Myocardial-restricted ablation of the GTPase RAD results in a pro-adaptive heart response in mice

Abstract

Existing therapies to improve heart function target β-adrenergic receptor (β-AR) signaling and Ca²⁺ handling and often lead to adverse outcomes. This underscores an unmet need for positive inotropes that improve heart function without any adverse effects. The GTPase Ras associated with diabetes (RAD) regulates L-type Ca²⁺ channel (LTCC) current (I_Ca,L). Global RAD-knockout mice (gRAD^-/-) have elevated Ca²⁺ handling and increased cardiac hypertrophy, but RAD is expressed also in noncardiac tissues, suggesting the possibility that pathological remodeling is due also to noncardiac effects. Here, we engineered a myocardial-restricted inducible RAD-knockout mouse (RAD^Δ/Δ). Using an array of methods and techniques, including single-cell electrophysiological and calcium transient recordings, echocardiography, and radiotelemetry monitoring, we found that RAD deficiency results in a sustained increase of inotropy without structural or functional remodeling of the heart. I_Ca,L was significantly increased, with RAD loss conferring a β-AR-modulated phenotype on basal I_Ca,L Cardiomyocytes from RAD^Δ/Δ hearts exhibited enhanced cytosolic Ca²⁺ handling, increased contractile function, elevated sarcoplasmic/endoplasmic reticulum calcium ATPase 2 (SERCA2a) expression, and faster lusitropy. These results argue that myocardial RAD ablation promotes a beneficial elevation in Ca²⁺ dynamics, which would obviate a need for increased β-AR signaling to improve cardiac function.

Keywords: GTPase; RGK GTPase; calcium channel; cardiomyocyte; gene knockout; heart failure; positive inotrope; transgenic mice.

Figure 1.

Myocardial deletion of RAD. A, graphic representation of the RRAD conditional targeting strategy. Flanking exons 3 and 4 of RRAD with loxP sites generated the floxed allele (fl). Cre-mediated recombination of the loxP sites results in the cKO allele (Δ). The approximate locations of the PCR primers used for genotyping (a and b) are shown. B, genotyping PCR of tail genomic DNA yielded the following genotypes: RRAD^wt (lane 1), RRAD^w/fl (lane 2), and RRAD^fl/fl (lane 3). Amplicons were generated by primers a and b as shown in A. C, Western blot analysis for RAD in protein lysates from wildtype (WT), global RAD knockout (gRADKO), RAD^fl/fl, and RAD^Δ/Δ (2 weeks post-tamoxifen treatment) from total heart (top panel) and spleen (bottom panel). Note that RAD expression is retained in the spleen of RAD^Δ/Δ mice following tamoxifen treatment. D, qRT-PCR analysis in mRNA samples from RRAD^fl/fl and RRAD^Δ/Δ hearts. MW, 100 bp DNA ladder. E, quantification by Western blotting of RAD levels in the heart 4 weeks post-tamoxifen treatment (n = 7); ***, p = 0.001.

Figure 2.

Cardiomyocyte-restricted RAD deletion does not induce markers of myocardial pathology. A, qRT-PCR for ANF and RCAN1 mRNA expression levels. RAD^fl/fl (n = 6) and RAD^Δ/Δ (n = 5) as shown. B, heart weight was not significantly different between RAD^Δ/Δ and RAD^fl/fl. C, representative pictures of PicroSirius Red staining with fibrosis quantification in D. RAD^fl/fl (n = 5) and RAD^Δ/Δ (n = 7) are shown. E, Western blot analysis with quantification in F demonstrates no difference in CTGF expression. Lanes were run on the same gel, but noncontiguous lanes are marked by the black line. Rad^fl/fl (n = 7) and RAD^Δ/Δ (n = 9) are shown. **, p = 0.008. G, aortic pressure recordings, 4 continuous days from baseline (prior to tamoxifen), and from the same mice 2 weeks after tamoxifen administration (RAD^Δ/Δ). Data are average of 3 mice. Data were fitted to a sine wave with a fixed cycle length of 24 h. Active (nighttime) and resting (daytime) blood pressures were 129 ± 2 mm Hg (baseline), 132 ± 5 mm Hg (0.5 months); and 118 ± 2 mm Hg (baseline), 122 ± 6 (0.5 months), respectively.

Figure 3.

