Natural Cardiac Regeneration Conserves Native Biaxial Left Ventricular Biomechanics After Myocardial Infarction in Neonatal Rats

Abstract

After myocardial infarction (MI), adult mammals exhibit scar formation, adverse left ventricular (LV) remodeling, LV stiffening, and impaired contractility, ultimately resulting in heart failure. Neonatal mammals, however, are capable of natural heart regeneration after MI. We hypothesized that neonatal cardiac regeneration conserves native biaxial LV mechanics after MI. Wistar rat neonates (1 day old, n=46) and adults (8–10 weeks old, n=20) underwent sham surgery or permanent left anterior descending coronary artery ligation. At 6 weeks after neonatal MI, Masson’s trichrome staining revealed negligible fibrosis. Echocardiography for the neonatal MI (n=15) and sham rats (n=14) revealed no differences in LV wall thickness or chamber diameter, and both groups had normal ejection fraction (72.7% vs 77.5%, respectively, p=0.1946). Biaxial tensile testing revealed similar stress-strain curves along both the circumferential and longitudinal axes across a full range of physiologic stresses and strains. The circumferential modulus (267.9 kPa vs 274.2 kPa, p=0.7847), longitudinal modulus (269.3 kPa vs 277.1 kPa, p=0.7435), and maximum shear stress (3.30 kPa vs 3.95 kPa, p=0.5418) did not differ significantly between the neonatal MI and sham groups, respectively. In contrast, transmural scars were observed at 4 weeks after adult MI. Adult MI hearts (n=7) exhibited profound LV wall thinning (p<0.0001), chamber dilation (p=0.0246), and LV dysfunction (ejection fraction 45.4% vs 79.7%, p<0.0001) compared to adult sham hearts (n=7). Adult MI hearts were significantly stiffer than adult sham hearts in both the circumferential (321.5 kPa vs 180.0 kPa, p=0.0111) and longitudinal axes (315.4 kPa vs 172.3 kPa, p=0.0173), and also exhibited greater maximum shear stress (14.87 kPa vs 3.23 kPa, p=0.0162). Our study is the first to show that native biaxial LV mechanics are conserved after neonatal heart regeneration following MI, thus adding biomechanical support for the therapeutic potential of cardiac regeneration in the treatment of ischemic heart disease.

1. INTRODUCTION

Ischemic heart disease currently afflicts 130 million people around the world and continues to be a leading cause of death for mankind, even as treatments for acute myocardial infarction (MI) have significantly improved over time (Virani et al., 2020). The biomechanical properties of the injured left ventricle (LV) play an essential role in determining long-term pump function after recovery from MI (Holmes et al., 2005). In adult mammals, including humans, MI causes massive cardiomyocyte death acutely, leading to LV wall thinning as necrotic cells are gradually replaced with collagen scar (Gupta et al., 1994; Lerman et al., 1983). Fibrotic stiffening of the LV wall reduces bulging of the akinetic infarcted region during each contraction, thereby attempting to maintain systolic function but at the expense of diastolic filling (Bogen et al., 1980; Parmley et al., 1973). Ultimately, the LV chamber dilates to sustain adequate stroke volume, but as LV wall stress rises according to Laplace’s Law, a positive-feedback loop is triggered, leading to further LV dilatation and finally pump failure (Pfeffer and Braunwald, 1990). Therapies that preserve the native biomechanics of the LV to prevent the onset of ischemic cardiomyopathy after MI therefore represent a significant unmet clinical need.

While adult mammalian hearts show very little baseline potential for cardiomyocyte proliferation (Bergmann et al., 2009; Senyo et al., 2013), neonatal mammalian hearts are capable of robust natural cardiac regeneration after MI. In particular, our team recently developed a neonatal rat MI model and showed that rats which experienced MI on postnatal day 1 (P1) gradually recovered nearly all lost myocardium over the course of three weeks, with complete normalization of LV function and geometry (Wang et al., 2020). Similar observations have also been reported previously in P1 mice (Das et al., 2019; Haubner et al., 2012; Porrello et al., 2013), P1 rabbits (Wang et al., 2021), P1 piglets (Ye et al., 2018; Zhu et al., 2018), and possibly human newborns (Haubner et al., 2016), in addition to adult zebrafish (Poss et al., 2002). These results collectively suggest that a conserved mechanism for natural heart regeneration may exist, which if reactivated in adult human patients after MI could provide a means toward circumventing the development of ischemic cardiomyopathy.

In order for natural heart regeneration to provide a durable long-term solution in preventing heart failure after MI, the biomechanical properties of the regenerated LV muscle must be understood. Using micropipette aspiration, one study previously reported that adult zebrafish hearts stiffened immediately after MI via cryoinjury, but recovered normal biomechanical properties within 5 weeks as the myocardium regenerated (Yu et al., 2018). Our team recently also showed that natural heart regeneration after MI in neonatal P1 mice preserves LV biomechanical properties using lenticular hydrostatic deformation testing (Wang et al., 2020). Whether myocardial biomechanics in both the circumferential and longitudinal axes are preserved by natural heart regeneration after MI in neonatal mammals, however, remains unknown.

Here, we performed planar biaxial tensile testing of passive, regenerated rat LV myocardium. We hypothesized that natural heart regeneration preserves native biaxial LV biomechanics at 6 weeks after MI in neonatal rats.

