Splenic Ly6Chi monocytes contribute to adverse late post-ischemic left ventricular remodeling in heme oxygenase-1 deficient mice

Abstract

Heme oxygenase-1 (Hmox1) is a stress-inducible protein crucial in heme catabolism. The end products of its enzymatic activity possess anti-oxidative, anti-apoptotic and anti-inflammatory properties. Cardioprotective effects of Hmox1 were demonstrated in experimental models of myocardial infarction (MI). Nevertheless, its importance in timely resolution of post-ischemic inflammation remains incompletely understood. The aim of this study was to determine the role of Hmox1 in the monocyte/macrophage-mediated cardiac remodeling in a mouse model of MI. Hmox1 knockout (Hmox1^-/-) and wild-type (WT, Hmox1^+/+) mice were subjected to a permanent ligation of the left anterior descending coronary artery. Significantly lower incidence of left ventricle (LV) free wall rupture was noted between 3rd and 5th day after MI in Hmox1^-/- mice resulting in their better overall survival. Then, starting from 7th until 21st day post-MI a more potent deterioration of LV function was observed in Hmox1^-/- than in the surviving Hmox1^+/+ mice. This was accompanied by higher numbers of Ly6C^hi monocytes in peripheral blood, as well as higher expression of monocyte chemoattractant protein-1 and adhesion molecules in the hearts of MI-operated Hmox1^-/- mice. Consequently, a greater post-MI monocyte-derived myocardial macrophage infiltration was noted in Hmox1-deficient individuals. Splenectomy decreased the numbers of circulating inflammatory Ly6C^hi monocytes in blood, reduced the numbers of proinflammatory cardiac macrophages and significantly improved the post-MI LV function in Hmox1^-/- mice. In conclusion, Hmox1 deficiency has divergent consequences in MI. On the one hand, it improves early post-MI survival by decreasing the occurrence of cardiac rupture. Afterwards, however, the hearts of Hmox1-deficient mice undergo adverse late LV remodeling due to overactive and prolonged post-ischemic inflammatory response. We identified spleen as an important source of these cardiovascular complications in Hmox1^-/- mice.

Keywords: Cardiac rupture; Heme oxygenase-1; Macrophages; Monocytes; Myocardial infarction.

Fig. 1

Lower early post-MI survival in WT than in Hmox1-deficient mice. a Flow cytometric detection of cells labeled with pimonidazole hydrochloride in the heart 90 min after LAD ligation (n = 4 mice/group). b ELISA for cTnI in plasma one day after LAD ligation (sham: n = 26 mice/group, MI: n = 36 mice/group). c Kaplan–Meier survival curves of Hmox1^+/+ and Hmox1^−/− mice subjected to sham and MI surgeries (sham: n = 18 mice/group, MI: n = 32–37 mice/group). *p < 0.05, **p < 0.01, ***p < 0.0001; *vs. appropriate sham control, ^#vs. other MI. d–f Photographs of an autopsied mouse that died due to cardiac rupture. d The chest cavity was filled with a large amount of clotted blood. e, f Photographs of the heart with a left ventricular perforation in the area of apex. MI myocardial infarction, LAD left anterior descending coronary artery, LVFWR left ventricular free wall rupture. g Western blot analysis of collagen type I in heart on day 4 after surgery

Fig. 2

More profound LV dysfunction in Hmox1^−/− than in Hmox1^+/+ mice after MI. At indicated time-points after MI or sham surgery a LV EF, b LV FS, c LV Vs, d LV Vd, e LV IDs, and f LV IDd were determined with TTE; n = 7–10 mice/group. g, h Evaluation of cardiomyocyte hypertrophy. g Representative fluorescent images of WGA (red) and DAPI (blue) in the peri-infarct (border zone) and remote myocardium (remote zone) of MI-operated Hmox1^+/+ and Hmox1^−/− mice. h Graph summarizing quantitative analysis of cardiomyocyte cross-sectional area (CSA) in the border and remote myocardium.*p < 0.05, **p < 0.01, ***p < 0.0001, ^†vs. appropriate sham control, ^#vs. other MI. Scale bar 50 μm

Fig. 3

Increased steady state and post-MI numbers of monocytes in the absence of Hmox1. a Schematic representation of different populations of blood monocytes based on the expression of CD43 and Ly6C markers. Flow cytometric analysis of b classical monocytes (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺⁺ CD43⁺), c intermediate monocytes (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺⁺ CD43⁺⁺) and d non-classical monocytes (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺ CD43⁺⁺) in the peripheral blood of mice after LAD ligation or sham surgery (n = 4–9 mice per group). Data represented as a number of cells per 1 µl of peripheral blood. *p < 0.05, **p < 0.01, ***p < 0.0001

Fig. 4

Expression of genes involved in inflammatory cell infiltration inversely correlates with Hmox1 expression. The qPCR for a Hmox1, b Mcp1, c Icam1, d Esel, e Vcam1 in the infarcted area (or corresponding area in sham-operated and intact controls) of the heart at indicated time points after surgery (n = 3–6 mice/group). *p < 0.05, **p < 0.01, ***p < 0.0001; *vs. appropriate sham control, ^†vs. appropriate intact control, ^#vs. other MI

