CaMKII activity contributes to homeometric autoregulation of the heart: A novel mechanism for the Anrep effect

Abstract

Key points: The Anrep effect represents the alteration of left ventricular (LV) contractility to acutely enhanced afterload in a few seconds, thereby preserving stroke volume (SV) at constant preload. As a result of the missing preload stretch in our model, the Anrep effect differs from the slow force response and has a different mechanism. The Anrep effect demonstrated two different phases. First, the sudden increased afterload was momentary equilibrated by the enhanced LV contractility as a result of higher power strokes of strongly-bound myosin cross-bridges. Second, the slightly delayed recovery of SV is perhaps dependent on Ca²⁺ /calmodulin-dependent protein kinase II activation caused by oxidation and myofilament phosphorylation (cardiac myosin-binding protein-C, myosin light chain 2), maximizing the recruitment of available strongly-bound myosin cross-bridges. Short-lived oxidative stress might present a new facet of subcellular signalling with respect to cardiovascular regulation. Relevance for human physiology was demonstrated by echocardiography disclosing the Anrep effect in humans during handgrip exercise.

Abstract: The present study investigated whether oxidative stress and Ca²⁺ /calmodulin-dependent protein kinase II (CaMKII) activity are involved in triggering the Anrep effect. LV pressure-volume (PV) analyses of isolated, preload controlled working hearts were performed at two afterload levels (60 and 100 mmHg) in C57BL/6N wild-type (WT) and CaMKII-double knockout mice (DKO^CaMKII ). In snap-frozen WT hearts, force-pCa relationship, H₂ O₂ generation, CaMKII oxidation and phosphorylation of myofilament and Ca²⁺ handling proteins were assessed. Acutely raised afterload showed significantly increased wall stress, H₂ O₂ generation and LV contractility in the PV diagram with an initial decrease and recovery of stroke volume, whereas end-diastolic pressure and volume, as well as heart rate, remained constant. Afterload induced increase in LV contractility was blunted in DKO^CaMKII -hearts. Force development of single WT cardiomyocytes was greater with elevated afterload at submaximal Ca²⁺ concentration and associated with increases in CaMKII oxidation and phosphorylation of cardiac-myosin binding protein-C, myosin light chain and Ca²⁺ handling proteins. CaMKII activity is involved in the regulation of the Anrep effect and associates with stimulation of oxidative stress, presumably starting a cascade of CaMKII oxidation with downstream phosphorylation of myofilament and Ca²⁺ handling proteins. These mechanisms improve LV inotropy and preserve stroke volume within few seconds.

Keywords: Anrep effect; Ca2+ handling proteins; CaMKII; LV contractility; cMyBP-C; elastance-time curve; end-systolic elastance (Ees); myofilament phosphorylation.

Figure 1:. Analysis of contractility of the Anrep effect in humans

Representative echocardiographic-derived pressure-volume (PV) loops and end-systolic PV relationships (ESPVR) of healthy volunteers at baseline (white) and after sustained handgrip exercise (red) are shown (A). Data are shown as mean±SD. End-systolic elastance (E_es) is correlated with either effective arterial elastance (E_a; B) or heart rate (C). E_es (D), E_a (E), end-systolic volume at 100 mmHg (ESV 100 mmHg; F), stroke volume (G) and ejection time (H). Open circles are shown at baseline (white) and at exercise (red). **p<0.01, ***p<0.001, ****p<0.0001 vs. baseline; n=14

Figure 2:. Basic hemodynamics of the Anrep effect in humans

Data are shown as mean±SD. Systolic (A) and diastolic (B) blood pressure (BP), heart rate (C), ejection fraction (D), PPI (E), stroke work (F), EDV (G) and E/e’ (H) of human healthy volunteers are shown at baseline (white) and during exercise (red). *p<0.05, ***p<0.001, ****p<0.0001 vs. baseline; n=14

Figure 3:. Anrep protocol and groundbreaking hemodynamics in mice

Scheme of experimental protocol (A) of steady and intermediate states are demonstrated with time periods of change in afterload levels (60 and 100 mmHg and vice versa). Original registrations of left ventricular (LV) pressure (red coloured) and volume (green area) of the PV relation (B) are presented at low chart speed. Representative graphs of the beat to beat changes of dp/dt_max (C), stroke volume (D) and stroke energy (E) vs. time are shown in response to increased afterload. Arrows indicate start of sudden increase in afterload.

