 2 (pH 7.4). Cardiomyocytes were field stimulated with an acute MyoPacer field stimulator (IonOptix) to electrically pace cellular contractions. The MyoPacer frequency setting for cardiomyocyte contractility measurement was 0.5 Hz, stimulation pulse duration was 3 ms, and voltage was 40 V. The cardiomyocyte being measured was displayed on the computer monitor via a MyoCam-S (IonOptix) digital acquisition camera, and the amplitude and velocities of shortening and relengthening were recorded.

Cell shortening and relengthening were assessed by using the following indices: peak shortening (PS), the shortest cell/sarcomere length of cardiomyocytes contracted on electrical stimulation, which is indicative of peak ventricular contractility; time-to-PS (TPS), the duration of myocyte shortening, which is indicative of contraction duration; time to 90% relengthening (TR₉₀), the time to reach 90% relengthening, which represents cardiomyocyte relaxation duration (90% rather than 100% relengthening was used to avoid the noisy signal present at baseline contraction); and maximum velocities of shortening (+dl/dt) and relengthening (−dl/dt), maximum slope (derivative) of the shortening and relengthening phases, which are indicators of maximum velocities of ventricular pressure increase and decrease (56, 57).

Intracellular Ca²⁺ handling

A separate cohort of cardiomyocytes was loaded with Fura-2/AM (0.5 μM) for 15 min, and fluorescence intensity was recorded with a dual-excitation fluorescence photomultiplier system (IonOptix). Cardiomyocytes were placed on an IX-70 inverted microscope (Olympus) and imaged with a Fluor ×40 oil objective. Cardiomyocytes were exposed to light emitted by a 75-W lamp and passed through either a 360- or 380-nm filter, while being stimulated at 0.5 Hz for contraction. Fluorescence emissions were detected between 480 and 520 nm, and qualitative change in Fura-2 fluorescence intensity (FFI) was inferred from the FFI ratio of the fluorescence intensity at the 2 wavelengths (360/380 nm). Fluorescence decay time (single exponential decay rate) was measured as an indicator of intracellular Ca²⁺ clearing rate (57, 58).

Wheat germ agglutinin staining

LV sections (5 μm) were stained with FITC-conjugated wheat germ agglutinin (WGA; MilliporeSigma, Burlington, MA, USA), and cardiomyocyte cross-sectional areas were calculated from randomly selected cells on a digital microscope with ImageJ (v.1.51K) software [National Institutes of Health (NIH), Bethesda, MD, USA; https://imagej.nih.gov/ij/] (59, 60).

Immunofluorescence staining

Paraffin-embedded sections of ventricular tissues (5 μm) were deparaffinized, and the antigen was retrieved in citrate buffer [90 mM sodium citrate, 9 mM citrate acid, and 0.5% Tween 20 (pH 6.0)]. The tissue sections were blocked in 5% bovine serum albumin for 1 h before incubation with antibodies against myosin heavy chain (MHC)-α and -β (4 μg/ml; Developmental Studies Hybridoma Bank at University of Iowa, Iowa City, IA, USA) at 4°C overnight, followed by incubation in a goat anti-mouse IgG1 Alexa Fluor 555 antibody (1:500; Thermo Fisher Scientific) at room temperature for 90 min. Tissues were visualized under a confocal microscope (Zeiss, Jena, Germany). The mean fluorescence intensity was calculated on the whole-section image with ImageJ software (NIH) (60, 61).

