Common Heart Failure With Preserved Ejection Fraction Animal Models Yield Disparate Myofibril Mechanics

Heart failure with preserved ejection fraction (HFpEF) is a complex, multiorgan syndrome. Cardiac manifestations include diastolic stiffening and impaired relaxation, normal resting systolic function but depressed systolic reserve, and modest hypertrophy. ¹ Although diastolic dysfunction remains a benchmark of HFpEF, the extent to which myofibrils, the contractile organelles of myocytes, contribute to this behavior remains unknown. HFpEF animal models historically emphasized hypertension and ventricular hypertrophy to achieve diastolic dysfunction, and recently have incorporated obesity and diabetes as they are increasingly prevalent. Popular rodent models include Zucker obese/spontaneously hypertensive rats ² and mice given a high‐fat diet (HFD) and the constitutive NO synthase inhibitor, Nω‐nitro‐l‐arginine methyl ester (ʟ‐NAME) (HFD+ʟ‐NAME). ³ However, neither model developed diastolic disease as severe as that observed in patients with HFpEF. Heightened diastolic pathology was achieved in larger Göttingen minipigs fed a HFD and treated with desoxycorticosterone acetate (DOCA) to induce volume retention/hypertension. ⁴ Although each model exhibited gross‐scale diastolic dysfunction, albeit to different extents, there are no data yet reported from myofibrils on their mechanical activation and relaxation properties. Thus, it remains unclear whether the mechanistic basis of global, organ‐level diastolic impairments observed among the models involves common underlying myofibrillar deficiencies. This has become salient as newer pharmaceuticals are targeting sarcomeric proteins to treat such diseases. Therefore, to test if shared defects in subcellular mechanics exist, and thereby potentially contribute to chamber‐level pathophysiology, we resolved the kinetic parameters of contraction and relaxation of individual myofibrils from each HFpEF animal model and its respective control.

The models, generated in independent laboratories, all had elevated ventricular diastolic filling pressure, normal‐range ejection fraction, myocardial hypertrophy and fibrosis, and obesity with glucose intolerance. ² , ³ , ⁴ Procedures followed were in accordance with institutional guidelines. Heart tissue from each model and control group was sectioned and frozen in liquid nitrogen. Frozen left ventricle tissue strips were incubated in a 4% triton skinning solution overnight at 4 °C and then homogenized to produce a concentrated suspension of myofibrils. Cell‐level measurements cannot time‐resolve relaxation kinetics, as even in single permeabilized cardiomyocytes, calcium diffusion is too slow to achieve this. However, given their small diameter, myofibrils promptly equilibrate with bathing solutions without significant diffusional constraints. ⁵ Individual myofibrils were electrostatically tethered between a glass probe connected to a Piezo‐length controller and a glass cantilever of known stiffness (0.031 N/m). Sarcomere length was set to 2.1 μm from an average of ≈1.7 μm. Cantilever displacement was measured by deflection of its shadow, incident on a photodiode, and converted into tension (force/cross‐sectional area). A double‐barreled pipette attached to a fast‐step motor permitted switching between relaxing and activating bathing solutions in <1 ms. Myofibrillar relaxation kinetics following sudden calcium removal are uniquely biphasic, and each phase can inform about deficits linked to regulatory and motor proteins. ⁵ Data supporting this study are available from the corresponding author on reasonable request.

Myofibrils from HFpEF minipigs exhibited no difference in resting tension (Figure [A] and [B]) but produced significantly less maximal active tension compared with control (Figure [A] and [C]). By contrast, resting and active tension generated by myofibrils from Zucker obese/spontaneously hypertensive rats and HFD+ʟ‐NAME mice were similar to controls, although myofibrillar activation was slower in the obese versus lean rats (Figure [D]). With respect to relaxation kinetics, we observed no differences in the duration or rate of the initial slow linear (Figure [A], [E], and [F]) or subsequent fast exponential relaxation phase (Figure [A] and [G]) in minipig myofibrils. Nevertheless, the initial linear relaxation phase was prolonged and slower in rodent HFpEF models versus controls, with Zucker obese/spontaneously hypertensive rat myofibrils additionally displaying a slower subsequent fast exponential‐decay rate. Collectively, these data show depressed systolic myofibrillar function only in the pig model and reduced relaxation rates exclusively in the rodent models. Shifts from fast α‐myosin to slower β‐myosin could partially explain differences in relaxation kinetics. ⁵ However, mass spectrometry analysis of the purified myofibrils revealed no significant increases in β‐myosin versus α‐myosin heavy chain protein content in any of the models, implying alternative underlying causes.