2. MATERIALS AND METHODS

2.1. Animal Care and Biosafety

Adult Wistar rats, including both pregnant females and healthy males, were obtained from Charles River Laboratories (Wilmington, MA, USA). Monitoring of pregnant mothers for parturition was performed every 12 hours at minimum. Neonatal pups were kept with their nursing mothers until weaned at 21 days old. Food and water were otherwise provided ad libitum. All animal experiments were performed in concordance with the Guide for the Care and Use of Laboratory Animals (United States National Institutes of Health, 8th Edition, 2011). All animal procedures were approved by the Institutional Animal Care and Use Committee at Stanford University (Protocol 28921). A flowchart illustrating the usage of animals in this study is provided in Supplemental Figure 1.

2.2. Neonatal Rat Myocardial Infarction

Surgery was performed for neonatal P1 rats, including both males and females, at 12–24 hours after observed parturition, as previously described (Wang et al., 2020). The sex of each individual neonate was not recorded. Briefly, P1 pups (n=46) were cooled with topical ice for 7 minutes to induce hypothermic circulatory arrest (Phifer and Terry, 1986). The anesthetized pup was then transferred to the operating table in supine position. Under microscope guidance, a fourth-interspace left anterior thoracotomy was performed, and the left anterior descending (LAD) coronary artery was permanently ligated 1 mm below the left atrial appendage using a 6–0 polypropylene suture. Sham surgery was performed in an identical manner, including hypothermic circulatory arrest and passage of the needle under the LAD artery, but the suture was not tied. Chest and skin closure were performed in layers using interrupted 6–0 polypropylene sutures. The pups spontaneously recovered on a 37 °C warm plate and were returned to their mother’s care after all surgeries for the litter were completed.

2.3. Adult Rat Myocardial Infarction

Surgery was performed for adult male rats (n=20, 8–10 weeks old, 270–330 g) as previously described (Stapleton et al., 2019; Steele et al., 2020; Wang et al., 2019). No adult female rats were used in any experiments. The rats were anesthetized using 3% inhaled isoflurane (Fluriso, VetOne, Boise ID, USA) and then intubated with a 16 G angiocatheter. The rat was placed in right lateral decubitus position on the operating table, and continuous 1–2% isoflurane was delivered during mechanical ventilation (Harvard Apparatus VentElite, Holliston, MA, USA) for maintenance of anesthesia. A fourth-interspace left lateral thoracotomy was performed, and the LAD artery was permanently ligated 1–2 mm below the left atrial appendage using a 6–0 polypropylene suture. Sham surgery was performed in an identical manner, including passage of the needle under the LAD artery, but the suture was not tied. Chest and skin closure were performed in layers using interrupted 4–0 and 6–0 polypropylene sutures, respectively. Finally, isoflurane was weaned, and the rat was extubated and recovered.

2.4. Confirmation of Neonatal Myocardial Infarction

A subset of neonatal rats (n=11) was allocated for confirmation of MI. One hour after fully recovering from LAD ligation surgery, these P1 rats were deeply anesthetized using 4% inhaled isoflurane and euthanized by decapitation. Median sternotomy was performed and 50 μL fluorescent-labeled lectin (0.33 mg/mL, DyLight 649-labeled Lycopersicon esculentum tomato lectin, Vector Laboratories, Burlingame, CA, USA) was injected into the right ventricle. The heart was then explanted. In the dark, the hearts were incubated for 45 minutes in 4% paraformaldehyde at 4°C, washed 3 times with phosphate-buffered saline (PBS), and finally placed in Vectashield (Vector Laboratories, Cat: H-1000) for 1–2 hours at room temperature before storing at −20°C until use. The whole hearts were placed on a double-concave microscope slide (Sail brand, Cat: 7104) with the anterior wall facing up. The heart was then covered and flattened with a microscope coverslip (Fisherbrand, Cat: 12545F, Pittsburgh, PA, USA). Images were acquired using an inverted Zeiss LSM 780 multiphoton laser scanning confocal microscope (Zeiss, Oberkochen, Germany). Scans of the whole heart were performed using 10X and 20X objectives. Images were captured, processed, and analyzed using Zeiss Zen software.

