Enhanced efferocytosis of apoptotic cardiomyocytes through myeloid-epithelial-reproductive tyrosine kinase links acute inflammation resolution to cardiac repair after infarction

Abstract

Rationale: Efficient clearance of apoptotic cells (efferocytosis) is a prerequisite for inflammation resolution and tissue repair. After myocardial infarction, phagocytes are recruited to the heart and promote clearance of dying cardiomyocytes. The molecular mechanisms of efferocytosis of cardiomyocytes and in the myocardium are unknown. The injured heart provides a unique model to examine relationships between efferocytosis and subsequent inflammation resolution, tissue remodeling, and organ function.

Objective: We set out to identify mechanisms of dying cardiomyocyte engulfment by phagocytes and, for the first time, to assess the causal significance of disrupting efferocytosis during myocardial infarction.

Methods and results: In contrast to other apoptotic cell receptors, macrophage myeloid-epithelial-reproductive tyrosine kinase was necessary and sufficient for efferocytosis of cardiomyocytes ex vivo. In mice, Mertk was specifically induced in Ly6c(LO) myocardial phagocytes after experimental coronary occlusion. Mertk deficiency led to an accumulation of apoptotic cardiomyocytes, independently of changes in noncardiomyocytes, and a reduced index of in vivo efferocytosis. Importantly, suppressed efferocytosis preceded increases in myocardial infarct size and led to delayed inflammation resolution and reduced systolic performance. Reduced cardiac function was reproduced in chimeric mice deficient in bone marrow Mertk; reciprocal transplantation of Mertk(+/+) marrow into Mertk(-/-) mice corrected systolic dysfunction. Interestingly, an inactivated form of myeloid-epithelial-reproductive tyrosine kinase, known as solMER, was identified in infarcted myocardium, implicating a natural mechanism of myeloid-epithelial-reproductive tyrosine kinase inactivation after myocardial infarction.

Conclusions: These data collectively and directly link efferocytosis to wound healing in the heart and identify Mertk as a significant link between acute inflammation resolution and organ function.

Keywords: efferocytosis; inflammation; macrophages; myocardial infarction; phagocytosis.

Figure 1. Macrophages phagocytose cardiomyocytes (CMs) and Mertk is specifically required for CM efferocytosis

(A) Adult mouse CMs were isolated, fluorescently labeled (red), and induced to apoptosis. Dying CM apoptotic bodies were overlaid onto primary mouse MΦs and percent efferocytosis enumerated in Mertk+/+ and Mertk-/- MΦs. First image is a magnification of a MΦ ingesting a CM apoptotic body. In parallel, apoptotic Jurkat cells were co-cultivated with MΦs at equivalent ratios for efferocytosis quantitation. (B) Engulfment of CMs was measured after co-cultivating dying CMs with MΦs from CD36-/-, LRP deficient, or MER-/- primary MΦs. (C) Quantitation of efferocytosis after transfection of Mertk into Mertk-deficient HEK cells. MT=empty pIRES2 vector. (D) Bar graph showing TNFα mRNA expression by qPCR in Mertk+/+ vs Mertk-/- MΦs in the absence or after phagocytosis of CMs. Each column is mean± SEM. # indicates unpaired t test, p < 0.05. Asterisk indicates p < 0.05 relative to control (2-way ANOVA followed by Bonferroni post hoc test).

Figure 2. Identification and kinetics of Mertk expression post MI in experimental mice

(A) Semiquantitative RT-PCR of Mertk in heart extracts of Mertk+/+ and Mertk-/- mice 7days post occlusion (MI) of the left coronary artery. The first 2 lanes are from non surgically treated hearts. (B) Bar graph shows real-time PCR of Mertk mRNA in Left Ventricle (LV) and Right Ventricle (RV) at indicated days post MI. C = non-infarcted hearts. 7S = 7 days after sham infarction. Data were normalized to glyceraldehyde-3-phosphate dehydrogenase (Gapdh). (C) Bar graph of qPCR after laser capture microdissection of Mertk mRNA in remote (R) and ischemic (I) zones. (D) Immunoblot of MERTK protein levels in LV vs RV at indicated days (d) after MI. Densitometric analysis is plotted to the right. (E) Images show immunohisotchemical analysis of MERTK expression border zones of infarct. (F) Flow cytometric analysis was performed for Ly6c and MERTK levels on CD45-HI, CD11b-HI, Ly-6G-LO cells from myocardial extracts at indicated days before (PRE) vs. after MI. * indicates p < 0.05 after 2-way ANOVA followed by Bonferroni post hoc test. # indicates p < 0.05 after unpaired t test relative to control.

