Chapter 2

Hepatic Stellate Cell Culture Models

Krista Rombouts,    UCL Institute for Liver and Digestive Health, Division of Medicine, University College London, Royal Free Hospital, London, UK

Hepatic stellate cells (HSCs) are perisinusoidal resident vitamin A-storing cells and are considered to be the most relevant profibrogenic cells operating in acute and chronic liver diseases. During liver injury, HSCs undergo phenotypic transformation from non-proliferating and non-contractile "quiescent" cells into "activated" promitogenic, profibrogenic, and proinflammatory myofibroblasts-like cells. Multiple etiologies of liver injury can lead to fibrosis, resulting from an imbalance between the production and degradation extracellular matrix (ECM). Thus, activated HSCs orchestrate an impressive response to different types of liver injury, leading to net deposition of ECM into the interstitium as a wound-healing reaction. In the last four decades several new and exciting in vitro models have been developed to mimic the normal microenvironment of HSCs in order to understand crucial molecular mechanisms of their activation that would be necessary to develop new clinical therapies. This chapter recapitulates the history of HSC isolation and provides an overview of various cell culture models. It also compares the differences between two-dimensional and three-dimensional in vitro models to investigate the characteristics of cultured quiescent and activated HSCs.

Keywords

Hepatic stellate cells (HSCs); cell culture; co-culture; two-dimensional culture (2D); three-dimensional culture (3D); extracellular matrix (ECM); precision-cut liver slices (PCLS); bioscaffold; Matrigel

2.1 Isolation of Hepatic Stellate Cells

In the past four decades several approaches such as density gradient separation [1-3], fluorescent cell sorting [4] and explant culture [5] have been employed to isolate hepatic stellate cells (HSCs). Separating HSCs from the total hepatic cell suspension allows culture of purified cell preparation to study, in a much defined way, the primary molecular signaling pathways associated with fibrogenesis. Explant culture and fluorescence-activated cell sorting (FACS) analysis are two cell isolation procedures that avoid the usage of density gradients. Nevertheless, density gradient centrifugation remains the method of preference for many investigators to isolate human and rodent HSCs. This method involves ex vivo digestion of the liver tissue by enzymes such as collagenase, pronase, and deoxyribonuclease (DNase) to dissociate the hepatic cells from the surrounding ECM. This is then followed by sequential centrifugation steps at different g force to separate HSCs from the other hepatic cell types. This is followed by a refined multi-step or single-step gradient centrifugation of the cell suspension through a density gradient of Nycodenz [6,7], Larex/Stractan/arabinogalactan [1,8], Optiprep [9], or Percoll [10,11] (Figure 2.1). Several alternatives have been optimized to isolate HSCs in combination with hepatic cells such as hepatocytes [12], liver sinusoidal endothelial cells (LSECs) [11], and Kupffer cells [13,14].

Figure 2.1 HSCs were isolated by employing an ex vivo enzymatic digestion of the liver tissue by enzymes such as collagenase, pronase, and DNase, which allows the dissociation of the hepatic cells from the surrounding ECM. This is then followed by sequential centrifugation steps at different g force to separate the various cell types from the HSC population. Next, a refined multistep or single-step gradient centrifugation allows to select and to purify HSCs. Different gradients and density solutions were used such as Percoll (35-50-90%) (A) and Optiprep (11.5-17%) (B).

HSCs obtained thus can be further enriched by sorting based on high side scatter of incident light [15]. Moreover, immediately after the density gradient centrifugation, purity of HSCs can be analyzed by taking advantage of the quickly fading greenish autofluorescence containing vitamin A-enriched lipid droplets under 328-nm illumination. Over the years, several markers for HSCs have been identified to ascertain purity of the cell preparation as well as in situ identification including vimentin, desmin, and glial fibrillary acidic protein (GFAP), nestin, synaptophysin, nerve growth factor (NGF) receptor p75, and alpha-smooth muscle actin (a-sma) [16]. However, desmin and GFAP are the gold standard immediately after isolation or early culture, and upon activation almost all HSCs express a-sma and nestin, which makes it easier to evaluate purity of the cell preparation. In addition, staining for macrophages (ED1 and ED2 in rat and F4/80 in mouse), epithelial cells (cytokeratin 19), and endothelial cells (SE1) is performed to further assure the proportion of contaminating cells.

