Ischemic damage elicits repair responses that provoke the removal of cell debris, recruit immune cells, and initiate angiogenesis. Macrophages play a central part in the reparative response. Following heart damage, macrophages actively participate in clearing cell debris and promote inflammatory signals to provoke angiogenesis. In the reparative phase, macrophages are shifted toward anti-inflammatory action to induce stable scar formation. These timely activations of macrophages are crucial for compensative and reparative process. However, excessive inflammation results in adverse remodeling with cardiac dysfunction. Inflammation and adverse outcomes are known to be linked in cardiovascular disease. However, the results of clinical trials targeting cardiovascular inflammation have been disappointing. For instance, the representative anti-inflammatory therapy CANTOS trial with the monoclonal antibody canakinumab targeting interleukin-1β (IL-1β) reduced plasma levels of IL-6 and CRP and the endpoint of cardiovascular death. On the other hand, canakinumab did not enter the cardiovascular therapeutic arena because of the potential risk for fatal infection. Failure of anti-inflammatory therapy may be related to a lack of deeper understanding of the pathophysiological roles of inflammation dynamics with multiple phenotypes of immune cells.
IKKs have canonical components (IKKα, β, and γ) and an alternative kinase (IKKε) according to their structural features. IKKα is involved in the termination of NF-κB-dependent inflammation signaling. RelA and c-Rel are phosphorylated by IKKα and then are cleared from target gene promoters to result in negative regulation of inflammatory macrophages. When IKKα function is impaired, IKKα fails to terminate macrophage activation, thus increasing local inflammation20. IKKβ suppresses the inflammatory macrophage phenotype to prevent over-exuberant activation of macrophages during inflammation21. IKKγ may exert a protective function in the parenchymal compartment during the pathogenesis of pancreatitis. In chronic pancreatitis, deletion of IKKγ in epithelial cells aggravates inflammation and fibrosis and delays recovery22.
The role of IKKε is reported to be diverse in the disease context. In obese mice, IKKε deficiency prevents the initial macrophage inflammatory response to a high-fat diet4. In an aortic banding model, IKKε deficiency increases cardiac hypertrophy and fibrosis11. In ApoE KO mice, IKKε deficiency inhibits high-fat diet-induced inflammation and obesity23. High-fat diet-induced atherosclerosis in ApoE KO mice is reduced in IKKε KO mice5. Also in ApoE KO mice, ablation of IKKε enhances and prolongs NLRP3 inflammasome priming and metaflammation. IKKε-deficient BMDMs and peritoneal macrophages show higher responses to LPS stimulation, and IKKε and TBK1 have been suggested as counter-inflammatory kinases to attenuate the inflammatory response24. In terms of the seemingly incompatible results of the impact of IKKε on cardiovascular disease models, the outcomes seemed to be highly dependent on obesity. In other words, the findings suggest that IKKε may act as an inflammation brake in lean mice models and an inflammation accelerator in models with obesity. In this study, we used C57BL/6 mice fed standard chow and IKKε KO macrophages showed higher inflammatory response than WT macrophages (Fig. 1f). Likewise, LPS-induced iNOS induction was enhanced in IKKε knockdown RAW264.7 cells (Fig. 1a). These initial results indicated that IKKε might be involved in inflammation resolution, and we analyzed RNA transcription profiles of cardiac macrophages.
For cellular dissection of inflammation-related features in cardiac macrophages during MI, clustering and integration analyses were performed using Seurat (Fig. 2a). Unsupervised clustering identified key cardiac cell types (Fig. 2b), each with distinct cluster-specific markers (Fig. 2c). To characterize macrophages and fibroblasts post-MI, we initially isolated fibroblast and macrophage subsets. Although the proportion of Cd68(+) macrophages remained similar, Col1a1 expression and the percentage of Col1a1(+)Cd68(+) cells increased in the IKKε KO group (Fig. 2e, f). Macrophages in the IKKε KO group expressed higher levels of fibrotic genes (Fig. 2g). This specific macrophage subpopulation was larger in the IKKε KO group compared to the WT group, suggesting more frequent occurrence of in the IKKε KO group (Fig. 2h). Differentially expressed genes in the Macrophage 3 cluster included fibrotic genes, such as Col1a1 and Col1a2, where GO enrichment analysis revealed that the top genes are involved in fibroblast apoptosis process and its regulation which align with key signatures of MMT. These findings led us to investigate cell transition in macrophages toward fibrotic cells.