Western blot analysis

Fetal heart tissues were homogenized in urea-thiourea buffer [8 M urea, 2 M thiourea, 75 mM DTT, 3% SDS, 0.05% bromophenol blue, and 0.05 M Tris (pH 6.8)] as described by Guo et al. (62). Total protein was separated by SDS-PAGE gel and transferred onto a PVDF membrane. The membrane was probed with the following antibodies: anti-phospho-cardiac troponin I (p-cTn-I) (Ser^23/24), anti-tropomyosin, anti-sarco/endoplasmic reticulum Ca-ATPase (Serca)-2, anti-PKA, anti-phosphorylated PKA (Thr¹⁹⁷), anti-Ca²⁺/calmodulin–dependent protein kinase II (CaMKII), anti-phosphorylated CaMKII (Thr²⁸⁶), anti-phospholamban (PLN), and anti-β-actin (loading control) (1:1000; Cell Signaling Technology, Danvers, MA, USA); anti-ryanodine receptor (RyR), anti-MHC-α, anti-MHC-β, anti-cTn-T, and anti-cTn-I (1:300; Developmental Studies Hybridoma Bank); anti-phosphorylated RyR-2 (Ser²⁸⁰⁸ and Ser²⁸¹⁴) (1:1500; Badrilla, Leeds, United Kingdom); and anti-FKBP12.6 and anti-phosphorylated PLN (Ser¹⁶) (1:500; Santa Cruz Biotechnology, Dallas, TX, USA). Horseradish peroxidase–coupled secondary antibodies were used for membrane incubation. After immunoblot, the films were developed with ECL Western blot substrate (Bio-Rad, Hercules, CA, USA) and exposed to CL-Xposure film (Thermo Fisher Scientific). Density analysis of immunoblot bands was performed with NIH ImageJ by normalizing to loading control β-actin or pan proteins for phosphorylated proteins (60).

Statistical analysis

Prism software (GraphPad, La Jolla, CA, USA) was used for statistical analysis. Results are expressed as means ± sem. Statistical significance was determined with an unpaired 2-sided t test analysis of differences between the 2 specified groups: control and MO. Significance was set at values of P < 0.05.

RESULTS

MO changes in fetal heart LV biometrics at late gestation

MO increased fetal LV free wall weight and thickness (Fig. 1A, B). To determine hypertrophy at the cellular level, immunostaining of paraffin-embedded sections with WGA staining was performed. The cross-sectional area of LV cardiomyocytes was increased in fetuses of mothers with MO (Fig. 1C, D). These results reveal that MO leads to fetal heart hypertrophy by late gestation, suggesting that cardiac dysfunction in F1 of obese mothers later in life could originate from changes in cardiac geometry during early heart development.

Figure 1 .

LV biometric properties in 0.9 gestation fetuses of control and MO mothers. A) LV weight (n = 5 control fetuses and n = 4 MO fetuses). B) LV thickness (n = 7 control fetuses and n = 7 MO fetuses). C) Cardiomyocyte cross-sectional area, ∼1000 cells (∼178–258 per animal) randomly selected from 5 control and 5 MO fetal hearts. D) Representative images with WGA staining of cardiomyocyte cross-sectional area of LV tissues. Means ± sem. *P < 0.05, **P < 0.01, ***P < 0.001 vs. control.

MO impairs fetal cardiomyocyte contractile properties

To determine the nature of changes in cellular contractile properties, we isolated fetal cardiomyocytes from the control and MO groups. MO increased fetal cardiomyocyte resting cell length from 67.3 ± 1.3 to 72.7 ± 1.3 μm (Fig. 2A), with little effect on cardiomyocyte sarcomere length (Supplemental Fig. S1A). Cardiomyocyte contractility based on cell-length measurements indicated that MO decreased PS and V_max of shortening and relengthening (±dL/dt) associated with prolonged TR₉₀ (Fig. 2B–G). Contractile dynamics based on sarcomere length measurements of MO fetal cardiomyocytes showed changes similar to those obtained from cell-length measurements (Supplemental Fig. S1). These results indicate that MO increased the resting cell length of fetal cardiomyocytes, although it suppressed cardiac contractility, as evidenced by a reduction of 30% in PS. Decreased PS was observed, along with slower speed of shortening and relengthening (±dL/dt) and prolonged TR₉₀. On the other hand, the duration of shortening, indicating the time from resting cell length to the maximum cell shortening of fetal cardiomyocytes, was unaffected by MO (Fig. 2F). Duration of TPS was shown to be unaffected, according to both the cell and sarcomere measurements (Supplemental Fig. S1E).

Figure 2 .

