Effect of captopril on post-infarction remodelling visualized by light sheet microscopy and echocardiography

Abstract

Angiotensin converting enzyme inhibitors, among them captopril, improve survival following myocardial infarction (MI). The mechanisms of captopril action remain inadequately understood due to its diverse effects on multiple signalling pathways at different time periods following MI. Here we aimed to establish the role of captopril in late-stage post-MI remodelling. Left anterior descending artery (LAD) ligation or sham surgery was carried out in male C57BL/6J mice. Seven days post-surgery LAD ligated mice were allocated to daily vehicle or captopril treatment continued over four weeks. To provide comprehensive characterization of the changes in mouse heart following MI a 3D light sheet imaging method was established together with automated image analysis workflow. The combination of echocardiography and light sheet imaging enabled to assess cardiac function and the underlying morphological changes. We show that delayed captopril treatment does not affect infarct size but prevents left ventricle dilation and hypertrophy, resulting in improved ejection fraction. Quantification of lectin perfused blood vessels showed improved vascular density in the infarct border zone in captopril treated mice in comparison to vehicle dosed control mice. These results validate the applicability of combined echocardiographic and light sheet assessment of drug mode of action in preclinical cardiovascular research.

Figure 1

Echocardiographic evaluation of cardiac function and remodelling in a mouse model of myocardial infarction. (a) Schematic study outline. (b) Ejection fraction (EF) at the time of inclusion in week 0. (c) EF after 4 weeks of dosing with either vehicle or captopril. (d) EF change over study period (week 4-week 0)/week 4*100%. (e) Left ventricle internal diameter in diastole (LVIDd) at the time of inclusion in week 0. (f) LVIDd at week 4. (g) LVIDd change over study period. (h) Left ventricle anterior wall dimension/thickness (LVAW) in diastole and (i) systole. (j) Left ventricle posterior wall dimension/thickness (LVPW) in diastole and (k) systole. Data is presented as mean ± s.e.m. n = 10–12. One-way ANOVA with Tukey’s post hoc test. Significance: *p < 0.05, **p < 0.01, ***p < 0.001. #: Significant (p < 0.05) after removal of single non-responder in LAD—Captopril. LAD: left anterior descending artery ligation.

Figure 2

Light sheet imaging and automated analysis of myocardial infarction. (a) Representative examples of light sheet imaged hearts. 3D overview image is shown on the left. Second panel from left—digital 2D section is taken from 3D reconstructed heart (long axis horizontal plane). Third panel from left demonstrates deep learning-based segmentation of the left ventricle (LV) chamber (in yellow). (b) Diastolic LV chamber volume quantified by deep-learning based analysis. (c) LV apex volume (quantified below plane fitted through the center of deep learning segmented chamber). (d) Light sheet quantification of LV chamber volume in diastole correlates with echocardiography-based LV end-diastolic volume (r = 0.73, p < 0.0001). Data is presented as mean ± s.e.m., n = 8–10. One-way ANOVA with Tukey’s post hoc test. Significance: **p < 0.01, ***p < 0.001. Scale bars: 1 mm. LAD: left anterior descending artery ligation.

Figure 3

Light sheet analysis of ventricular wall dilation and infarct size. (a) Schematic illustration of apical conicity index calculation in Sham—vehicle heart. The index represents the ratio between the fitted triangle and striped area. (b) Same as above but shown for LAD—vehicle heart. Left ventricle (LV) wall dilation leads to increased area outside the fitted triangle. (c) Statistical comparison of cardiac conicity index. (d) Ejection fraction (EF) at week 4 shows negative correlation to conicity index (r = -0.76; p < 0.0001). (e) Representative example of light sheet imaged heart with infarct zone separated by line. (f) Segmented volume of cardiac wall with thickness below 400 µm (from sample shown in E). (g) Quantification of infarcted LV wall volume with thickness below 400 µm. Significance: *p < 0.05, **p < 0.01, ***p < 0.001. Scale bars: 1 mm. LAD: left anterior descending artery ligation.

Figure 4

Light sheet imaging of left ventricle chamber topology. (a) Surface rendering of sham-operated light sheet imaged heart. (b) Magnified image demonstrates high optical resolution, intact papillary muscles (Pm) and dense distribution of trabeculae carneae (Tc). (c) Surface rendering of LAD-ligated light sheet imaged heart. (d) Magnified image demonstrates MI-induced changes in left ventricle (LV) chamber. Loss of trabeculae carneae (Tc) is evident in the dilated infarct area (Inf). (e) LV chamber surface is color-coded, with blue colours denoting concave surface and red convex surface areas (ant.—anterior view; post.—posterior view). Green indicates smooth surfaces. Outward dilation of the LV chamber is visible in LAD ligated hearts. Indentations left by papillary muscles are indicated by arrows. Lack of curvature is visible in the posterior view of the LAD—vehicle heart, with papillary muscle and trabecular indentations absent from the apex. Ejection fraction (%) is shown for the selected hearts performed 5 weeks after surgery. (f) Quantitative analysis of LV chamber mean curvature. Analysis was carried out in the lower apical half the LV chamber and the group mean value is shown on the frequency distribution histogram. Data is presented as mean ± s.e.m., n = 8–10. Scale bars: (a,c) 1 mm; (b,d) 300 µm.

Figure 5

Captopril improves vascularization in the border zone. (a) 2D view of lectin perfused vasculature in the left ventricle (LV) wall. (b) Automatically segmented blood vessels in the same image. (c) Lectin perfused vasculature and overlay with automated detection (in white) of capillaries and arterioles. (d) 3D view of lectin perfused vasculature in the LV anterior wall (LV aw, view from the inside of the LV chamber towards the anterior wall). (e) Automated analysis of lectin signal in the same 3D image stack. (f) Wall thickness map corresponding to the same area. Thicker wall (> 500 µm) is shown in red, border zone with thickness between 300–500 µm in yellow and LV wall with thickness < 300 µm in blue. (g-h) Magnified view from the anterior side of LV. (i) 2D section through the LV wall, demonstrating wall thickness categories. (j) There is no difference in vascular density in the infarct zone (wall thickness < 300 µm) between vehicle and Captopril treated groups. (k) Captopril treatment improves vascular density in the border zone with wall thickness 300–500 µm. (l) No difference is visible between the groups in cardiac tissue with wall thickness above 500 µm. (m) Vascular density in posterior heart wall (no differences were observed between the study groups). One-way ANOVA with Tukey’s post hoc test. Significance: *p < 0.05. Data is presented as mean ± s.e.m., n = 8–9. Scale bars: (a–c), 100 µm; (d–i), 500 µm. LAD: left anterior descending artery ligation.