Figure 4:. PV analysis and systolic function in response to increased afterload

Original PV loops of multiple beat analysis (A) demonstrate ESPVR of WT mice during “aortic occlusion” (black, ESPVR 1) and “aortic release” (grey, ESPVR 2), comparing ESPVR from low to high (60–100 mmHg) and from high to low (100–60 mmHg) afterload levels, respectively. Representative PV loops (B) and corresponding single beat ESPVR measurements (identical coloured; C) of WT mice are shown at different steady and intermediate states. SV (D), E_es (E), dP/dt_max (F), ESV 100 mmHg (G), time to Emax (H) of WT mice c. Representative elastance vs. time graphs of WT mice are also demonstrated for SS1, IMA and SS2 states (H). Data are shown as mean±SD (panels D-H). *p<0.05, **p<0.01, ***p<0.001, **** p<0.0001 vs. SS1; ∇ p<0.05 IMA vs. SS2; p-value in graphs represents ANOVA test; n=10

Figure 5:. Energetics in response to increased afterload

Stroke work (A), V₀ (B), end-diastolic volume (C), ejection fraction (D), mechanical efficiency (E) and potential energy (F), total pressure-volume area (G) and wall stress (F) of WT mice are shown at different steady and intermediate states. *p<0.05, **p<0.01, ***p<0.001, **** p<0.0001 vs. SS1; ∇ p<0.05 IMA vs. SS2; p-value in graphs represents ANOVA test; n=10

Figure 6:. Diastolic function and protein kinase activity

Peak filling rate (A), dP/dt_min (B), end-diastolic pressure (C), tau (τ) (D), and coronary flow (E) (n=10). PKA- (F), PKG- (G) PKC-activity (H) of WT mice (n=6–7) are shown at different steady and intermediate states. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 vs. SS1; ∇ p<0.05 IMA vs. SS2; p-value in graphs represents ANOVA test.

Figure 7:. Stroke volume, oxidative stress parameters, and kinases expression and activity

Anrep effect in WT mice treated with (n=6; black circles) and without (n=7; white circles) streptomycin (80μM) indicated by SV at SS1, IMA and SS2 (A, left graph), Sustained stretch after baseline (BL=SS1, LVEDP= 8mmHg) with increased LVEDP to 15mmHg (STR) and slow force response after 5min of LVEDP of 15mmHg (SFR) with and without streptomycin treatment (A, right graph). H₂O₂ level in homogenate, H₂O₂ level in mitochondria and H₂O₂ level in cytosol (B) GSH concentration (C), expression, activity over time,, activity after 60min and oxidation of Ca²⁺/calmodulin-dependent protein kinase-II (CaMKII) (D-G), activities of myosin light chain kinase (MLCK; H) are given at different steady states (SS1, SS2 and SS3). Data are shown as mean±SD. * p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 vs. SS1 or BL in WT; ×p<0.05 SS3 vs. SS2 in WT; ††p<0.01, †††p<0.001, ††††p<0.0001 vs. SS1 or BL in WT with streptomycin; ‡p<0.05 SFR vs. STR in WT; #p<0.05 WT SFR vs. WT SFR with streptomycin; p-value in graphs represents ANOVA test.

Figure 8:. Phosphorylation of intracellular CaMKII targets

Heatmap of protein phosphorylation (P; A) of SRF serine (Ser)-103-P (B), calcineurin (CaN) Ser-197-P (C), HDAC4 Ser-467-P (D), H3 Ser-10-P (E), HSF1 Ser-230-P (F), DRP1 Ser-616-P (G), CREB-1 Ser-142-P (H), and ATF-1 Ser-63-P (I) are shown in steady state groups (SS1, SS2 and SS3). Insets indicate corresponding representative blots. Data are shown as mean±SD. *p<0.05, **p<0.01, ***p<0.001, ****p<0.0001 vs. SS1; p-value in graphs represents ANOVA test.

Figure 9:. Gene expression of proteins phosphorylated by CaMKII

Scheme of intracellular targets of CaMKII signaling (A). ATF-1: activating transcription factor-1; cMLC2: cardiac myosin light chain 2; cMyBP-C: cardiac myosin binding protein-C; cTnI: cardiac troponin I; CREB-1: cAMP-response element binding protein 1; DRP1: dynamin related protein 1; H3: histone 3; HDAC4: histone deacetylase 4; HSF1: heat shock factor 1; MAPK1: mitogen-activated protein kinase 1; NFAT: nuclear factor of activated T-cells; PLB: phospholamban; RyR2: ryanodine receptor 2; SERCA2a: sarcoplasmic/endoplasmic reticulum Ca2+ ATPase 2a; SRF: serum response factor. Heatmap of gene expression (B) and relative gene expression of CREB1 (C), NFATC1 (NFAT, cytoplasmic 1; D), HDAC4 (E), phospholamban (PLN; F), and MAPK1 (G) are shown in steady state groups (SS1, SS2 and SS3). **p<0.01 vs. SS1; Data are shown as mean±SD; p-value in graphs represents ANOVA test.