2.5. Masson’s Trichrome Staining and Infarct Size Analysis

A subset of neonatal rats (n=6) and adult rats (n=6) were allocated for histological infarct analysis. At 6 weeks post-surgery for neonatal rats and 4 weeks post-surgery for adult rats, these animals were deeply anesthetized using 4% inhaled isoflurane and euthanized by cervical dislocation. A median sternotomy was performed, and the heart was arrested by injection of potassium chloride (1 mEq/kg) into the right ventricle. The hearts were explanted, flushed with PBS, filled with optimum cutting temperature compound (OCT, Fisher Healthcare, Cat: 23730571, Houston, TX, USA), frozen in OCT using 2-methyl butane on dry ice, and stored at −80 °C until use. Hearts were sectioned along the LV short-axis plane with 10 μm thickness. For neonatal rats, three sections were selected per heart, spanning the level of the ligation suture to the apex. For adult rats, one representative section was selected per heart at the level of the mid-lower papillary muscles. In preparation for Masson’s trichrome staining (American MasterTech, Cat: KTMTR2PT, Lodi, CA, USA), the samples were thawed at 4 °C for 10 min, then rehydrated with ethanol and fixed in Bouin’s solution (Polysciences Inc., Cat: 16045–1, Warrington, PA, USA) for 105 minutes at 60 °C. The samples were washed with water until the tissue cleared, and then incubated with hematoxylin solution for 5 minutes, and Scarlet-Acid Fuchsin for 15 minutes. The samples were washed again, incubated with phosphotungstic-phosphomolybdic acid solution for 10 minutes, and stained with Aniline Blue for 5 minutes. Finally, the samples were incubated with 1% acetic acid for 5 minutes between washes, and then dehydrated and mounted in Cytoseal (Thermo Fisher Scientific, Cat: 8310–16, Waltham, MA, USA). The completed slides were imaged using an EVOS XL Core Imaging System (Thermo Fisher Scientific, Cat: AMEX-1000). Infarct size was analyzed using ImageJ software (Schneider et al., 2012) by determining the percentage of blue fibrotic area relative to the entire LV area for each section. For neonatal hearts, the average percentage infarct across all three sections examined is reported.

2.6. Echocardiography

A subset of neonatal rats (n=29) and adult rats (n=14) were allocated for echocardiography followed by biomechanical testing. Transthoracic echocardiography was performed at 6 weeks after surgery for neonatal rats and at 4 weeks after surgery for adult rats. The animals were anesthetized using 3% inhaled isoflurane, and then transferred to the imaging station in supine position. Anesthesia was maintained during imaging using 1–2% isoflurane via nosecone. B- and M-mode images of the LV in left parasternal short- and long-axis views were acquired using a Vevo 2100 Imaging Station equipped with a MicroScan MS250 13–24 MHz transducer (VisualSonics Inc., Toronto, Canada). Image analysis was performed using Vevo LAB software, including measurements of LV wall thickness in diastole over the anterolateral LAD territory (LVWTd), LV internal diameter in systole (LVIDs) and diastole (LVIDd), and LV ejection fraction (LVEF) at the level of the mid-low papillary muscles. Echocardiographic imaging and analysis were each performed by a single individual to reduce interobserver bias.

2.7. Planar Biaxial Tensile Testing

After echocardiography at 6 weeks post-surgery for neonatal rats (n=29) and 4 weeks post-surgery for adult rats (n=14), these animals were deeply anesthetized using 4% inhaled isoflurane. A median sternotomy was performed, and the heart was arrested in diastole by injection of University of Wisconsin (UW) cold cardioplegic storage solution (Bridge to Life Ltd., Columbia, SC, USA) into the right ventricle. The heart was explanted, and the LV was isolated and prepared for biaxial testing as previously described (Wang et al., 2019). Briefly, the right ventricle and interventricular septum were first removed. Next, the LV wall was unfurled circumferentially, and the lateral LV region, representing non-LAD territory, was excised. The final sample represented an approximately 10 mm x 10 mm square block of anterolateral LV wall supplied by the LAD artery. To avoid contracture after explant, all samples were stored in UW solution at 4°C until testing, which was within 6 hours of explant in all cases.

Planar biaxial tensile testing was performed using a displacement-controlled biaxial load frame, as previously described (Azadani et al., 2012; Mookhoek et al., 2017; Wang et al., 2019). Each of the four edges of the LV sample was mounted onto the corresponding arm of the biaxial stretcher using two hooked 5–0 polypropylene sutures per side, ensuring that the circumferential and longitudinal axes of the LV sample were aligned with the principal axes of deformation. At the center of each LV sample, five black marker beads (250–355 μm, MO-SCI Corporation, Rolla, MO, USA) were positioned in a 4 mm x 4 mm square, without the use of fixative agents. With the sample in a fully relaxed state, the load cells (model 31/3672–02, 1000 g; Honeywell Sensotec, Columbus, OH, USA) for both axes of the stretcher were zeroed. Tissue deformation during stretching was monitored using a charge-coupled device camera (model TM 9701; 300 frames/sec, 0.1 pixels/mm; Pulnix, Sunnyvale, CA, USA) positioned orthogonally above the sample. The same biaxial displacement-controlled protocol was employed for all samples. First, the pre-load was set to 0.05 N in both axes. Next, each sample underwent 20 preconditioning cycles of 4% strain, followed immediately by three consecutive test cycles up to 40% peak strain, all performed at 1% strain/sec. This testing protocol captures the full range of physiologic strains observed in rat hearts (Weytjens et al., 2008). The force imparted on the LV tissue along each principal direction during stretching was recorded by the respective load cell. The strain experienced by the LV tissue along each principal direction was calculated from the relative movement of the marker beads using LabVIEW (National Instruments, Austin, TX, USA). For all samples, only data from the final stretching cycle was used for analysis.