Mechanical contractile properties based on cell-length measurement of LV cardiomyocytes from 0.9 gestation fetuses of control and MO mothers. Resting cell length (A); PS (B); representative contractile trace (C); maximum velocity of shortening (D); maximum velocity of relengthening (E); TPS (F); and TR₉₀ (G). Means ± sem (n = 140 cells: 13–33 cells/animal from 5 control fetal hearts; n = 100 cells: 11–32 cells/animal from 7 MO fetal hearts). **P < 0.01, ***P < 0.001 vs. control.

MO impairs intracellular Ca²⁺ homeostasis in fetal cardiomyocytes

To explore the possible mechanisms responsible for the impaired fetal cardiomyocyte contractile properties in response to MO, intracellular Ca²⁺ levels were assessed with the Fura-2 fluorescence technique. MO elevated resting intracellular Ca²⁺ levels from 1.31 ± 0.02 to 1.36 ± 0.01. Peak intracellular Ca²⁺ ratios increased from 1.34 ± 0.03 to 1.44 ± 0.01. FFI rose (ΔFFI) from 0.039 ± 0.002 to 0.080 ± 0.004 (Fig. 3A–C). However, the intracellular Ca²⁺ clearance rate was unaffected by MO (Fig. 3D). These results revealed that MO led to elevated resting intracellular Ca²⁺ and a greater release of Ca²⁺ in response to electric stimulation (ΔFFI) in fetal cardiomyocytes, but that intracellular Ca²⁺ uptake rate was similar in MO compared with control fetal cardiomyocytes. These findings suggest that intracellular Ca²⁺ is overloaded in MO fetal cardiomyocytes, which could be a mechanism leading to impaired fetal cardiomyocyte contractile properties.

Figure 3 .

Intracellular Ca²⁺ transient properties of LV cardiomyocytes from 0.9 gestation fetuses of control and MO mothers. Resting FFI (A); peak FFI (B); electrically stimulated rise in FFI (∆FFI) (C); and intracellular Ca2⁺ decay rate (single exponential) (D). Means ± sem (n = 69 cells: 11–23 cells/animal, from 4 control fetal hearts; n = 103 cells: 12–30 cells/animal, from 5 MO fetal hearts). *P < 0.05, ***P < 0.001, ****P < 0.0001 vs. control.

Changes in Ca²⁺-handling proteins in fetal cardiomyocytes of obese mothers

Western blot analysis (Fig. 4A) revealed similar levels in total RyR between MO and control fetal hearts. However, phosphorylation of RyR2 at Ser²⁸⁰⁸ and Ser²⁸¹⁴ was significantly enhanced in MO fetal myocardium (Fig. 4B–F). Increased expression of the RyR2 binding protein FK 506 binding protein 12.6 was observed (Fig. 4G). Serca-2 and PLN are 2 major proteins in the Ca²⁺-uptake pump, and changes in these proteins may impair Ca²⁺ uptake (63). Expression of Serca-2 and phosphorylated PLN (Ser¹⁶) were unchanged (Fig. 4H, I, K). However, expression of pan PLN was increased in MO fetal myocardium (Fig. 4J). These results suggest that Ca²⁺ was leaking from the Ca²⁺-release channel, but because there was no change in the Ca²⁺ uptake pump, the intracellular Ca²⁺ level in MO fetal cardiomyocytes was elevated.

Figure 4 .

Western blot of Ca²⁺ pump proteins in SR from control and MO 0.9 gestation fetal heart LVs. SDS-PAGE gel image of pRyR2(Ser²⁸⁰⁸), pRyR2(Ser²⁸¹⁴), RyR(pan), FKBP12.6, Serca-2, pPLN(Ser¹⁶), and PLN(pan) (A); pRyR2(Ser²⁸⁰⁸) expression (B); pRyR2(Ser²⁸¹⁴) expression (C); RyR (pan) expression (D); pRyR2(Ser²⁸⁰⁸)-to-RyR(pan) ratio (E); pRyR2(Ser²⁸¹⁴)-to-RyR(pan) ratio (F); FKBP12.6 expression (G); Serca2 expression (H); pPLN(Ser¹⁶) expression (I); PLN(pan) expression (J); and pPLN-to-PLN(pan) ratio (K). Means ± sem.(B–D and G–J, n = 6 control fetal hearts and n = 7 MO fetal hearts, normalized to β-actin expression; E, F, and K normalized to relative pan protein expression). *P < 0.05 vs. control.

