Figure 1.

Murine and Human Macrophage Populations in the Healthy and Inflamed Liver. (A) Two distinct populations of hepatic macrophages have been defined to date in the murine liver, the resident KCs, which make up the majority of the hepatic macrophage population and a smaller population of macrophages found in the liver capsule (21). In inflammation, these populations are often accompanied by populations of recruited macrophages. These can also exist in at least two subsets, recruited KCs which can persist in the tissue to generate resident KCs (4, 22–26) and a population of macrophages that are lost from the liver upon resolution of inflammation (15) here termed recruited-temporary macrophages. In the human liver, we still do not fully understand the distinct populations of macrophages present and how these relate to those found in mice. To date no counterpart for the murine capsule macrophages has been identified, however, this may be due to difficulties in isolating cells from the capsule, particularly from smaller liver biopsies which do not harbour significant amounts of capsule tissue. While all macrophages in the human liver can be identified on the basis of their expression of CD68, scRNA-seq studies have revealed that these can be further split into distinct subsets (27–30). Two populations of macrophages have been identified in healthy human liver tissue, which are distinguished by their expression of MARCO and TIMD4. Here we speculate (shown in grey) that the MARCO ⁺ TIMD4 ⁺ cells would be the counterparts of the murine resident KCs while the cells lacking expression of these two genes could be considered as recruited-temporary macrophages. In inflammation, a population of MARCO ⁺ TIMD4 ^- macrophages has also been identified thus it is tempting to align these with murine recruited KCs (30), however this also requires validation. In addition to these genes delineating subsets of human hepatic macrophages, CD32 has also recently been suggested to be a good protein marker to distinguish between these macrophages (31). (B) Due to their recent identification within the hepatic macrophage pool, the precise nature of the recruited-temporary macrophage population also remains unclear. In certain inflammatory settings, such as NASH and fibrosis/cirrhosis, it has recently been shown that these cells express genes including Spp1, Gpnmb (mouse) Trem2 and Cd9 (mouse and human). In mouse, these cells were termed lipid-associated macrophages (LAMs) while in human they were called scar-associated macrophages (SAMs) (4, 30, 32). Alignment of the LAMs and SAMs showed significant overlap (4) suggesting these could indeed be equivalent populations. In other inflammatory settings much less is known about these cells, and hence it is unclear if these cells also have a LAM/SAM phenotype or if their phenotype is dependent on the inflammatory stimulus. In acute liver injury in mice, they have been suggested to express genes associated with a function in resolution as well as genes associated with extracellular matrix (15). In human acute liver injury, a population of MerTK⁺CD163⁺ macrophages have been reported which may represent human recruited-temporary macrophages (33). Moreover, to date, the specific functions of these cells are largely speculative. To date, limited data is available regarding recruited-temporary macrophages in infection in mice and humans and thus their potential function(s) remain speculative.