2.8. Constitutive Modeling and Data Analysis

The experimental Cauchy stress ($T^{exp}$) in the circumferential (θ) and longitudinal (L) directions was calculated by equations (1a) and (1b) below. F represents the force reported by the load cells, t represents the tissue thickness determined by echocardiography (LVWTd), and λ represents the ratio of tissue length during deformation to that at rest (l).

$$T_{\theta \theta }^{exp}={\lambda }_{\theta }\frac{F_{\theta }}{t{l}_L}$$ (1a)

$$T_{LL}^{exp}={\lambda }_L\frac{F_L}{t{l}_{\theta }}$$ (1b)

Green strain (𝐸) was computed using the equations below:

$$E_{\theta \theta }=\frac{1}{2}\left({\lambda }_{\theta }^2-1\right)$$ (2a)

$$E_{LL}=\frac{1}{2}\left({\lambda }_L^2-1\right)$$ (2b)

The LV tissue was assumed to be an anisotropic, incompressible, non-linear hyperelastic material. As such, the biomechanical properties of the LV samples were examined using a four-parameter Fung model. The Fung strain energy function, 𝑊, was described by the equations below, where $c,c_{\theta \theta },c_{\theta L}$, and $c_{LL}$ represent the Fung model parameters.

$$W(Q)=\frac{1}{2}c\left(e^Q-1\right)$$ (3a)

$$Q(E)=c_{\theta \theta }{E}_{\theta \theta }^2+2c_{\theta L}{E}_{\theta \theta }{E}_{LL}+c_{LL}{E}_{LL}^2$$ (3b)

Using the Fung model, the theoretical Cauchy stress ($T^{model}$) was defined via the following equations:

$$T_{\theta \theta }^{model}={\lambda }_{\theta }^{2}c{e}^Q\left(c_{\theta \theta }{E}_{\theta \theta }+c_{\theta L}{E}_{LL}\right)$$ (4a)

$$T_{LL}^{model}={\lambda }_L^{2}c{e}^Q\left(c_{\theta L}{E}_{\theta \theta }+c_{LL}{E}_{LL}\right)$$ (4b)

A non-linear least-squares regression algorithm was used to mathematically fit the experimental stress data for each sample to equations (4a) and (4b), thereby yielding the Fung model parameters for each individual sample. In both the circumferential and longitudinal directions, the average Fung model stress-strain curve for each experimental group was generated as the average of the fitted curves among the individual samples. All computations were performed using MATLAB version 9.4 (MathWorks, Natick, MA, USA).

Using the stress-strain curves derived from the Fung model for each individual sample, the tangent modulus along each principal direction was calculated for each sample at a stress level of 19.1 kPa, representing the average peak wall stress of the rat LV at 4 weeks after MI (Ferferieva et al., 2018). In addition, using a physiologic strain level of 20% in each principal direction (Espe et al., 2017), the maximum shear stress ($T_{max}$) of each LV sample was approximated using the formula below, derived from Mohr’s circle (Dally and Riley, 1991).

$$T_{max}=\frac{\left|T_{\theta \theta }^{exp}-T_{LL}^{exp}\right|}{2}$$ (5)

While the stress level of 19.1 kPa and the strain level of 20% may serve as representative physiologic values, in order to conservatively encompass the full range of possible wall stresses and strains experienced by the neonatal and adult rat heart, the tangent modulus was further evaluated for wall stresses up to 50.0 kPa and the maximum shear stress was further evaluated for strains up to 40%.

Finally, the degree of anisotropy for each LV sample was calculated as shown in equation (6) below, where a material is isotropic if $c_{\theta \theta }=c_{LL}$ and Anisotropy = 1; a material is anisotropic with stiffer behavior in the circumferential direction if $c_{\theta \theta }>c_{LL}$ and Anisotropy < 1; and a material is anisotropic with stiffer behavior in the longitudinal direction if $c_{\theta \theta }<c_{LL}$ and Anisotropy > 1.

$$Anisotropy=\frac{c_{LL}+c_{\theta L}}{c_{\theta \theta }+c_{\theta L}}$$ (6)

2.9. Statistical Analysis

After surgery was completed, all investigators were blinded to the experimental group assignments until data collection was finalized. All statistical analyses were performed using Stata version 14.2 (StataCorp LLC., College Station, TX, USA). Continuous variables were reported as mean ± standard error and compared using two-sample t-tests. P-values less than 0.05 were considered statistically significant. Experimental data and Fung model data from our biaxial tensile testing experiments are available at: https://purl.stanford.edu/pg377kj4997.

3. RESULTS

3.1. LAD Ligation Generates Large Acute Myocardial Infarcts in P1 Rats

A subset of P1 rats in the sham group (n=5) and MI group (n=6) were sacrificed on the day of surgery to confirm successful LAD ligation via fluorescent lectin perfusion. Using whole-heart confocal microscopy, the entire anterior LV surface of the sham hearts was bright with fluorescent lectin signal and the LAD was visibly intact (Figure 1A). The LAD territory of MI hearts, however, was devoid of lectin signal distal to the ligation suture for all 6 hearts examined (Figure 1B).

Figure 1. Confirmation of Neonatal Myocardial Infarction.

Fluorescent lectin perfusion was used to confirm successful left anterior descending (LAD) coronary artery ligation immediately after surgery on postnatal day 1 (P1). Representative whole-mount confocal microscopy images of the anterior left ventricle surface are shown. (A) For sham hearts (n=5), fluorescent lectin signal is observed across the entire heart surface and the LAD is visibly intact (white arrow). (B) For hearts in the myocardial infarction group (MI, n=6), the LAD territory distal to the ligation suture (yellow arrow) is devoid of lectin signal (dotted blue line), confirming successful LAD ligation.