Then we explored phosphorylated proteins to identify the critical substrate of IKKε. The level of phosphorylated p38 was significantly lower in the IKKε KO macrophages than in the WT macrophages, and treatment of macrophages with SB, a p38 inhibitor, resulted in enhanced inflammatory responses. To analyze the features of enhanced inflammation in macrophages, we compared the molecular signatures by using scRNA-seq in non-myocytes isolated from both IKKε KO mice and WT mice at 4 days after MI. scRNA-seq analyses indicated that macrophages acquire fibrotic gene signatures that result in phenotypic and functional changes and may lead to fibrosis. This unexpected finding led us to turn our attention to cell transition.
Responses to cardiac injury involve dynamic interactions between individual cells that collectively regulate heart architecture and function. Cell populations are in continuous communication with their counterparts in response to injury signals. In the infarcted myocardium, macrophages proliferate and are dynamically recruited from bone marrow and spleen with diverse phenotypes. Resident fibroblasts rapidly move to the infarcted lesion to be activated into myofibroblasts, contributing to the wound healing process. Myofibroblasts originate from a number of sources25, including epithelial cells26,27, endothelial cells28, fibroblasts, pericytes28, and bone marrow29. Upon acute injury, cell phenotypes are not stable; these cells can change their functions with mixed identities in their microenvironment. This phenomenon is known as cell transition and is triggered in response to pathophysiological stimuli. Cell transition including epithelial-mesenchymal transition (EMT), endothelial to mesenchymal transition (EndMT), and mesenchymal-to-endothelial transition (MEndoT), is the process by which cells change their state during embryonic development, homeostasis, and tissue repair. In terms of the fibrotic changes in the myocardium, EndMT is induced by TGFβ1 to contribute to the accumulation of cardiac fibroblasts in a pressure overload mouse model, and the administration of BMP-7, a member of the TGFβ1 superfamily, significantly inhibits EndMT and the progress of cardiac fibrosis by targeting the fibroblasts carrying an endothelial imprint28.
MMT is a complex process in which macrophages partially lose their phenotypic and functional characteristics and acquire myofibroblast-like features. MMT was identified as mechanism by which ongoing inflammation causes progressive fibrosis in diseased tissues. In the early stages of benign prostatic hyperplasia, CD68(+)CD163(+) anti-inflammatory macrophages have been shown to be converted to the myofibroblast phenotype and to promote stromal fibrotic tissue remodeling30. We observed MMT in the infarcted myocardium in WT mice and more frequently in IKKε KO mice. More importantly, collagen-producing MMT was detected in human failing heart tissues. As far as we know, this is the first report showing MMT in human heart tissues. Although the mechanisms underlying cardiac MMT are still largely unknown, MMT may provide several advantages to the restoration of tissue structure and function. That is, the alternative types of macrophages may additively support the formation of a stable fibrotic scar to maintain cardiac pumping.
We have previously reported that 5AZ administration exerts cardioprotective effects through anti-inflammation in an angiotensin II-infusion cardiac infusion mouse model and rat MI model18,19. After finding that 5AZ treatment restored p-p38 in the stimulated macrophages, we examined whether the mechanism of 5AZ might be related with normalizing MMT. Interestingly, MMT was reduced in the IKKε KO group by 5AZ administration to a level similar to that of the WT group. These encouraging findings warrant further development of precision targeting therapeutics for cardiovascular and fibrotic diseases.
Development of cardiovascular therapeutics has been limited by an incomplete understanding of the diverse phenotypes of macrophages and origin of fibroblasts in the heart. Because the relative contribution of inflammation is likely dependent on contextual factors, we need to stratify cardiac pathophysiology and we need to understand to what extent inflammation contributes to appropriate MMT in order to identify therapeutic targets. Collectively, our data showed that IKKε-p38 axis in macrophages may be related with endogenous regulators of MMT. Macrophages are a heterogeneous population with an important role in heart homeostasis and the modulation of MMT may be a key signaling mechanism of cardiac therapeutics.