*3.1. SCF–KIT Axis as Chemotactic Guidance*

Initial work aimed to identify the factor(s) governing MC infiltration (chemotaxis). There have been several factors identified that govern MC migration [105]. Those include two key factors that act as a chemoattractant for MCs and that are important in the process of liver fibrosis. The first one and the most potent is TGF-β1, which is increasingly expressed in the liver under fibrogenic conditions [106,107]. However, aside migration/chemotaxis, TGF-β1 has an impact on several critical functions of MC biology, including proliferation, apoptosis, effector synthesis, and degranulation (see below). In the second line, stem cell factor (SCF), which is an essential factor for MC proliferation, survival, and differentiation is also a potent chemoattractant for MCs and circulating MC progenitors [108–110]. The SCF transcript can be differentially spliced, resulting in alternative transcripts, which differ in the presence or absence of exon 6. Both encoded proteins are membrane-bound but the longer one, including exon 6, can be more easily proteolytically cleaved to generate soluble SCF [111]. SCF is produced by fibroblasts and endothelial cells, and both the membrane-bound and soluble SCF bind to the surface tyrosine kinase membrane receptor KIT (CD117), which leads to the recruitment of MC progenitors and activation of MCs in the tissue.

Recruitment of MCs to the liver requires the liberation of progenitors into the circulation and entry into the tissue at the destination site. In this scenario, extravasation of circulating MC progenitors is facilitated by the interaction of MC progenitors with endothelial cells [11]. This contact can be mediated by the α4β7 integrin, as has been shown for murine MC homing/recruitment to the small intestine [11]. For human MCs, attachment to endothelial cells was shown to rely on the α4β1 integrin subunits [112]. As mentioned above, endothelial cells express the MC chemoattractant SCF.

In order to substantiate SCF as the basis for MC recruitment during liver disease, the expression of SCF in normal as well as diseased livers from patients suffering from PBC or PSC was analyzed using RT-qPCR and ELISA showing that SCF was increased. The corresponding SCF mRNA could be detected in human and rat primary HSCs. The expression of SCF increased during the culture, activation, and transdifferentiation of primary rat HSCs. In turn, isolated human skin MCs adhered in co-culture to HSC monolayer cells, most likely via membrane-bound SCF, since this interaction could be blocked by anti-SCF antibodies [110]. HSCs do not constitutively express SCF, but it can be induced by MCs in a TNF-α-dependent fashion [113]. In addition to HSCs, SCF is produced by keratinocytes, airway epithelial cells, and endothelial cells. In a recent paper, another critical player in fibrosis, i.e., cholangiocytes, have been shown to express and secrete SCF, in contrast to hepatocytes, which do not express SCF [114]. Inhibition of SCF expression in *Mdr2*−/<sup>−</sup> mice decreases MC recruitment in vivo. In vitro deprivation of SCF in cholangiocytes reduces MC migration and HSC activation. As a consequence, histamine levels, biliary reaction, and fibrosis were reduced in *Mdr2*−/<sup>−</sup> mice [115]. Moreover, it has been shown that the SCF-mediated migration of MCs depends on the activity of the sheddase/metallproteinase ADAM10 [115].

In addition to migration/recruitment, recombinant human SCF, as well as NIH3T3 fibroblasts (in co-culture), have the capability to induce the differentiation of MC precursors isolated from human fetal liver [116–118]. Indeed, mouse NIH3T3 fibroblast cells express SCF constitutively [119]. Analysis of glucocorticoids (dexamethasone) on a co-culture of MCs with NIH3T3 fibroblasts or HSCs showed that dexamethasone blocks NIH3T3-mediated MC proliferation, while it has no effect on HSC-mediated MC proliferation, because the latter induces SCF expression in HSC at a post-translational level [119].

In a recent work, Kim and colleagues analyzed extracellular vesicles (EVs) of patients suffering from SM. They found that EVs show an MC signature, including the proteins tryptase, FcεRI, and KIT [72]. Incubation of HSCs in vitro with these purified EVs led to the transfer of functional KIT to stellate cells. This transfer caused an increase in HSC proliferation, cytokine production, and differentiation. These effects were blocked by KIT inhibition, while in in vivo experiments, the application of EVs to mice increased α-smooth muscle actin (α-SMA) production and the presence of human KIT in mouse HSCs. These results suggest that KIT expression/presence on HSCs causes activation of these cells [72]. A mutual involvement of Kupffer cells in the recruitment of MCs to the liver was demonstrated in an endotoxin liver injury model that causes hepatocyte damage, an inflammatory reaction, and MC infiltration. Treatment of those animals with an agonist for the liver X receptor (LXR; GW3965), highly expressed in Kupffer cells, leads to reduced liver affection and reduction in MC count [120]. If this effect is mediated by Kupffer cells, or if the agonist has a direct impact on MCs is questionable since the LXR agonist GW3965 directly blocks at least inflammatory cytokine production of BMMC [121].