3.2. Minimal Scar Formation after P1 Myocardial Infarction

A subset of P1 rats in the sham group (n=3) and MI group (n=3) were sacrificed at 6 weeks after surgery to assess for collagen scar formation using Masson’s trichrome staining. Averaging across three levels from below the ligation suture to the apex, we observed no difference in collagen area relative to total LV area between the sham and MI groups (0.62±0.07% vs 0.61±0.06%, p=0.9466, Figure 2A–C).

Figure 2. Scar Assessment After Myocardial Infarction.

Masson’s trichrome staining was performed at 6 weeks after surgery for postnatal day 1 (P1) rats, and at 4 weeks after surgery for adult rats. Negligible fibrosis was observed for (A) the P1 sham (n=3) and (B) the P1 MI groups (n=3), with (C) no significant difference in collagen area relative to total left ventricular area. Compared to (D) the adult sham hearts (n=3), the (E) adult MI hearts (n=3) exhibited transmural infarcts, with (F) significantly greater relative collagen area. Scale bars represent 5 mm. *** indicates p<0.001.

A subset of adult rats in the sham group (n=3) and MI group (n=3) were sacrificed at 4 weeks after surgery. Whereas the hearts in the sham group exhibited minimal relative collagen area (0.75±0.10% of LV area), the hearts in the MI group all showed prominent fibrotic remodeling (33.39±2.90% of LV area, p=0.0004, Figure 2D–F).

3.3. Cardiac Function and Geometry are Preserved after P1 Myocardial Infarction

All P1 rats which were planned for biomechanical testing underwent echocardiography at 6 weeks after surgery. As shown in Table 1, no difference in cardiac geometry was observed between the sham group (n=14) and MI group (n=15), including LV wall thickness (LVWTd 1.34 mm vs 1.36 mm, p=0.4398) and LV chamber size (LVIDs 3.12 mm vs 3.56 mm, p=0.2137; LVIDd 5.92 mm vs 6.25 mm, p=0.3838). Cardiac function was also maintained at a normal level at 6 weeks after P1 MI (LVEF 72.7%), with no difference compared to that of P1 sham rats (LVEF 77.5%, p=0.1946).

Table 1.

Echocardiographic Assessment of Cardiac Geometry and Function. Echocardiography was performed 6 weeks after surgery for postnatal day 1 (P1) rats and 4 weeks after surgery for adult rats. LVIDd, left ventricle internal diameter in diastole; LVIDs, left ventricle internal diameter in systole; LVWTd, left ventricle wall thickness in diastole; MI, myocardial infarction.

P1 Rats	P1 Sham (n=14)	P1 MI (n=15)	P-value
LVWTd (mm)	1.34 ± 0.02	1.36 ± 0.02	0.4398
LVIDs (mm)	3.12 ± 0.27	3.56 ± 0.22	0.2137
LVIDd (mm)	5.92 ± 0.25	6.25 ± 0.27	0.3838
Ejection Fraction (%)	77.5 ± 3.0	72.7 ± 2.0	0.1946
Adult Rats	Adult Sham (n=7)	Adult MI (n=7)	P-value
LVWTd (mm)	1.85 ± 0.05	1.20 ± 0.08	<0.0001
LVIDs (mm)	3.20 ± 0.15	5.83 ± 0.41	0.0001
LVIDd (mm)	6.37 ± 0.31	7.61 ± 0.37	0.0246
Ejection Fraction (%)	79.7 ± 0.8	45.4 ± 3.9	<0.0001

All adult rats which were planned for biomechanical testing underwent echocardiography at 4 weeks after surgery. Compared to adult sham rats (n=7), the adult MI rats (n=7) exhibited significant LV wall thinning (LVWTd 1.85 mm vs 1.20 mm, p<0.0001), contractile impairment (LVIDs 3.20 mm vs 5.83 mm, p=0.0001), and chamber dilation (LVIDd 6.37 mm vs 7.61 mm, p=0.0246), as well as profoundly depressed cardiac function (LVEF 79.7% vs 45.4%, p<0.0001).

3.4. Biaxial Left Ventricular Biomechanics are Preserved after P1 Myocardial Infarction

Experimental stress-strain data plots for the P1 sham (n=14) and P1 MI rats (n=15) are shown in Supplemental Figure 2. The average stress-strain curves derived from the individual Fung model curves are presented in Figure 3A–B, encompassing the full range of LV wall stresses experienced by healthy and infarcted rat hearts (Ferferieva et al., 2018; Fomovsky et al., 2011). Along both principal directions, the stress response to strain was non-linear and there was substantial overlap of the P1 sham and P1 MI curves, indicating a similar biomechanical response between the two groups. Indeed, the average Fung model parameters for the P1 rats, including the circumferential ($c_{\theta \theta }$), longitudinal ($c_{LL}$), and coupling ($c_{\theta L}$) coefficients, were all similar between the P1 sham and MI groups (Table 2). The degree of anisotropy was also similar (P1 sham 0.78 vs P1 MI 0.80, p=0.8600). At a representative post-MI LV wall stress level of 19.1 kPa (Ferferieva et al., 2018), the tangent modulus for the P1 sham and P1 MI groups were similar in both the circumferential (274.2 kPa vs 267.9 kPa, p=0.7847) and longitudinal directions (277.1 kPa vs 269.3 kPa, p=0.7435, Figure 3C). At a representative LV strain level of 20% (Espe et al., 2017), the maximum shear stress for the P1 sham and P1 MI hearts was similar (3.95 kPa vs 3.30 kPa, p=0.5418, Figure 3D). Indeed, at all physiologically relevant wall stresses (up to 50.0 kPa) and strains examined (up to 40%), there was no significant difference between the P1 sham and P1 MI hearts in terms of tangent modulus or maximum shear stress, respectively.