Myocardial RAD deletion results in a rapid, stable gain of cardiac function without pathological structural remodeling in vivo. A, representative M-mode short-axis echocardiography of female mice 1-week through 15 months following tamoxifen injection. Scale bars of depth: RAD^fl/fl baseline = 0.69 mm; RAD^fl/fl 15 months = 0.67 mm; RAD^Δ/Δ baseline = 0.69 mm; RAD^Δ/Δ 15 months = 0.76 mm; all images have a duration of 1 s. B, ejection fraction; C, left ventricular inner dimension; D, left ventricular anterior; and E, posterior wall thickness. Dimensions (C–E) are shown in diastole; RAD^fl/fl (n = 7) and RAD^Δ/Δ (n = 17; by genetic modification, p = 0.0002; F = 20; by time p = 0.04; F = 2; by gene × time, p = 0.04; F = 2). F–H, echocardiography of male and female mice 3 months after tamoxifen treatment. F, ejection fraction. Sidak's multiple comparison test showed RAD^Δ/Δ increased in EF (p < 10⁻³ males, p < 10⁻⁴ females; by genetic modification, p < 10⁻⁴, F = 49; by sex, p < 0.01; F = 8; no interaction). RAD^fl/fl 2-way ANOVA shows female gender contributed 8% to variance suggesting more sensitivity than males to RAD^Δ/Δ (p < 0.06). G, anterior wall thickness and H, posterior wall thickness during diastole was not different by RAD^Δ/Δ or by male versus female. I, linear regression of EF as a function of HR shows HR differences do not account for elevated EF in RAD^Δ/Δ. The dashed lines show 95% confidence limits. The slope is not different between RAD^Δ/Δ and RAD^fl/fl; elevation of EF is significantly different (p < 10⁻⁴). **, p = 0.006; ****, p < 10⁻⁴.

Figure 4.

Myocardial RAD deletion results in increased calcium current (I_Ca,L). A, representative family of I_Ca,L traces, V_test ranging from −75 to +45 mV in 10-mV increments, with voltage protocol schematic shown above. B, representative family of I_Ca,L traces from Vpre ranging from −50 to 0 mV in 10-mV increments for determining steady-state availability, with voltage protocol schematic shown above. Scale bars: 500 pA and 200 ms. Scale bars common to panels A and B. C, current/voltage curve shows that I_Ca,L density is increased in cardiomyocytes from RAD^Δ/Δ compared with RAD^fl/fl. D, conductance transform of the current/voltage curve demonstrates higher maximal conductance in RAD^Δ/Δ compared with RAD^fl/fl with quantification shown in E. F, conductance-voltage curve normalized to maximal conductance illustrates that RAD^Δ/Δ cardiomyocytes have shifted the activation midpoint. Smooth curves are Boltzmann distribution fitted to data. Steady-state availability was not different. Note that voltage for the availability curve is the value for the 500-ms pre-pulse potential step (see schematic in A). G, activation midpoint is significantly negative-shifted in RAD^Δ/Δ. Current data are from RAD^fl/fl (n = 7 mice, n = 18 cells) and RAD^Δ/Δ (n = 7 mice, n = 17 cells); ****, p < 10⁻⁴. H, qRT-PCR for Ca_V1.2 showed no change in expression. Data were displayed from RAD^fl/fl (5 mice) and RAD^Δ/Δ (5 mice). I, Western blot analysis demonstrates no difference in Ca_V1.2 expression between RAD^fl/fl and RAD^Δ/Δ. Lanes were run on the same gel, but noncontiguous lanes are marked by the black line; total protein stained with Coomassie Blue and quantification are shown on the right. Western blotting data from RAD^fl/fl (7 mice) and RAD^Δ/Δ (9 mice).

Figure 5.

RAD^Δ/Δ enhances cellular calcium handling. A, representative calcium transients from isolated ventricular cardiomyocytes loaded with fura2-AM, RAD^fl/fl (top) and RAD^Δ/Δ (bottom) paced at 1 Hz. B, amplitude of the transients from RAD^Δ/Δ are higher than in RAD^fl/fl. C, the velocity at which calcium enters the cytosol (upstroke of the transient) is faster in RAD^Δ/Δ than in RAD^fl/fl. D and E, calcium transient decay is faster in RAD^Δ/Δ than in RAD^fl/fl. Measurements in E used fluo4-AM and high speed 2-dimensional imaging. Measurements in A–D used fura2-AM, RAD^fl/fl (n = 10 mice, n = 38), RAD^Δ/Δ (n = 8 mice, n = 49). For E, RAD^fl/fl (n = 10 mice, n = 69) and RAD^Δ/Δ (n = 8 mice, n = 67). F, Western blot analysis demonstrates increased expression of SERCA2a, quantification is shown on the right (p = 0.035). Lanes were run on the same gel, but noncontiguous lanes are marked by the black line. Below, total protein stained with Coomasie Blue. RAD^fl/fl (7 mice) and RAD^Δ/Δ (8 mice) are shown. *, p = 0.014; **, p = 0.008; ****, p < 10⁻⁴.

Figure 6.

