Background: Sterile inflammation following myocardial infarction leads to the depletion of circulating innate immune cells. Compensatory activation of bone marrow-derived emergency hematopoiesis (EH) maintains early inflammatory monocyte recruitment, vital for macrophage (Mφ )-driven efferocytosis. Unsupervised myelopoiesis however can aggravate cardiac inflammation and amount to maladaptive remodeling. The targeted reduction of hematopoietic activation may therefore present a novel approach to preserve cardiac function. All-trans retinoic acid (ATRA) is a pleiotropic modulator of EH and innate immunity shielding hematopoietic stem cells from activation and driving survival and differentiation of myeloid cells. 

Purpose: This study aimed to investigate the effects of ATRA on EH and its implications in cardiac wound healing.  

Methods: MI was induced by permanent coronary ligation in C57BL/6 mice and treated with daily injections of ATRA (30mg/kg) or DMSO as vehicle control up to five days, starting 24h after ligation. Flow cytometry (FACS) was used for cell cycle analysis of HSCs and immunophenotyping of leukocytes in bone marrow (BM), blood and heart. Cytokine expression was analyzed by qPCR in bulk infarct and sorted immune cells of bone marrow and cardiac origin. Echocardiography, Immunohistochemistry (IH) and masson trichrome (MT) staining evaluated the functional and histological infarct development

Results: On day 2 after MI, EH was significantly reduced in ATRA-treated mice as compared to vehicle controls by means of cell cycle activity and single cell proliferation potential of HSCs, as well as downstream myeloid cells in BM, blood and heart. Consequently, mRNA-expression of key inflammatory cytokines, IL-1βand TNFα, was diminished in the infarct tissue during this early phase. ATRA had no effect on BM mobilization of leukocytes nor on the expression of recruitment factors on cardiac endothelial cells. These early anti-inflammatory changes however failed to prevent LV-ventricular remodeling and to preserve cardiac function 21 days after MI. Interestingly, IH revealed a 2-fold increase in CD11b-positive myeloid cells in remote myocardium (RZ) that coincided with elevated expression of IL-6, TNFa and TGFb. Strikingly, MT-staining, performed 21 days after MI, demonstrated an almost 3-fold increase in collagen deposition in the RZ of ATRA treated mice in contrast to vehicle controls. As the possible culprit, bulk RNA-sequencing revealed dense expression of retinoic acid receptors on murine cardiac monocytes and Mφ. By qPCR, non-canonical activation of recruited ATRA-primed monocyte-derived Mφ, was found to propagate a pro-inflammatory phenotype with higher expression of MMP2 and MMP9 in sorted cardiac Mφ. Furthermore, prominent IL-1βand IL-6 expression in M2-polarized BM-derived Mφ indicated an impaired anti-inflammatory phenotype after ATRA treatment. 

Conclusion: Despite a beneficial reduction of EH after MI, short-term treatment with ATRA induced profound and persisting changes in the cytokine expression of monocytes and Mφ of BM and cardiac origin, with increased inflammatory and fibrotic activity. This prevented improvements in cardiac function due to myeloid and fibrotic expansion in remote myocardium. Our data emphasize the importance of the individual inflammatory behavior of immune cells participating in cardiac wound healing to enhance scar care and cardiac outcome after MI.