Fig. 5

Lack of Hmox1 affects composition of macrophages in ischemic cardiac muscle. a Representative fluorescent images of CD11b (green) and 4′,6-diamidino-2-phenylindole (DAPI; blue) in the peri-infarct region of MI-operated mice on day 21 after surgery. Scale bar 50 μm. b Schematic representation of different populations of cardiac macrophages and monocytes based on the expression of MHC II and Ly6C markers. Flow cytometric analysis of CD45⁺ Ly6G⁻ CD11b⁺ c monocytes (MHC-II^lo Ly6C⁺⁺) and different subsets of macrophages: d MHC-II⁺ Ly6C⁺⁺, e MHC-II⁺ Ly6C⁺, f MHC-II⁺⁺ Ly6C⁺, g MHC-II⁺ Ly6C⁺⁺ CD11c⁺, h MHC-II⁺ Ly6C⁺ CD11c⁺, i MHC-II⁺⁺ Ly6C⁺ CD11c⁺ in the myocardium of mice after LAD ligation or sham surgery (n = 3–6 mice/group) in indicated time points. Data represented as a number of cells detected in heart. *p < 0.05, **p < 0.01, ***p < 0.0001

Fig. 6

Lack of Hmox1 is associated with higher steady state and post-MI numbers of selected hematopoietic stem and progenitor cell populations in bone marrow. Flow cytometric analysis of a SKL cells (CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻), b long-term HSC (LT-HSC; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁻ CD48⁻ CD150⁺), c short-term HSC (ST-HSC; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁺ CD48⁻ CD150⁺), d hematopoietic progenitors (HPC; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁺ CD48⁻ CD150⁻), e multipotent hematopoietic progenitors (MPP; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁺ CD48⁺ CD150⁻), f progenitor cells lacking Sca-1 (KL; CD45⁺ Sca-1⁻ c-kit⁺ Lin⁻), g granulocyte-monocyte progenitor cells (GMP; CD45⁺ Sca-1⁻ c-kit⁺ Lin⁻ CD34⁺ CD48⁺⁺ CD150⁻) in the bone marrow of mice after LAD ligation or sham surgery (n = 4–10 mice/group). Data represented as a number of cells in bone marrow isolated from tibiae and femora. *p < 0.05, **p < 0.01, ***p < 0.0001

Fig. 7

Increased steady state and decreased late post-MI monocyte numbers in spleens of Hmox1^−/− mice. Flow cytometric analysis of a classical (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺⁺ CD43⁺), b intermediate (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺⁺ CD43⁺⁺) and c non-classical (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺ CD43⁺⁺) monocytes in the spleens of mice after LAD ligation or sham surgery (n = 4–10 mice/group). Data represented as a number of cells per spleen. *p < 0.05, **p < 0.01, ***p < 0.0001

Fig. 8

Lower numbers of selected hematopoietic stem and progenitor cell populations in spleens of Hmox1^−/− mice 21 days after MI. Flow cytometric analysis of a SKL cells (CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻), b long-term HSC (LT-HSC; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁻ CD48⁻ CD150⁺), c short-term HSC (ST-HSC; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁺ CD48⁻ CD150⁺), d hematopoietic progenitors (HPC; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁺ CD48⁻ CD150⁻), e multipotent hematopoietic progenitors (MPP; CD45⁺ Sca-1⁺ c-kit⁺ Lin⁻ CD34⁺ CD48⁺ CD150⁻), f progenitor cells lacking Sca-1 (KL; CD45⁺ Sca-1⁻ c-kit⁺ Lin⁻) and g granulocyte-monocyte progenitor cells (GMP; CD45⁺ Sca-1⁻ c-kit⁺ Lin⁻ CD34⁺ CD48⁺⁺ CD150⁻) in the spleens of mice after LAD ligation or sham surgery (5–10 mice/group). Data represented as a number of cells per spleen. *p < 0.05, **p < 0.01, ***p < 0.0001

Fig. 9

Splenectomy changes the post-MI patterns of monocytes/macrophages and LV function in Hmox1^+/+ and Hmox1^−/− mice. LV EF in splenectomized vs. non-splenectomized a Hmox1^+/+ and b Hmox1^−/− mice monitored for 21 days post-sham or -MI surgery. Flow cytometric analysis of c classical (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺⁺ CD43⁺), d intermediate (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺⁺ CD43⁺⁺) and e non-classical (CD45⁺ CD11b⁺ Ly6G⁻ NK1.1⁻ Ly6C⁺ CD43⁺⁺) monocytes in the blood. Subsets of cardiac macrophages: f MHC-II⁺ Ly6C⁺⁺, g MHC-II⁺ Ly6C⁺, h MHC-II⁺⁺ Ly6C⁺, i MHC-II⁺ Ly6C⁺⁺ CD11c⁺, j MHC-II⁺ Ly6C⁺ CD11c⁺, k MHC-II⁺⁺ Ly6C⁺ CD11c⁺ at 21 days after LAD ligation or sham surgery (n = 3–9 mice/group). Data represented as c–e a number of cells per 1 µl of peripheral blood and f–k a number of cells detected in heart. *p < 0.05, **p < 0.01, ***p < 0.0001. ^#vs. corresponding cardiac surgery in non-splenectomized mice