Next, we examined the upstream kinases CaMKII and PKA that phosphorylate Ser²⁸⁰⁸ and Ser²⁸¹⁴ of RyR2. MO did not affect expression of total PKA and CaMKII in the fetal myocardium, but phosphorylation of both kinases was elevated in MO fetal hearts (Fig. 5), suggesting that MO leads to activation of both PKA and CaMKII but not to altered total expression of these 2 kinases.

Figure 5 .

Western blot of kinases active in Ca²⁺ signaling in control and MO 0.9 gestation fetal heart LVs. SDS-PAGE gel image of pPKA (Thr¹⁹⁷), PKA(pan), pCaMKII(Thr²⁸⁶) and CaMKII(pan) (A); pPKA(Thr¹⁹⁷) expression (B); PKA(pan) expression (C); pPKA(Thr¹⁹⁷)-to-PKA(pan) ratio (D); pCaMKII(Thr²⁸⁶) expression (E); CaMKII(pan) expression (F); and pCaMKII(Thr²⁸⁶)-to-CaMKII(pan) ratio (G). Means ± sem. (B, C, E, and F, n = 6 control fetal hearts and n = 7 MO fetal hearts, normalized to β-actin expression; D, and G to relative pan protein expression). *P < 0.05 vs. control.

Changes in contractile and regulatory proteins in the MO fetal myocardium

To explore the effect of MO on the contractile protein Myosin and the regulatory proteins troponin and tropomyosin, we performed immunofluorescent staining with specific antibodies: anti-MHC-α and -β. MO reduced MHC-α fluorescence intensity but increased that of MHC-β (Fig. 6A–D). Furthermore, Western blot results confirmed the reduced MHC-α and elevated MHC-β protein expression in the MO fetal heart (Fig. 6E–H). These results suggest that MO decreases the levels of high-ATP hydrolysis MHC-α associated with fast twitch and increases the levels of low-ATP hydrolysis MHC-β associated with slow twitch. These changes indicate that MO slows fetal heart contraction, in agreement with our findings in single cardiomyocytes.

Figure 6 .

MHC expression in control and MO 0.9 gestation fetal heart LVs. Representative images of immunofluorescence staining of cardiac MHC-α (red), MHC-β (red) in fetal LV tissue, nucleus stained by DAPI (blue) (A); mean fluorescence intensity of MHC-α (B); mean fluorescence intensity of MHC-β (C); fluorescence intensity of MHC-β:MHC-α ratio (D); SDS-PAGE gel image of MHC-α and -β (E); MHC-α expression (F); MHC-β expression (G); MHC-β:MHC-α expression ratio (H). Means ± sem. (F and G, n = 6 control fetal hearts and n = 7 MO fetal hearts, normalized to β-actin expression; H, normalized to MHC-α protein expression). *P < 0.05 vs. control.

We next examined the expression and phosphorylation status of the regulatory proteins, troponin complex and tropomyosin. Western blot analysis (Fig. 7A) revealed that expression of cTn-T was increased by ∼60%, phosphorylation of cTn-I was increased by ∼130%, and cardiac tropomyosin was increased by 110% in fetal myocardium from the MO group (Fig. 7B–F). No changes were observed for cTn-C expression in the fetal myocardium (Fig. 7G). These results indicated that MO affected the composition of the myofilament regulatory troponin-tropomyosin complex.

Figure 7 .

Western blot of troponin complex protein markers in control and MO 0.9 gestation fetal heart LVs. SDS-PAGE gel image of cTn-T, p-cTn-I, cTn-I, cardiac tropomyosin and cTn-C (A); cTn-T expression (B); p-cTn-I expression (C); cTn-I expression (D); p-cTn-I:cTn-I ratio (E); tropomyosin expression (F); and cTnC expression (G). Means ± sem. (B–D, F, and G, n = 6 control fetal hearts and n = 7 MO fetal hearts, normalized to β-actin expression and E, normalized to relative pan protein expression). *P < 0.05 vs. control.

