Recapitulation of the intergrated liver sinusoidal environment in vitro using a multicomponent cell culture system

Lead Research Organisation: University of Birmingham
Department Name: Clinical and Experimental Medicine


The liver is the largest organ in the human body and carries out many functions. These including cleaning the blood of toxins and microbes and processing food into proteins to build tissues. The liver is made up of cells the most of important of which are hepatocytes and these are the engines of liver function. The access of hepatocytes to factors in the blood is controlled by another cell type, the endothelial cell, which lies above the hepatocyte in contact with the blood. The endothelial cells from channels called sinusoids which penetrate throughout the liver. This increased the amount of blood the liver cells are exposed to allowing them to efficiently remove drugs and waste products. The endothelial cells lining the sinusoids form a barrier between the hepatocytes and are specifically designed for this specialised function within the liver. For example they contain small pores or fenestrations which allow them to filter particles from blood and they have molecules on their surface which allow them to remove factors from the blood. During liver disease the endothelial cells change and increase their ability to remove bacteria and to attract white blood cells to fight infection. It is important to be able to study liver cells in the laboratory to help us understand human liver biology and what goes wrong in disease. If we can do this in the test tube using cells from human beings this will reduce the requirement for animal experimentation. However if the cells do not behave normally once they are grown in culture our models will not be useful. We have developed ways of getting liver cells out of liver tissue which is removed from a patient when they have a liver transplant. We can grow the different liver cell types in the test tube and have developed a way to grow the different cell types together called coculture. This allows us to grow an artificial 3D structure which resembles the liver sinusoid. Our early results show us that this system allows the cells to behave in a manner similar to how they behave in the body. However we need to do more experiments to determine exactly how similar these cells are to those found in humans. We want to find out (a) whether cocultured cells are more like normal cells in the body than cells grown in isolation and whether they function as normal cells would be expected to do; (b) whether the coculture method stops the liver cells becoming like diseased cells in culture; (c) whether the success of the coculture depends on the type of cells we use (e.g. if we take them from a diseased liver or if they are extracted from a normal donor liver which wasn¿t suitable for transplantation; (d) finally we want to understand the molecules and interactions which cause our endothelial cells to stay like normal cells when they are cocultured with hepatocytes? These studies are important because we will learn new information about how the human liver develops and functions both in health and disease. Our culture system could also be used to test new drugs for toxic effects on the liver, thereby reducing the need for animal experiments and perhaps as a basis for the development of an artificial liver to help treat people with liver failure.

Technical Summary

Human hepatic sinusoidal endothelial cells (HSEC) represent a unique endothelial population with exceptional phenotype related to their specialised functions. The cells have fenestrations, lack tight junctions, produce minimal basement matrix and express a unique profile of surface proteins and adhesion molecules under physiological conditions. In disease the sinusoidal nature of the endothelium changes and the cells resort to a more classical vascular or capillarised phenotype which regulates the recruitment of increased numbers of leukocytes into the liver. Isolation of HSEC from the liver microenvironment and long term culture is associated with phenotypic dedifferentiation which mirrors capillarisation in vivo. Thus cultured HSEC may not truly represent resting sinusoidal endothelial cells, hampering the development of physiological in vitro sinusoidal models. There is thus a need to develop more physiological, integrated in vitro systems which recapitulate the in vivo sinusoidal microenvironment. We have developed such a model by coculturing primary human HSEC with primary human hepatocytes on either side of a porous cell culture insert. Our preliminary data show that the phenotype of HSEC is altered by coculture with either primary human hepatocytes or HepG2 cells and that this process is contact dependent. Coculture changed the expression of several endothelial adhesion molecules including CD31, ICAM-1 and VCAM-1 on cocultured HSEC and this was associated with a marked increase in the ability of endothelial cells to recruit lymphocytes under physiological blood flow conditions. The adhesion promoting effect of coculture was dependent upon physical contact between the hepatocytes and HSEC and was specific to hepatocytes since CHO cells did not induce the same response. In the proposed studies we shall further develop the model and determine the optimal coculture conditions for promoting a sinusoidal endothelial cell phenotype and the paracrine signal involved. Initially we determine the effects of different coculture conditions on gene expression, phenotype and morphology of both the HSEC and hepatocytes using fluorescent and electron microscopy, flow cytometry and Affymetrix gene arrays. The phenotypic analysis will include determination of expression of general and tissue-specific endothelial markers including CD31, VE-Cadherin, VAP-1, vWF, ICAM-1 and ¿2, VCAM-1, FcRII and II as well as molecules that confer specialised functions on HSEC including CD36, DC-SIGN, L-SIGN, SR-A, SR-B1, Clever-1, and LYVE-1. In parallel we will analyse whether bi-directional signalling between HSEC and hepatocytes has implications for expression of hepatocyte-specific genes such as albumin, HNF-4 and AFP and functional activity of CYP450 isoforms. We shall compare the ability of normal and diseased hepatocytes and the hepatocyte cell line HepG2 cells to modulate the phenotype of primary human sinusoidal endothelial cells as well as determining whether coculture can induce a sinusoidal phenotype on a generic a non-hepatic endothelial cell (HUVEC). We will determine the molecular mechanisms responsible for the regulation of HSEC phenotype in coculture by manipulating the physical conditions of coculture and by using a combined proteomic and gene array analysis to look for the specific factors involved.
Description We have used our coculture systems to investigate the regulation of the specialised phenotype of sinusoidal endothelial cells. By using techniques including flow cytometry, immunocytochemistry, ELISA, multiplex and analysis of gene