The principle of using the density gradient centrifugation to isolate HSCs from other hepatic cell types is based on the presence of intracellular vitamin A-containing lipid droplets in HSCs. Indeed, HSCs are a major storage site of retinoids including vitamin A and play a cardinal role in their storage and controlled release. These lipid droplets differ in number and diameter and vary between species and under different physiological conditions [16,17]. The presence or absence of the lipid droplets is of major importance as the "activation" of resting vitamin A-rich HSCs into myofibroblast-like phenotype observed in chronic liver diseases is associated with the loss of retinoids and an increase in ECM synthesis [18-20]. This should be taken into account when isolating HSCs from the diseased fibrotic liver, induced, for example, by bile duct ligation (BDL) or carbon tetrachloride (CCl4) injections, and requires an increased collagenase and pronase concentration to digest the ECM and when using the density gradient centrifugation based upon the presence of retinoids lipid droplets [7,21].

The freshly isolated HSCs show prominent dendritic cytoplasmic processes, and the presence of lipid droplets. During culture over the following days, the morphology of HSCs gradually changes and displays a slightly more myofibroblast-like phenotype with heterogeneous retinoid droplet size (Figure 2.2).

Figure 2.2 Phase contrast microscopic images of HSCs in culture. Twenty-four hours after isolation the freshly isolated HSCs show prominent dendritic cytoplasmic processes and the presence of lipid droplets. The HSC morphology gradually displays a slightly more myofibroblast-like phenotype during the subsequent days in culture. Images were taken after HSC isolation at 24 h, 7 days in culture, and after the first passage. Magnification 4×, 10×, 20×, and 40×.

2.2 Single Cell Culture

Many cell culture models exist with different complexity. The most described and used model is the single monolayer culture of HSCs. The reason to use this in vitro model is that the primary quiescent HSCs found in the normal healthy liver spontaneously "transdifferentiate" into activated myofibroblast-like cells when cultured on non-coated plastic culture dishes in the presence of fetal bovine serum. Hence HSC "activation" refers to the transformation of the resting vitamin A-rich cell into a proliferating, fibrogenic, and highly contractile phenotype [22,23]. This in vitro model is well established and represents the process similar to that observed in chronic liver diseases. Thus, the single monolayer culture of quiescent HSCs has been used extensively to determine the role of specific proteins and genes during HSC activation in vitro (Figure 2.3). Moreover, this in vitro model enables researchers to investigate and compare the possible anti-fibrogenic effects of compounds between quiescent HSCs, transdifferentiated HSCs, and fully activated HSCs [6,24]. This information can then be extended/applied to the in vivo models [25]. Therefore, quiescent HSCs isolated from the normal, healthy liver and grown in a monolayer culture is an optimal tool to characterize the molecular mechanisms of the progression of HSC "transdifferentiation." As a consequence, this model has been used to determine the specific therapeutic target protein and signaling pathway(s) in HSCs and to determine which HSC phenotype (i.e., a quiescent vs. activated HSC) is targeted or is sensitive to the putative drug [6,26-28].

Figure 2.3 Confocal images of rat HSCs (A) and human HSCs (B) at different time points in culture. Cells were cultured on glass coverslip and stained with different antibodies. (A) Activated rat HSC demonstrates a strong presence of focal adhesion kinase (FAK, green color) and partial co-localization with vinculin (red color). (B) A mitotic human HSC is shown with B-actin localized in the cell body and small cellular processes, whereas phosphorylated myristoylated alanine-rich C-kinase substrate (MARCKS) is present as a centrosome protein. Both proteins are key players during mitosis in human HSC biology.

2.3 In Vitro- Versus In Vivo-Activated HSCs

Almost three decades ago, attempts were made to isolate in vivo-activated HSCs from the rodents in which fibrosis was induced by administration of CCl4...