Figure 3. Left Ventricular Biomechanics After Neonatal Rat Surgery.

Biomechanics data derived from the Fung model at 6 weeks after either sham surgery (n=14) or myocardial infarction (MI, n=15) for postnatal day 1 (P1) rats. Average Fung model stress-strain curves are shown for (A) the circumferential axis, and (B) the longitudinal axis, with shaded region representing standard error. No significant difference in (C) the tangent modulus at 19.1 kPa stress, and (D) the maximum shear stress at 20% strain, was observed between the neonatal sham and MI groups.

Table 2.

Left Ventricular Biomechanical Properties After Neonatal Rat Surgery. Fung model parameters $\left(c,c_{\theta \theta },c_{\theta L},c_{LL}\right)$ and biomechanical properties were determined at 6 weeks after either sham surgery or myocardial infarction (MI) for postnatal day 1 (P1) rats.

P1 Biomechanical Properties	P1 Sham (n=14)	P1 MI (n=15)	P-value
$c$	4.40 ± 0.95	4.00 ± 0.55	0.7167
$c_{\theta \theta }$	6.05 ± 0.81	5.65 ± 0.70	0.7064
$c_{\theta L}$	5.48 ± 0.57	5.31 ± 0.70	0.8510
$c_{LL}$	3.06 ± 0.55	3.63 ± 0.88	0.5950
Anisotropy	0.78 ± 0.08	0.80 ± 0.07	0.8600
Circumferential Modulus (kPa)	274.2 ± 14.0	267.9 ± 17.7	0.7847
Longitudinal Modulus (kPa)	277.1 ± 15.5	269.3 ± 17.6	0.7435
Maximum Shear Stress (kPa)	3.95 ± 0.64	3.30 ± 0.83	0.5418

Experimental stress-strain data plots for the adult sham (n=7) and adult MI rats (n=7) are shown in Supplemental Figure 3. The average stress-strain curves derived from the individual Fung model curves are presented in Figure 4A–B. Along both principal directions, the adult MI curves exhibited a markedly stiffer behavior than the adult sham curves, which was reflected by significant differences in the average Fung model parameters $c_{\theta \theta }$ (p=0.0320) and $c_{\theta L}$ (p=0.0341), as well as near-significant differences among the other parameters (Table 3). Although there was no difference in anisotropy between the two groups (adult sham 0.73 vs adult MI 0.72, p=0.9357), the tangent modulus at 19.1 kPa was significantly greater for the adult MI group than the adult sham group in both the circumferential (180.0 kPa vs 321.5 kPa, p=0.0111) and longitudinal directions (172.3 kPa vs 315.4 kPa, p=0.0173, Figure 4C). Finally, at 20% strain, the maximum shear stress for the adult MI group was significantly greater than that for the adult sham group (3.23 kPa vs 14.87 kPa, p=0.0162, Figure 4D).

Figure 4. Left Ventricular Biomechanics After Adult Rat Surgery.

Biomechanics data derived from the Fung model at 4 weeks after either sham surgery (n=7) or myocardial infarction (MI, n=7) for adult rats. Average Fung model stress-strain curves are shown for (A) the circumferential axis, and (B) the longitudinal axis, with shaded region representing standard error. A significantly greater (C) tangent modulus at 19.1 kPa stress, and (D) maximum shear stress at 20% strain, was observed for the adult MI group. * indicates p<0.05.

Table 3.

Left Ventricular Biomechanical Properties After Adult Rat Surgery. Fung model parameters $\left(c,c_{\theta \theta },c_{\theta L},c_{LL}\right)$ and biomechanical properties were determined at 4 weeks after either sham surgery or myocardial infarction (MI) for adult rats.

Adult Biomechanical Properties	Adult Sham (n=7)	Adult MI (n=7)	P-value
$c$	17.03 ± 4.07	7.20 ± 1.95	0.0502
$c_{\theta \theta }$	2.19 ± 0.29	9.07 ± 2.83	0.0320
$c_{\theta L}$	1.81 ± 0.36	8.50 ± 2.77	0.0341
$c_{LL}$	1.04 ± 0.14	4.01 ± 1.62	0.0935
Anisotropy	0.73 ± 0.08	0.72 ± 0.08	0.9357
Circumferential Modulus (kPa)	180.0 ± 5.4	321.5 ± 46.9	0.0111
Longitudinal Modulus (kPa)	172.3 ± 7.6	315.4 ± 51.3	0.0173
Maximum Shear Stress (kPa)	3.23 ± 0.48	14.87 ± 4.14	0.0162