Figure 3. Chemokine and cytokine mRNA and inflammatory cells in hearts from Mertk+/+ vs. Mertk-/- mice

(A) Flow cytometric analysis of CD45-HI, CD11b-HI, Ly-6G-LO, Ly6c and CD11c cells in myocardium at indicated days (D) after MI. Enumeration in bar graph indicates percent monocytes that are Ly6c-HI vs. Ly6c-LO prior (P) or after MI at indicated days. (B) Time course (days = d) of chemokine mRNA expression in control and Mertk-/- infarcts. Measured were CCL2, TNF-α, IL-6, and IL-10. Data were obtained from 6 mice per genotype per timepoint. * indicates p < 0.05 relative to +/+ control.

Figure 4. Quantitation of cardiomyocyte (CM) apoptosis and association with CD68+ phagocytes in Mertk+/+ and Mertk-/- hearts post MI

(A) Photomicrographs of TUNEL nuclear (HOechst) staining in CMs (DESmin+). The arrows point to TUNEL (red) positive CM (green) nuclei (blue). TUNEL analysis is in the infarct border zone at indicated days (d) post MI. Bar = 200 micrometers. At 3days, bar graphs are quantitation of TUNEL positive CMs post MI relative to sham operated mice. At 5 days post MI, analysis is of Remote (R) and Border Zone (BZ) myocardium. 7day post MI analysis is within the BZ. (B) In situ efferocytosis analysis at 5 days post MI. Red phagocytes are CD68+. Phagocytes that also stain for CM Desmin are scored as efferocytosis as described in text. Bar = 100 micrometers. Similar findings were found with the alternative CM α-actinin. Asterisk = p < 0.05 relative to +/+ control.

Figure 5. Acute myocardial infarct size is increased in Mertk-/- mice post MI

Evan's Blue and TTC staining was performed in hearts at 3 (A) and 7 (B) days (d) post MI in Mertk+/+ and Mertk-/- mice as described in the Materials and Methods. AAR is Area At Risk and INF is Infarct. LV is left ventricle. * indicates p < 0.05 relative to +/+ control.

Figure 6. Quantitation of scar formation in remodeled hearts post myocardial infarction (MI) in Mertk deficient mice

(A) Representative photomicrographs of transverse cross-sections of trichrome-stained hearts from of control and Mertk-/- hearts at 28 days post MI. Picrosirus red staining was also performed and examined under brightfield and polarized light. (B) Quantitation of collagen levels after trichrome staining and PSR staining (polarized light) in the remote (R) and infarct (I) areas compared to non-infarcted control (c). (C) LV wall dimensions in Mertk+/+ and Mertk-/- hearts. * indicates p < 0.05 relative to +/+ control.

Figure 7. Assessment of heart function by echocardiography after myocardial infarction (MI) in Mertk deficient mice

(A) Echocardiography analysis of volumes and % FS and % EF. ESV is end systolic volume. EDV is end diastolic volume. EF is fractional shortening. FS is fractional shortening. Images are M-mode tracings. (B) Assessment of FS and EF after bone-marrow transplant of either Mertk+/+ or Mertk-/- myeloid cells post MI relative to non-surgically treated control (c) mice. * indicates p < 0.05 relative to +/+ control.

Figure 8. Identification of a soluble MER profile post MI

Immunoblot of myocardial extracts 5 days post MI in WT mice. Three separate mice (MI1, MI2, and MI3) were subjected to ligation of the left anterior descending artery. Subsequently, myocardial extracts were interrogated for immunoreactivity against a monoclonal anti-MER (ectodomain specific) antibody. MI3 was also treated with a glycanse prior to electrophoresis.