RAD^Δ/Δ enhances sarcomere shortening and increases the tension-integral. A, representative sarcomere length traces of RAD^fl/fl and RAD^Δ/Δ paced at 1 Hz. B, diastolic sarcomere length was not different between RAD^fl/fl and RAD^Δ/Δ. C, sarcomere fractional shortening was higher in RAD^Δ/Δ than in RAD^fl/fl. D, the velocity of shortening was faster in RAD^Δ/Δ than in RAD^fl/fl. E, integral of sarcomere length is larger for RAD^Δ/Δ than in RAD^fl/fl. RAD^fl/fl (n = 7 mice, n = 15 cells) and RAD^Δ/Δ (n = 8 mice, n = 24 cells) are shown. ***, p < 0.001; ****, p < 10⁻⁴.

Figure 7.

ISO does not alter RAD^Δ/Δ I_Ca,L. A, representative I_Ca,L traces at V_test of −5 mV before (black line) and after treatment with ISO (colored line; blue for Rad^fl/fl and red for RAD^Δ/Δ). Scale bars: 500 pA and 200 ms. B, current/voltage curves superimposing basal and acute 1 μm ISO treatment for RAD^fl/fl (B, i) and RAD^Δ/Δ (B, ii). C, fractional change of maximal conductance between baseline and ISO response of RAD^fl/fl (mean = 32.68%) and RAD^Δ/Δ (mean = 0.84%). D, conductance-voltage curve normalized to maximal conductance for ISO-treated cells demonstrates RAD^Δ/Δ cardiomyocytes have negative shifted activation midpoint relative to RAD^fl/fl in ISO. Smooth curves are Boltzmann distribution fitted to data. Steady-state availability was not different. E, difference between activation midpoints between baseline and ISO response of RAD^fl/fl (mean = −4.47) and RAD^Δ/Δ (mean = −0.19). RAD^fl/fl (n = 4 mice) and RAD^Δ/Δ (n = 4 mice) are shown. *. p = 0.02.

Figure 8.

β-AR–mediated responsiveness retained following RAD deletion in vivo. Echocardiography of mice scanned at baseline (filled symbols) and after injection of isoproterenol (open symbol with connecting line). Ejection fraction significantly increased with addition of isoproterenol in both RAD^fl/fl and RAD^Δ/Δ at 1- and 12-months after tamoxifen administration. RAD^fl/fl (n = 7) and RAD^Δ/Δ (n = 17) are shown. ****, p < 10⁻⁴.

Figure 9.

Cardiomyocyte cytosolic Ca²⁺ handling responds to β-AR–mediated activation in RAD^Δ/Δ. A, representative calcium transients from RAD^fl/fl and RAD^Δ/Δ cardiomyocytes paced at 1 Hz in 1 μm ISO. B, fractional change (ISO/baseline) of Ca²⁺ transient amplitude. C, upstroke velocity, and D, rate of decay. RAD^fl/fl (n = 10 mice, n = 38 cells) and RAD^Δ/Δ (n = 8 mice, n = 49 cells) are shown. E, Western blot analysis of total and phosphorylated PLN, RAD, and calsequestrin (15% SDS-PAGE). Lower panel, quantification shows no significant effect of myocardial Rad deletion. RAD^fl/fl (n = 6 mice) and RAD^Δ/Δ (n = 9 mice) are shown. F, acute ISO increased PLN Ser-16^P in RAD^Δ/Δ (4–20% SDS-PAGE). RAD^Δ/Δ (n = 5 mice) is shown. For all panels: *, p < 0.05, **. p = 0.005; ****, p < 10⁻⁴.

Figure 10.

Sarcomere dynamics and proteins respond to β-AR–mediated activation in RAD^Δ/Δ. A, representative sarcomere traces of RAD^fl/fl and RAD^Δ/Δ paced at 1 Hz after treatment with 1 μm ISO. B, fractional change (ISO/baseline) in fractional shortening, and C, relaxation velocity. ISO significantly increases shortening and relaxation velocity in RAD^Δ/Δ. D, integral of sarcomere length is larger in Rad^Δ/Δ than in RAD^fl/fl. E, ISO treatment increased the tension-integral index in RAD^fl/fl but not in RAD^Δ/Δ. RAD^fl/fl (n = 3 mice, n = 8 cells) and RAD^Δ/Δ (n = 5 mice, n = 16 cells) are shown. F, Coomassie stain and ProQ phosphostain analysis demonstrating increased phosphorylation of proteins from RAD^Δ/Δ hearts after treatment with ISO of cMyBPC and TnI, quantification to the right. Saline, n = 6 mice; and ISO, n = 6 mice were used. For all panels: *, p = 0.01; **, p = 0.0002; ****, p < 10⁻⁴.