DISCUSSION

The findings of this study indicate that MO impairs fetal myocardial contractile function, and intracellular Ca²⁺ homeostasis. The compromised cardiomyocyte contractile function and intracellular Ca²⁺ handling seen in fetuses from MO ewes were accompanied by changes in contractile and regulatory proteins. Previous studies suggested that MO programs cardiac remodeling and disrupts cardiac homeostasis in fetal sheep heart (16, 18, 25, 33, 51), although little information is available relating to activity of individual cardiomyocytes. Maternal overfeeding during early to midgestation has been reported to increase ventricular free wall weights and wall thickness in fetal sheep (49, 50). Moreover, MO induces cardiac hypertrophy in the mouse, human, and sheep fetal heart (25–27). Our findings are consistent with these earlier findings. We provide further evidence of cardiac remodeling and hypertrophy indicated by elongated single cardiomyocyte length and enlarged cardiomyocyte cross-sectional area. In vitro, whole-heart Langendorff studies indicated that contractile function in the MO fetal heart is compromised under high workloads (16). However, contractile profiles of individual fetal sheep cardiomyocytes have not been studied. Our results revealed that MO-reduced contractile function in fetal cardiomyocytes decreased cell shortening and slowed the speed of contraction. These findings provide important new cellular evidence of the nature of MO-induced cardiac dysfunction in single cardiomyocytes.

Intracellular Ca²⁺ release and uptake play essential roles in contractile function of the normal mammalian cardiac muscle cell (63). Two major Ca²⁺-dependent mechanisms—the availability of Ca²⁺ to the myofilaments and myofilament responsiveness to activation by intracellular Ca²⁺—are key regulators of the cardiac contractile state (64, 65). Ca²⁺ availability is regulated by the sarcoplasmic reticulum (SR). Imbalance between release and uptake of Ca²⁺ results in altered intracellular Ca²⁺ levels. When Ca²⁺ release exceeds its uptake, cells will eventually become overloaded with Ca²⁺, leading to cardiomyocyte contractile dysfunction (64, 66–69). The cellular measurement of Ca²⁺ showed that the resting and the peak levels were both higher in MO fetal cardiomyocytes because of increased Ca²⁺ release. However, the Ca²⁺ clearance rate was unaffected by MO. The higher intracellular Ca²⁺ would increase the time needed for Ca²⁺ reuptake into the SR in MO cardiomyocytes, which is consistent with prolonged cell relengthening time. Increased intracellular Ca²⁺ levels suggest Ca²⁺ overload in MO fetal cardiomyocytes, which may be associated with impaired activity of the Ca²⁺-release channel. Studies have shown that phosphorylation by PKA of RyR2 at Ser²⁸⁰⁸ and phosphorylation by CaMKII of RyR2 at Ser²⁸¹⁴ can reduce the stability of the Ca²⁺ release channel RyR2 and lead to a Ca²⁺ leak from the SR associated with cardiac contractility dysfunction and heart failure (70–74). Our data show increased activity of PKA and CaMKII and high phosphorylation levels at both Ser²⁸⁰⁸ and Ser²⁸¹⁴ of RyR2, which further confirm impaired contractile function in MO fetal cardiomyocytes at the molecular level.