expression (RT-PCR/gene arrays/ SAGE) we have confirmed that

i) HSEC express a unique profile of phenotypic markers which underlie their specialized functions in vivo. These include but are not restricted to DC-SIGN, L-SIGN, L-sectin, stabilin-1, LYVE-1, Mannose receptor, FcRs (CD16, CD32),

CD36, SSAO. They also express low levels (or lack) other markers which are more abundant upon vascular EC including CD99, CD31, Lysyl oxidase, vWF, Collagen 1a, Vimentin, Laminin a1 and b4.

ii) HSEC secrete a number of soluble mediators which influence the local inflammatory environment including (IL-1b, IL-1RA, IL-2R, IL-4, IL-6, I:-7, IL-8, IL-12, bFGF, GCSF, CXCL9-11, HGF, CCL3, CCL4 and VEGF)

iii) As HSEC dedifferentiate in culture they produce a more mature basal lamina by increasing production of Collagen 1A, Collagen IV and Laminin. They also loose expression of FcRs, VAP-1, DC-SIGN, L-SIGN, CD36, SR-B1

and elevate expression of CD31 and CD34.

Coculture with hepatocytes dramatically alters the phenotype of overlying endothelial cells and restores or maintains sinusoidal phenotype. Thus expression of markers such as CD36, ICAM-1, DC-SIGN, L-SIGN and SR-B1 is elevated in

coculture and the expression of soluble factors such as IL-6, Il-8, GCSF, HGF, CXCL10, CXCL9, CCL2 and VEGF is modulated. We have also demonstrated that the phenotype of hepatocytes is modified in coculture (with particularly

pronounced changes in albumin production and expression of CYP2E1, ADH1a and ADH5)

Interestingly we have demonstrated that hepatocytes cell lines are almost as potent as primary hepatocytes at modulating the phenotype of overlying HSEC. However if they were cultured with vascular endothelium (HUVEC, Dermal EC or

Sapphenous vein EC) we did not observe expression of characteristic HSEC markers such as L-SIGN and DC-SIGN and soluble production of IL-8, IL-15, IFNg, GMCSF, CCL3 and CCL5 was significantly different than cocultures using HSEC
Exploitation Route Our model systems are of interest to the pharmaceutical industry and have potential for use in toxicity testing and other applications We have devised model systems which can be used by other scientists interested in understanding the biology of the liver. We have published data which will be useful to scientists in biomedical disciplines and we have generated libraries of gene expression data which are available via public databases.
Sectors Chemicals,Healthcare,Pharmaceuticals and Medical Biotechnology

Description Coordinated regulation of glucose and lipid uptake and metabolism by hepatocytes and hepatic sinusoidal endothelial cells
Amount £72,540 (GBP)
Funding ID BB/G529824/1 
Organisation Biotechnology and Biological Sciences Research Council (BBSRC) 
Sector Public
Country United Kingdom
Start 09/2008 
End 05/2013