Exploiting dynamic enhancer landscapes to decode macrophage and microglia phenotypes in health and disease

Abstract

The development and functional potential of metazoan cells is dependent on combinatorial roles of transcriptional enhancers and promoters. Macrophages provide exceptionally powerful model systems for investigation of mechanisms underlying the activation of cell-specific enhancers that drive transitions in cell fate and cell state. Here, we review recent advances that have expanded appreciation of the diversity of macrophage phenotypes in health and disease, emphasizing studies of liver, adipose tissue, and brain macrophages as paradigms for other tissue macrophages and cell types. Studies of normal tissue-resident macrophages and macrophages associated with cirrhosis, obese adipose tissue, and neurodegenerative disease illustrate the major roles of tissue environment in remodeling enhancer landscapes to specify the development and functions of distinct macrophage phenotypes. We discuss the utility of quantitative analysis of environment-dependent changes in enhancer activity states as an approach to discovery of regulatory transcription factors and upstream signaling pathways.

Keywords: ATAC-seq; Alzheimer's disease; Disease associated microglia; Hematopoietic progenitor cells; Hematopoietic stem cells; Kupffer Cell; Lipid associated macrophages; Macrophage; NFkB; Nonalcoholic steatohepatitis; PU.1; Scar associated macrophages; TLR4; Trem2; enhancer; epigenetic; histone acetylation; lipopolysaccharide; microglia; obesity.

Figure 1.. Nature and nurture of tissue resident macrophages.

Yolk sac-derived hematopoietic progenitor cells (HPCs) engraft tissues during early fetal development. This process continues following transition of hematopoiesis from the yolk sac to fetal liver, resulting in all tissues containing embryonically derived resident macrophage populations at the time of birth. Each tissue provides a distinct environmental context that dictates the resulting resident macrophage phenotype, exemplified by brain microglia, liver Kupffer cells and adipose tissue macrophages. Microglia are self-renewing and remain exclusively HPC-derived throughout life in the healthy brain. Resident macrophages are replaced to variable extents in other tissues over time by HSC-derived monocytes that exhibit convergent but non-identical patterns of differentiation with the embryonically derived macrophages.

Figure 2.. Mechanisms driving the selection and activation of signal-dependent, macrophage-specific enhancers.

Panel A. Combinations of lineage determining transcription factors (LDTFs) with the ability to interact with their DNA recognition motifs in the context of closed chromatin initiate enhancer selection through collaborative binding interactions with each other and dozens of other co-expressed collaborative transcription factors (CTFs). These binding interactions establish regions of open chromatin that enable access to broadly expressed signal-dependent transcription factors (SDTFs) such as NFκB (left branch of pathway). Activation of NFκB by LPS results in its binding to poised enhancers containing κB recognition motifs, resulting in enhancer activation and control of target genes in a macrophage-specific manner. A variation of this pathway involves collaborative interactions with active repressors, which recruit co-repressor complexes that maintain enhancers in a poised-repressed state (right branch of pathway). Examples include RbpJ in the absence of Notch signaling and unliganded nuclear receptors such as LXRs. Notch signaling, or LXR agonists, convert Notch and LXRs from repressors to activators, respectively, resulting in co-activator recruitment and enhancer activation. Panel B. Example of activation of primed enhancers (Yellow stripes) associated with the Ptges gene in bone-marrow derived macrophages. Treatment with a TLR4 agonist leads to p65 (NFκB) binding at genomic locations occupied by PU.1 and C/EBPβ under vehicle treatment conditions. Binding of p65 is associated with gain of H3K27ac and increased expression of Ptges. Data from (Link et al., 2018).

Figure 3.. Environment-independent and environment-dependent macrophage enhancers and super enhancers.

Panel A. Genes encoding core macrophage LDTFs are associated with super enhancers in which the component enhancers are bound by combinations of each factor, e.g., PU.1, C/EBPs and AP-1 family members. This property is exemplified by the binding of PU.1, C/EBPβ and cJun at a super enhancer (yellow bar) associated with the Spi1 gene, which encodes PU.1. These relationships establish a mutually reinforcing, environment independent transcriptional network that maintains expression of core macrophage LDTFs and drives expression of core macrophage genes, such as TLRs, phagocytic receptors, etc. Panel B. Many tissue-specific macrophage enhancers and super enhancers are selected and activated by combinations of transcription factors that are environment-dependent, exemplified by Smad4 and LXRα in Kupffer cells. These factors must be expressed and active in order for monocytes to acquire a Kupffer cell phenotype following Kupffer cell depletion.

Figure 4.. Inference of transcription factors driving changes in macrophage fate and state in response to environmental signals.

Panel A. Transitions in cell fate and state are typically driven by selection and activation of hundreds to thousands of enhancers that in turn regulate gene expression. This dynamic response can be quantitively assessed at a genome-wide level using assays of open chromatin (e.g., through the use of ATAC-seq) and assays that are surrogates of enhancer activity (e.g., ChIP-seq for H3K27ac). Enriched DNA recognition motifs in the set of regulated enhancers represent binding sites for the dominant transcription factors responsible for transitions in enhancer function. The identities of these factors may enable inference of upstream signaling pathways. Panel B. Enriched motifs and inferred transcription factors and upstream signaling pathways in enhancers that are activated during monocyte to Kupffer cell differentiation. Motifs for SMADs, RBPJ and LXRs implicate TGFβ-receptor, Notch and LXRs as signaling pathways important for Kupffer cell differentiation. TGFβ and the Notch ligand DLL4 are potential environmental signals that are provided by sinusoidal endothelial cells. Hepatocytes produce desmosterol that is an activating ligand for LXRs.

Figure 5.. Inference of transcription factors driving changes in macrophage phenotypes in disease.

Panel A. Temporal sequence of events in which resident or recruited macrophages sense disease-associated signals, leading to induction of Trem2 and other Trem2-independent genets. Expression of Trem2 enables additional sensing of disease-associated signals and regulation of downstream gene expression, presumably through Syk/PI3K-dependent mechanisms. Trem2-independent and Trem2-dependent gene expression contribute to generation of disease associated macrophage phenotypes, exemplified by DAMs, LAMs and SAMs. Panel B. Inference of transcription factors associated with transition of Kupffer cells to a ‘SAM-like’ phenotype in a mouse model of NASH. A NASH inducing diet result in gain of H3K27ac at >4000 enhancers and loss of H3K27ac at >3500 enhancers. These changes are associate with upregulation of Trem2 and other LAM/SAM genes and downregulation of Kupffer cell identity genes. Motifs associated with gained H3K27ac include AP1/ATF, NFAT, RUNX and EGR motifs. Panel C. NASH-induced changes in Trem2-associated enhancers. Genome browser tracks for Kupffer cell ATAC-seq, H3K27ac, ATF3 and LXR are illustrated under control and NASH conditions in the vicinity of the Trem2 gene. The NASH diet leads to marked increases in ATF3 and LXR binding that is correlated with selection of a 3’ downstream NASH-specific enhancer and increased H3K27ac at the Trem2 promoter, a poised intronic enhancer and the NASH induced 3’ enhancer.