Cardiomyocyte contractility is also regulated by the sensitivity of myofilaments to Ca²⁺. Evidence has indicated a role for MHC isoform composition and cardiac troponin in myocardial contractile dysfunction (75–80). It has been shown that the specific pattern of cardiac MHC isoform expression significantly affects cardiomyocyte contractile properties (81–85). In particular, isoform switching from MHC-α to -β has been reported as being associated with suppressed cardiac contractile function in failing hearts (75, 77, 79, 80, 82, 83). Therefore, the downregulated MHC-α and up-regulated MHC-β seen in MO fetal hearts could lead to reduced and slower cardiomyocyte contractility. Cardiac troponin proteins play a critical role in transducing Ca²⁺ binding to initiate cardiac muscle contraction (86–88). McAuliffe and Robbins (89) have identified only 1 cTn-T isoform in the fetal sheep heart. Although several studies have been conducted on the Ca²⁺ sensitivity and cTn-T, these studies have shown that variations in isoforms of cTn-T affect Ca²⁺ sensitivity and cardiac contractility (90–92). However, there are no reports of how various conditions affect contractility in fetal sheep in the presence of only 1 form of troponin. Data presented herein, show that increased expression of cTn-T is associated with reduced cardiomyocyte contractility. Further studies are needed to determine how MO-induced upregulation of cTn-T affects the function and structure of the troponin complex, and thereby, cardiac contractility. Increased phosphorylation of cTn-I increases the Ca²⁺ necessary to achieve half-maximum myofibrillar ATPase and decreases the amount of Ca²⁺ available to myofilaments (93). Although we failed to note any change in cTn-C, the major element in the troponin complex binding Ca²⁺ to transduce mechanosensing to the myofilament, the increased phosphorylation of cTn-I in MO fetal heart could help to explain the reduced contractility, potentially by reducing enzyme activity of myofibrillar ATPase. Further studies are necessary to determine the mechanisms of MO-induced fetal cardiomyocyte Ca²⁺ insensitivity.

Another major goal of our study was to establish a reliable method to assess fetal cardiac mechanistic function at the level of the individual fetal cardiomyocyte. The challenge in mechanical assessment of single fetal cardiomyocyte function is the ability to isolate viable fetal cardiomyocytes and buffer the cells in optimal conditions to respond to field electric stimulation. Fetal sheep cardiomyocyte isolation methods have been established for over 15 yr (52), and several elegant studies have demonstrated endocrine control of the cell cycle and growth in fetal sheep cardiomyocytes (94–97). However, our data are the first, to our knowledge, on contractility of individual fetal cardiomyocytes.

There is a need to develop and refine methods to evaluate mechanisms that control fetal sheep cardiomyocyte contractility. Reported methods maintain cardiomyocytes in a Ca²⁺ free environment (52, 54, 94–96). Calcium plays a critical role in cardiac contraction (98, 99). Therefore, the isolated Ca²⁺-free fetal sheep cardiomyocytes could not contract in response to electric stimuli. To assess the contractile function and intracellular Ca²⁺ handling, normal Ca²⁺ levels have to be restored in the fetal sheep cardiomyocyte (58, 100–102). In published studies on cardiomyocyte contractility in adult mice and rats, the concentration of extracellular Ca²⁺ increased to 1.25 mM (57–59, 103). However, different extracellular Ca²⁺ concentrations have been used in different studies: Thompson et al. (101) used 1.8 mM in isolated cardiomyocytes from guinea pigs, Chandrashekhar et al. (100) used 0.2 mM in rat cardiomyocytes, and Griffiths et al. (102) raised the extracellular Ca²⁺ to 2 mM in isolated rat cardiomyocytes. In our study, we first followed the Ca²⁺ recovery steps as described in Ren and Wold (58), with initial Ca²⁺ concentration as 100 μM and raised to 1.25 mM in 5 steps of double increments. However, the viability of fetal sheep cardiomyocytes decreased dramatically from ∼70–20%, which indicates that the fetal sheep cardiomyocytes are very sensitive to extracellular Ca²⁺. Because there are no reports investigating isolated fetal cardiomyocytes in other species, it remains to be determined whether there is similar sensitivity to Ca²⁺ concentrations in other experimental species, both precocial and altricial. Based on these observations of the sensitivity of the fetal sheep cardiomyocyte to Ca²⁺, we modified the Ca²⁺ recovery procedure to start with 3 μM and ended with 1 mM with 9 gradual incremental steps. This approach helped to maintain fetal sheep cardiomyocyte viability at ∼60%. The levels of Ca²⁺ in the contractile buffer also affected myocyte contractility. The conventional level of 1 mM Ca²⁺ in the contractile solution did not stimulate fetal sheep cardiomyocytes; however, in this study, 2 mM Ca²⁺ was determined to be the ideal concentration for the contraction solution. Higher Ca²⁺ concentrations (≥4 mM) in the contraction solution may reduce viability and result in the death of fetal sheep cardiomyocytes during the measurement of contractility. The Ca²⁺ concentration in the contraction solution determines the extracellular Ca²⁺ concentration during the measurement of contractile function. In our case, the working concentration of extracellular Ca²⁺ was 2 mM, which is the normal extracellular concentration of Ca²⁺ (104, 105). The methods we developed for Ca²⁺ recovery and assessment of contractility in fetal sheep cardiomyocytes showed that an extracellular Ca²⁺ level of 1 mM in the fetal myocyte maintains cell viability. In addition, we showed the effects of a working concentration of extracellular Ca²⁺ of 2 mM, which is the normal physiologic condition of cardiomyocytes. These guidelines will be of value in future studies.