Figure 10:. Phosphorylation of Ca²⁺ handling proteins

Representative western immunoblotting in arbitrary units for Ca²⁺ handling proteins (A). Quantification of sarcoplasmic/endoplasmic reticulum Ca²⁺ ATPase 2a (Serca2A) expression (B), phospholamban (PLB) expression, (C), PLB serine (Ser)-16 and threonine (Thr)-17 phosphorylation (D and E, respectively), and ryanodine receptor 2 (RyR2) Ser-2814-P (G, F) is shown. Data are depicted as mean±SD. *p<0.05, ** p<0.01 vs. SS1; ×p<0.05 vs. SS2; p-value in graphs represents ANOVA test.

Figure 11:. Myofilament contractility and phosphorylation

Single cardiomyocyte force measurements (panels A-D) are represented at different steady states (n=16–24 cells/steady state from 3–5 hearts/steady state) Ca2+-activated tension vs. pCa relationships are shown (A). Normalized force vs. pCa relationships (B) with the corresponding pCa50 values (bars) indicate Ca2+-sensitivity of force production (C). Representative recordings are shown for the single exponential fit of rate constants of force redevelopment (ktr) (D). Bars indicate ktr at saturating [Ca²⁺] (at pCa 4.5; ktr,max). Phosphorylation of cMyBP-C Ser-282-P, cTnI Ser-23/24 and cMLC2 Ser-19-P are shown by Western immunoblotting (E, F and G, respectively). Insets indicate corresponding representative blots. Data are given as mean±SD. *p<0.05 SS2 vs. SS1; †p<0.05 SS3 vs. SS1; ‡ SS3 vs. SS2 (panels A-B). ** p<0.01, *** p<0.001, **** p<0.0001 vs. SS1; †p<0.001 vs. SS2; ††p<0.01, †††p<0.001, ††††p<0.0001 vs. SS2 (panels C-F); p-value in graphs represents ANOVA test followed with Bonferroni adjusted t-test and paired Student’s t-test for a subsequent incubation with GSH.

Figure 12:. Cardiomyocyte passive stiffness (Fpassive) and titin phosphorylation

Single cardiomyocyte passive force (F_passive) vs. sarcomere length relationships are depicted at different steady states (n=8–10 cells/steady state from 3 hearts/steady state) (A). Total (at Ser/Thr amino acid residues; B) and CaMKII-mediated site-specific (at Ser-4043; C) phosphorylation of titin is shown by Western immunoblotting. Insets indicate corresponding representative blots. F_passive at SS1 (D), SS2 (E) and SS3 (F) with and without GSH. Data are given as mean±SD. *p<0.05, **P<0.01 SS2 vs. SS1; †p<0.05 SS3 vs. SS1; ‡p<0.05 SS3 vs. SS2. p-value in graphs represents ANOVA test.

Figure 13:. Systolic function of DKOCaMKII at different afterload levels

E_es (A), end-systolic pressure (P_es; B), ESV 100 mmHg with changes from SS1 to SS2 as Δ_{ESV 100 mmHg} (C), stroke volume with changes from IMA to SS2 as Δ_{stroke volume} (D), dP/dt_max (E), and ejection fraction (F) of DKO^CaMKII mice (n=8, yellow) and corresponding wild type WL mice (grey, n=7) are shown at SS1, IMA and SS2. Data are given as mean±SD. **p<0.01, ***p<0.001, **** p< 0.0001 IMA/SS2 WL vs. SS1 WL; †p<0.05, †††p<0.001, ††††p<0.0001 IMA/SS2 DKO^CaMKII vs. SS1 DKO^CaMKII; ‡ p<0.05, ‡‡ p<0.01 DKO^CaMKII vs. WL; #p<0.01 DKO^CaMKII vs. WL;

Figure 14:. Elastance-time graphs and diastolic function of DKOCaMKII

Representative elastance vs. time graphs of wild type littermates (WL) (A) and DKO^CaMKII mice (B) are shown at SS1, IMA and SS2. Time to E_max (C), tau (τ) (D) and dP/dt_min (E) values of DKO^CaMKII mice (yellow, n=8) and corresponding wild type WL mice (grey, n=7) are given at SS1, IMA and SS2. Data are shown as mean±SD. ****p< 0.0001 IMA/SS2 WL vs. SS1 WL; †p<0.05, †††p<0.001, ††††p<0.0001 IMA/SS2 DKO^CaMKII vs. SS1 DKO^CaMKII; ‡ p<0.05 DKO^CaMKII vs. WL.