4. DISCUSSION

Research aimed at improving the treatment of coronary artery disease and MI is actively being pursued from multiple angles, including pharmacologic strategies (Li et al., 2017), revascularization techniques (Goldstone et al., 2018), tissue engineering approaches (von Bornstädt et al., 2018; Wanjare et al., 2019), cell therapy (Menasché, 2018), cytokine-mediated angiogenesis (Steele et al., 2020), and natural heart regeneration (Lam and Sadek, 2018). Because myocardial biomechanics profoundly influence the fate of long-term LV function and geometry after MI (Holmes et al., 2005), our team has continuously explored how new therapies designed to stimulate myocardial repair affect the biomechanical properties of the recovering LV, including after epicardial implantation of engineered tissues (Atluri et al., 2013; Shudo et al., 2015), and after intramyocardial injection of angiogenic cytokines (Hiesinger et al., 2012; MacArthur et al., 2013; Trubelja et al., 2014; Wang et al., 2019). These therapies may not eliminate all collagen scar formation after MI, but still improve or potentially even normalize LV biomechanics, albeit to varying degrees. Modulus values from biomechanical tensile testing of rat LV myocardium after MI, with or without therapeutic intervention, are summarized in Table 4. While healthy adult rat LV myocardium exhibits modulus values in the 140–220 kPa range, infarcted adult rat LV myocardium is much stiffer, with modulus values in the 250–400 kPa range, or even greater. Therapies such as angiogenic cytokines and tissue-engineered implants have been reported to maintain the modulus value of infarcted rat LV myocardium in the 200–250 kPa range, reflecting attenuation of LV biomechanical derangements after severe ischemic injury.

Table 4.

Summary of Biomechanical Tensile Testing Experiments for Rat Left Ventricular Myocardium After Infarction. Studies reporting rat left ventricle (LV) tensile modulus are summarized, including type of tensile testing and type of therapy after myocardial infarction (MI), if any. Our current work represents the first study of rat LV biomechanics after MI in postnatal day 1 (P1) neonatal rats. C, circumferential; L, longitudinal. ~ denotes data values estimated from the source manuscript.

Reference	Tissue/Therapy	Method	Modulus
Current Work	P1 neonatal rat LV 6 weeks after MI (with natural regeneration) Adult rat LV 4 weeks after MI No therapy	Biaxial tensile testing	Neonate Sham: C: 274 kPa L: 277 kPa Neonate MI (with natural regeneration): C: 268 kPa L: 269 kPa Adult Sham: C: 180 kPa L: 172 kPa Adult MI: C: 322 kPa L: 315 kPa
Wang et al., 2019	Adult rat LV 4 weeks after MI ± angiogenic cytokine therapy	Biaxial tensile testing	Sham: C: 220 kPa L: 209 kPa MI: C: 264 kPa L: 258 kPa MI + therapy: C: 212 kPa L: 195 kPa
Sirry et al., 2018	Adult rat LV 4 weeks after MI No therapy	Biaxial tensile testing	MI: C: 1218 kPa L: 487 kPa
Shudo et al., 2015	Adult rat LV 4 weeks after MI ± cell sheet therapy	Uniaxial tensile testing	Sham: ~140 kPa MI: ~400 kPa MI + therapy: ~250 kPa
Atluri et al., 2013	Adult rat LV 6 weeks after MI ± cell scaffold therapy	Uniaxial tensile testing	Sham: 193 kPa MI: 304 kPa MI + therapy: 239 kPa
MacArthur et al., 2013	Adult rat LV 4 weeks after MI ± angiogenic cytokine therapy	Uniaxial tensile testing	Sham: 195 kPa MI: 301 kPa MI + therapy: 252 kPa

In this study, we investigated for the first time the biaxial mechanical properties of naturally regenerated LV myocardium in a mammalian model. We observed that, at 6 weeks after MI in newborn rats, natural heart regeneration resulted in negligible scar formation and preserved LV biomechanics in both the circumferential and longitudinal directions, such that both the histological and mechanical properties of the regenerated myocardium were indistinguishable from that of sham controls. While our results support the findings of previous biomechanical studies performed using micropipette aspiration in an adult zebrafish model of heart regeneration after cryoinjury (Yu et al., 2018) and using lenticular hydrostatic deformation testing in a neonatal mouse model of heart regeneration after coronary ligation (Wang et al., 2020), our current study utilized planar biaxial tensile testing, which remains the gold standard for the study of myocardial biomechanics (Voorhees and Han, 2015). While other testing methods such as micropipette aspiration and lenticular hydrostatic deformation testing cannot account for the intrinsic anisotropy of the LV muscle (Demer and Yin, 1983; Gupta et al., 1994; Omens et al., 1993), biaxial testing is able to deform the sample along two orthogonal axes at once, thereby assessing mechanical properties in two dimensions and resolving any coupled behavior due to anisotropy. Thus, biaxial testing permits a more complete biomechanical characterization than other testing methods.