Figure . Kinetic and mechanical properties of myofibrils from heart failure with preserved ejection fraction (HFpEF) models.

A, Averaged myofibril tension over time (solid line) and SE (shaded region) for each model, aligned to the time of solution change (insets: dashed lines and arrows indicate the time of solution change; SP and FP highlight the initial, slow linear and subsequent fast, exponential phases of relaxation, respectively). B, Resting tension is steady‐state myofibril tension produced at pCa 8 (log₁₀[Ca²⁺]) (gray bars represent mean values). C, Active tension is the difference between steady‐state maximal tension produced at pCa 4 and the resting tension. D, Myofibrillar activation rate as calculated from a monoexponential fit. E, The slow linear phase of relaxation is mediated by thin filament deactivation and cross‐bridge detachment. ⁵ The start of the slow phase is provided by the trigger signal of the solution exchange motor, and its end is determined by the peak residual error of a progressive linear fit of the subsequent data. F, The rate of the slow linear phase is calculated as the slope of the slow phase, normalized to the active tension of the myofibril. G, The fast exponential phase is primarily mediated by passive sarcomeric elements and is independent of thin filament activation. ⁵ The rate of the fast phase is determined by a monoexponential fit of the data starting from the end of the slow phase. N=3 animals for each group, and n=5 to 9 myofibrils from each animal. Significance was assessed using a 2‐tailed Welch t‐test between controls and experimental groups of each model (P values are listed above each comparison; P<0.05 in bold). HFD indicates high‐fat diet; ʟ‐NAME, Nω‐nitro‐l‐arginine methyl ester; WT, wild type; and ZSF1, Zucker obese/spontaneously hypertensive rat.

Our findings demonstrate that myofibril mechanics vary significantly between HFpEF animal models, suggesting that some may uniquely recapitulate distinct aspects of the disorder and particular subphenotypes observed among patients. ¹ Although all 3 models had signs of diastolic dysfunction, impaired myofibril relaxation appears to contribute mechanistically to chamber‐level dysfunction, only in the rodent models. In the minipig model, because myofibril relaxation was unperturbed, global changes in diastole must involve other cellular‐ and tissue‐level pathologies. This discordance likely stems from physiological differences inherent among the hearts of small and large mammals and the unique regimens deployed to engender HFpEF‐like phenotypes. No single experimental system can perfectly mimic human disease, and this is especially true with HFpEF, because of its complex multiorgan pathology and distinct subphenotypes. These intricacies thus necessitate the judicious selection, or even combinatorial use, of distinct animal models to better simulate and understand various disease facets, to help refine therapeutic targeting, and to enhance the translational impact of preclinical HFpEF research.

Sources of Funding

These studies were supported by National Institutes of Health (NIH) grant T32HL007227 and an American Heart Association/DC Women's Board grant 22POS915659 to Dr Fenwick; American Heart Association grant 23PRE1026275 and NIH grant F31HL168850 to V. P. Jani; NIH grant HL164478 to Dr Foster; NIH grant AA029984 to Dr Sharp; NIH grant HL159428 to Dr Goodchild; NIH grant TL1TR003106 to J. E. Doiron; NIH grants HL146098, HL146514, and HL151398 to Dr Lefer; NIH grants HL128215, HL147933, HL155765, and HL164586 to Dr Hill; NIH grants R35HL135827, R35HL166565, and support from Cytokinetics to Dr Kass; and NIH grant HL124091 to Dr Cammarato.

Disclosures

Dr Lefer is a consultant for Sulfagenix. Dr Kass receives research support from Amgen related to human heart failure with preserved ejection fraction studies. The remaining authors have no disclosures to report.