**1. Introduction**

Liver fibrosis is denoted by the excess deposition of extracellular matrix (ECM) components in response to chronic liver injury, such as viral hepatitis, cholestatic disorders, alcoholic liver disease or non-alcoholic fatty liver disease (NAFLD). With perpetuated injury, liver fibrosis might progress to cirrhosis and facilitate hepatocellular carcinoma (HCC) formation [1]. Fibrosis accounts for severe morbidity and mortality and has been shown to determine the outcome of patients with NAFLD [2].

Bone marrow-derived fibrocytes (FC) represent a unique cell type, sharing features of both hematopoietic and mesenchymal cells. While their secretion of collagens and other ECM components resembles fibroblasts, they express various leucocyte markers (e.g., CD34, CD45, CD11b, Ly6C, and F4/80) [3,4] and are hence commonly identified by the simultaneous expression of CD45 and collagen I [5]. Fibrocytes comprise ∼0.5% of peripheral blood leucocytes and rapidly enter the site of injury in physiological and pathological wound healing processes [3,6]. Since they had been explicitly described by Bucala et al. in 1994 [3], FC have been shown to participate in fibrotic diseases of the lung [7], kidney [8], heart [9], and colon [10,11]. Moreover, they are implicated in the pathogenesis of asthma [12], inflammatory bowel disease [10,13], and ocular disorders [14].

In recent years, the mechanisms of hepatic fibrogenesis have been studied extensively. Activated hepatic stellate cells (HSCs) and, to a lesser extent, portal fibroblasts were identified as the main source of contractile, α-SMA<sup>+</sup> myofibroblasts in the liver, which, being absent under healthy conditions, drive scar tissue formation during hepatic fibrogenesis [15–17]. HSCs are therefore commonly considered the key to understanding and treating liver fibrosis [15,18,19]. However, there is compelling evidence that FC, in fact, contribute to liver fibrosis. Fate-tracing studies demonstrated that fibrocytes specifically infiltrate the liver upon injury [20] and participate in fibrogenesis by the secretion of ECM components [21,22]. Furthermore, FC constitute a potential source of myofibroblasts. Although the transdifferentiation into myofibroblasts has been shown both in vitro [7,23] and in vivo [12], its relevance remains controversial.

Besides their direct contribution to fibrogenesis, FC exert a variety of paracrine functions (reviewed in references [24,25]), thus possibly influencing liver fibrosis. FC, on the one hand, express the fibrogenic mediators TGF-β and PDGF [24,26], which are essential for the activation and proliferation of myofibroblasts [1,27]. FC can acquire an inflammatory phenotype, characterized by the production of cytokines (e.g., TNF-α, IL-1β, and CCL2,-3,-4) and eicosanoids [26,28], and their capability of antigen presentation [29]. On the other hand, FC are able to promote the degradation of ECM components via the secretion of matrix metalloproteinases (MMPs) [30,31], regulate angiogenesis [32,33], and exert antimicrobial defense mechanisms [34].

Given the complex interplay of the aforementioned factors in the pathogenesis of liver fibrosis, the overall contribution of FC remains highly speculative. We therefore seek to characterize the role of FC on experimental liver fibrosis in vivo by specific depletion of these cells. Utilizing the well-established herpes simplex virus thymidine kinase (HSV-TK)/Valganciclovir (VCV) model, driven by a collagen I promotor, we depleted collagen I-expressing cells of bone marrow (BM) origin [35–38] in C57BL/6J mice. All collagen-producing cells herein co-express the HSV-TK, making them susceptible to killing by Valganciclovir. BM of such mice was transplanted into non-transgenic mice in order to limit the effect to cells of BM origin. Introducing the HSV-TK via bone marrow transplantation circumnavigates the issue of an unstable expression of the widely used markers CD34 and, as shown recently, CD45 in fibrocytes [5]. Additionally, the collagen-promotor driven expression of HSV-TK enables the killing of fibrocytes in various stages of their development or differentiation. Although a CD14<sup>+</sup> cell population, located in the bone marrow, is assumed to be the origin of fibrocytes [23], the exact differentiation pathways remain poorly understood. Thus, the results and possible side effects of an approach that interferes with alleged monocyte precursors in order to deplete fibrocytes, as reported recently [39], seem hardly predictable to us.

While different animal models of murine liver fibrosis have been described [40], with CCl4-induced liver injury certainly being the most popular, we chose to induce fibrosis with thioacetamide (TAA). TAA, administered via drinking water, causes a chronic-toxic, more slowly progressing fibrosis with only moderately increased serum transaminase levels, thus closely mimicking alcoholic liver fibrosis in humans [41,42]. The aim of the present study was two-fold: To determine whether the depletion of fibrocytes (1) ameliorates fibrosis, as indicated by decreased hepatic hydroxyproline content, and (2) attenuates liver cell damage, denoted by reduced serum alanine transaminase levels.