In summary, our study represents a characterization of MO effects on fetal cardiomyocyte contractility. The data suggest that MO induces cardiac functional and geometric anomalies in fetal sheep hearts via alteration of myofilament proteins and the SR Ca²⁺ channel. In fetuses of mothers with MO, there is a fetal myocardial MHC isoform switch from the high-ATPase activity MHC-α isoform to the low-ATPase activity MHC-β, along with an altered cTn complex. This switch could be one mechanistic cause of reduced contractility in MO fetal sheep cardiomyocytes. Our results also suggest that MO disrupts intracellular Ca²⁺ homeostasis in fetal hearts by elevating phosphorylation of the SR Ca²⁺-release channel RyR2, whereas the SR Ca²⁺uptake channel was not affected. This conclusion is supported by our observation of intracellular Ca²⁺ transience in fetal sheep cardiomyocytes. That upstream Ca²⁺ signaling molecules of RyR2, PKA, and CaMKII were activated in the MO fetal myocardium suggests a vital role for dysfunctional cellular Ca²⁺ signaling mechanisms in the MO fetal heart (Fig. 8). Although further studies are needed to understand mechanisms underlying MO-induced changes in fetal cardiac contractile function, our data provide additional mechanistic evidence at both the cellular and molecular levels which, if they persist, would contribute to programming of life course adult cardiovascular disease in F1 offspring of obese mothers.

Figure 8 .

The signaling pathways linking MO to fetal heart contractile dysfunction. MO activates phosphorylation of PKA and CaMKII which further activate SR Ca²⁺ release channel RyR2 at Ser²⁸⁰⁸ and Ser²⁸¹⁴, resulting in intracellular Ca²⁺ overload. On the other hand, MO decreases MHC-α and increases MHC-β. Meanwhile, MO increases cTn-T, phosphorylation of cTn-I, and tropomyosin, which lead to Ca²⁺ insensitivity of the myofilament. We conclude that intracellular Ca²⁺ overload and insensitivity are major causes of fetal cardiomyocyte contractile dysfunction and intracellular Ca²⁺ mishandling. Up arrows: up-regulation; down arrows: down-regulation.

Supplementary Material

This article includes supplemental data. Please visit http://www.fasebj.org to obtain this information.

Glossary

CaMKII

calmodulin-dependent protein kinaseII

cTn

cardiac troponin

FFI

Fura-2 fluorescence intensity

KB

Kraft-Brühe

LV

left ventricle

MHC

myosin heavy chain

MO

maternal obesity

NRC

National Research Council

PLN

phospholamban

PS

peak shortening

RyR2

ryanodine receptor 2

SR

sacroplasmic reticulum

Serca

sarco/endoplasmic reticulum Ca-ATPase

TPS

time to peak shortening

TR₉₀

time to 90% relengthening

WGA

wheat germ agglutinin

AUTHOR CONTRIBUTIONS

Q. Wang and W. Guo designed the experiments; Q. Wang and W. Guo analyzed and interpreted the data; Q. Wang, C. Zhu, M. Sun, and R. Maimaiti performed the experiments; S. P. Ford and J. Ren contributed new reagents or analytic tools; Q. Wang and W. Guo wrote the manuscript; and P. W. Nathanielsz and J. Ren revised the manuscript.