Development and mathematical modelling of zonated hollow fibre bioreactors for in vitro to in vivo extrapolation of systemic chemical toxicity

Bioartificial livers containing porcine hepatocytes, with a hollow fibre bioreactor are already used in the clinic as supportive devices, allowing the patients liver to regenerate properly upon acute liver failure, or to bridge the individual's liver functions until a transplant is possible. This application seeks to employ this technology for in vitro to in vivo extrapolation of systemic chemical toxicity.

The scientific/technical basis of the project led by Dr Steven Webb from Liverpool John Moores University is the combination of a unique mix of tissue engineering/bioreactor development with mechanism‐based toxicology. This will be underpinned by mathematical (mechanistic) modelling and systems biology (data led modelling), which can both help direct research and develop a model for extrapolation of the in vitro findings to in vivo systems. The application will define the physiological & pharmacological response of primary hepatocytes and human hepatic cell lines in a bioreactor. Should this approach prove successful, later models will incorporate co‐culture and multi‐tissue bioreactors in series. Incorporation of features such as 3D cell architecture, vectorial liquid flow & oxygen gradients are favourable for liver zonation enhancing xenobiotic metabolism & pharmacokinetics / biomarker assessment. The project will predominantly focus upon the development of the hepatocyte compartment as metabolic capability here is crucial for assessing effects of metabolites. The project will be underpinned and informed by a strong mathematical modelling/systems biology approach that will form a data framework consisting of circulating drug and metabolite levels, tissue/cellular burden of metabolites, glutathione and covalent binding levels, adaptive response (Nrf2/NFkB nuclear translocation), apoptosis and necrosis biomarkers, this would allow more accurate in vitro to in vivo extrapolation to both animals and man.

Full details about this CRACK IT Challenge can be found on the CRACK IT website.

Reddyhoff D et al. (2015). Timescale analysis of a mathematical model of acetaminophen metabolism and toxicity. JTB 386:132-146. doi:10.1016/j.jtbi.2015.08.021

Storm MP et al. (2016). Hollow Fiber Bioreactors for In Vivo-like Mammalian Tissue Culture. J. Vis. Exp. (111), e53431, doi:10.3791/53431 (2016).  Video: doi:10.3791/53431 (2016).



A multi-disciplinary team led by Dr Steven Webb, Liverpool John Moores University, has developed a zonated hollow fibre bioreactor (HFB) that attempts to more closely replicate the architecture and physiology of the liver for toxicology testing.

Current in vitro test systems employed by industry include simple liver-derived cell-based models that tend to be poorly predictive of the toxic effects of chemicals entering the systemic circulation. The need for more predictive models for systemic toxicity is therefore clear.

The liver is a multifunctional organ that is zonated in terms of areas of specific cell function and sensitivity to toxicants. Hepatocyte phenotype and metabolism is dependent upon the cell position along the liver lobule (periportal, central and perivenous zones), with differing exposure to substrate, oxygen and hormone gradients.

By combining experimental and mathematical modelling, the team created a zonated HFB liver model, the first in vitro liver model with separate sinusoid and bile canaliculi compartments, which is currently not possible with other culture systems. Mathematical modelling played a large role in the development of the HFB, for example, in predicting the optimal operating conditions of the HFB to best recapitulate the zonated liver physiology and reduced development time by approximately six to twelve months. It is also worth noting that this exact design may not have been reached at all without the mathematical modelling. A schematic of the HFB system is shown in Figure 1. The system incorporates physiologically relevant 3D cell architecture, liquid flow and oxygen gradients to promote liver zonation. 

Figure 1. Schematic of the hollow fibre bioreactor set up. A) The system is composed of a borosilicate glass module fitted with three plasma treated porous polystyrene fibres. Cells are seeded onto the outside of the fibres and media is perfused through the fibre lumen from the inlet port, thereby replicating the liver sinusoid. There is an oxygen gradient along the fibres to promote liver zonation. The device is set up to ‘split’ the flow between the lumen, which is collected as the ‘permeate’ and force some flow across the cell membrane wall which is collected as the retentate. This allows biochemical analysis to be performed on the feed, permeate and retentate to quantify the amount of compound, metabolites and waste products in each, and ultimately the conversion of the compound per unit number of cells. B) Zonation divisions of differing length within the HFB. 

The system is composed of a borosilicate glass module fitted with three plasma treated porous polystyrene fibres. Cells are seeded onto the outer fibre wall and the lumen of the fibre acts as the sinusoid (blood vessel), through which media is perfused. The junctions between the cells and the extracapillary space around the outside of the fibres then act as a bile canaliculi compartment.

The system was optimised using the HepG2/C3A cell line. Cells were seeded (2x106 cells) onto the outer fibre wall and cultured for seven days with continuous media flow. Zonation of the HFB was confirmed by western blotting with known zone specific markers (Figure 2). In addition, the cells are polarised and form a canalicular membrane, confirmed by staining for the multi-drug resistant protein 2 (MRP2), which co-stained with high levels of F-actin (phalloiden staining) found to line bile canalicular structures (Figure 3). 

Figure 2. Western blot of known zonation markers. 

Figure 3. Immunofluorescence images of the Multidrug Resistance Associated Protein 2 (MRP2), Phalloiden (stain for F-actin) and Hoechst (DNA stain).  

The HFB system was compared to 2D static monolayer cultures and a comprehensive functional analysis between the two culture conditions was carried out. There was greater production of albumin (Figure 4A) and an increase in activity of key xenobiotic metabolising enzymes (Figure 4B) with a 2.7 fold increase in CYP3A4 and a 5 fold increase in CYP2D6 enzyme activity in HepG2/C3A cells cultured in the HFB model under flow compared to static 2D culture. 

Figure 4. Functional comparison between 2D static and hollow fibre bioreactor flow culture. A) Albumin production and B) CYP activity. 

Based on the analysis of multiple morphological and functional parameters, the HFB is an improved model system for HEPG2/C3A cells. The next steps are to further validate the system with primary heptocytes and a wider range of compounds, plus  to manufacture an easy-to-use prototype device for user trial.

For more information about the HFB system please contact Dr Steven Webb.

Back to top
CRACK IT Challenge

Award date:

Dec 2011

Contract amount