Our team previously demonstrated that neonatal rats recover normal cardiac function and geometry by 3 weeks after MI, following a period of robust cardiomyocyte cell cycle activation and proliferation (Wang et al., 2020). Here, we followed neonatal rats to 6 weeks after MI and again observed no evidence of significant adverse LV remodeling or decline in cardiac contractility on echocardiography, and no evidence of significant scar formation histologically. In contrast, within only 4 weeks after MI in adult rats, the process of adverse remodeling and fibrotic scar formation led to significant LV tissue stiffening as expected. These results support the durability of natural heart regeneration as a potential therapy for ischemic heart failure after MI, as the process of chronic LV remodeling is determined collectively by infarct size, infarct healing, and LV wall stress (Pfeffer and Braunwald, 1990). In particular, previous reports have suggested that infarct size and collagen content, specifically, are the primary determinants of LV function and pump mechanics after MI (Fomovsky and Holmes, 2010; Sirry et al., 2016). It is therefore interesting to note that, in this study, we observed complete preservation of native LV biomechanical properties with natural heart regeneration which yielded no significant scar formation at 6 weeks after neonatal rat MI, while in a previous study we observed preservation of native LV biomechanics with angiogenic therapy which reduced but did not eliminate scar formation at 4 weeks after adult rat MI (Wang et al., 2019). These results suggest that there may be an infarct size threshold below which LV biomechanics are not significantly impacted. Whether the implications of such a threshold can apply to long-term LV function, however, remains unknown, and additional studies are required to determine whether a slow, chronic remodeling process ensues even if the early impact of a small infarct on LV mechanics is minimal. Similarly, although natural heart regeneration after neonatal MI produces little to no scar, long-term biomechanical analyses are needed to determine the long-term durability of this intriguing phenomenon.

Interestingly, while our results suggest that neonatal heart regeneration influences tissue biomechanics, other studies have demonstrated both in vitro and in vivo that matrix stiffness influences neonatal cardiomyocyte proliferative potential (Notari et al., 2018; Yahalom-Ronen et al., 2015). Specifically, these studies indicate that neonatal cardiomyocyte proliferation may be inhibited when cells are attempting to grow within a stiff extracellular matrix, and that cardiac regeneration may be potentiated if the cells are allowed to grow within a more compliant environment. This interconnected relationship between cardiomyocyte proliferation, heart function, and myocardial biomechanics after MI is further complicated by the fact that artificial stiffening of infarcted myocardium using injectable biomaterials or implantable patches represents another promising therapy for preserving LV function after MI (Fomovsky et al., 2012; Morita et al., 2011). When cardiac regenerative therapies are ultimately applied in the setting of adult MI, a careful biomechanical balance may be necessary to optimize cell proliferation and functional performance if fibrotic scar is present. The complete avoidance of scar formation after neonatal heart regeneration may circumvent this dilemma, thereby increasing the attractiveness of natural cardiac regeneration as a future therapy for adult MI.

Biaxial tensile testing is not without limitations. During stretching, the samples are subjected to a non-homogeneous stress profile in the areas around the loading hooks, leaving only the central area of the sample with a relatively homogeneous stress state. Small samples, including rat myocardial samples, may therefore be difficult to test using biaxial stretching. To ensure accurate data, we followed the methods described by a previous team that used finite element simulations to validate stress uniformity when performing biaxial testing of similarly-prepared rat myocardial samples after MI (10 mm x 10 mm samples, two loading hooks per edge) (Fomovsky and Holmes, 2010). By 6 weeks after P1 MI, we found that our myocardial samples were large enough for accurate biaxial testing based on these finite element simulation results. However, due to size limitations, we were not able to perform biaxial testing at significantly earlier timepoints after neonatal rat surgery. Similarly, because our samples likely have varying muscle fiber orientation throughout their thickness, we would expect some variation in the deformation and strain field throughout the sample thickness. As a result, while we utilized a Mohr’s circle-based approximation of maximum shear stress, calculation of the full stress tensor would provide a more precise assessment. Another limitation of our biaxial testing setup is the inability to perform non-equibiaxial testing, which would have provided a greater depth of insight regarding tissue anisotropy. Nevertheless, based on the results of our Fung model analysis, we preliminarily determined that the degree of anisotropy of regenerated LV myocardium after neonatal MI was similar to that of sham controls. Finally, our study utilized only adult male rats, and although both male and female neonatal rats were included, the sex of each neonate was not recorded. Our previous study of natural heart regeneration in neonatal rats demonstrated no difference between males and females in the hemodynamic, functional, structural, and histological recovery of the regenerating heart after P1 MI, which suggests that the biomechanical recovery may be similar for both sexes as well, although our current study was not designed to address this question.

Overall, our study is the first to show that native biaxial LV mechanical properties are conserved after neonatal heart regeneration. Our results contribute new biomechanical support for the therapeutic potential of natural cardiac regeneration for the treatment of ischemic heart disease. Further studies are needed to evaluate the long-term biomechanics of regenerated LV myocardium.

Supplementary Material

4

Supplemental Figure 1. Study Flowchart of Animal Usage. A flowchart is provided, illustrating the usage of postnatal day 1 (P1) rats and adult rats in this study. LAD, left anterior descending coronary artery; MI, myocardial infarction.

5

Supplemental Figure 2. Experimental Stress-Strain Data Plots After Neonatal Rat Surgery. At 6 weeks after neonatal rat surgery, experimental stress-strain data were collected along the circumferential (A-B) and longitudinal axes (C-D) for both the sham (n=14) and myocardial infarction (MI, n=15) groups.

6

Supplemental Figure 3. Experimental Stress-Strain Data Plots After Adult Rat Surgery. At 4 weeks after adult rat surgery, experimental stress-strain data were collected along the circumferential (A-B) and longitudinal axes (C-D) for both the sham (n=7) and myocardial infarction (MI, n=7) groups.