Pilot Grant - Optimising liver equivalents to model liver fibrosis.

Liver disease causes over 16,000 deaths in the UK each year. This number is predicted to double over the next 20 years due to poor lifestyle choices and a lack of drugs. Regardless of cause liver disease follows a similar course; damage - inflammation - fibrosis (scars) - cirrhosis (dense scars) and increased risk of liver cancer. If patients do not dramatically change their lifestyle (reduce fat and alcohol consumption) or in the case of viral infections respond to anti-viral drugs there are no current treatments for liver disease. Patients with an autoimmune disease (primary biliary cirrhosis, primary sclerosing cholangitis or Wilsons disease) have no therapeutic options. Liver disease represents a massive clinical and economic burden therefore understanding how liver scars (fibrosis) develop and go away is critical to identifying new medicines. Our Laboratories research focus over the past 10 years has been to understand what causes liver scarring (fibrosis) by culturing scar forming cells on plastic in petri dishes. We use this information to identify new anti-fibrotic (scar preventing or removing) drug targets and use archived tissue from rodent and human disease to validate and confirm these targets. Once potential medicines are identified we test these drugs in models of liver disease. Developing an in vitro 3D Liver chip system that mimics normal liver cell-cell interactions and models scarring would provide an opportunity to test new drugs in primary human cells; this would directly REDUCE the number of animals required for in vitro (cells made from animals) and even in vivo (whole animal) experiments. A 2nd advantage of this system is "controllability". LiverChip scaffolds are in individual chambers and can be manipulated experimentally with disease modulators e.g. damaging agents, inflammatory cells or potential anti-fibrotic drugs in a tightly controlled, accurate manner. This level of control will greatly assist our understanding of what causes the disease and the breakdown or success of drugs. If a medicine causes significant hepatoxicity (liver damage) or does not achieve the same anti-scar effects in this system as the 2D cultures on plastic (petri dishes) for reasons such as rapid drug metabolism (breakdown), this may negate the need to test the drug in an in vivo liver disease model. Conversely, growing hepatocytes with scar cells (co-culture) may increase drug toxicity through direct or indirect mechanisms. This would prevent unnecessary procedures which carry both moderate and severe bands. Once established this system could be used to monitor scar generation or removal +/- drug intervention or damage stimuli daily over many days. The current hepatocyte LiverChip cultures at Zyoxel are growing for >2 weeks therefore they can be treated repeatedly with damage agents or test compounds. Preliminary published (hepatocytes and sinusoidal endothelial cells) and unpublished data from Zyoxel suggests that we would expect similar results in co-culture systems (1). Sampling the culture media and measuring makers of hepatocyte damage (liver enzymes release) or release of scar factors (collagens or fibrogenic cytokines e.g. TGFb1) would generate constant data streams to monitor disease and liver damage as the experiment progresses. Other applications of the LiverChips could include biomarker, drug breakdown and transport analysis.

If successful we hope that this system would be adopted by other liver researchers and prove a useful predictive tool for pharmaceutical companies. Long term this technology could be adapted to model disease in other organs susceptible to scarring e.g. kidney or skin.

1. Domansky K, et al. (2010) Perfused multiwell plate for 3D liver tissue engineering. Lab Chip 10(1):51-58.

Achievements to date: Zyoxel optimised the LiverChip with a Massachusetts Institute of Technology research group and Pfizer. The LiverChip is designed to mimic the hepatic architecture and is perfused and oxygenated by microfluidic flow under conditions akin to normal liver physiology. Cultures are contained within individual chambers, providing the "controllability" to accurately manipulate culture conditions to address the experimental question. Hepatocytes are viable and retain phase I and II metabolising enzyme expression and functionality for >2 weeks. Published work revealed that the system can be used to co-culture hepatocytes and liver sinusoidal endothelial cells. In this pilot grant we will receive training from Zyoxel to establish the system in Newcastle, maximising our chance of success.

Experimental Plan

Histological readouts
1. Fluorescence staining of aSMA to visualise scar forming HM.
2. Sirius Red stain (collagen deposition) to assess fibrosis.

Biochemical readouts
1. Measure fibrogenic gene expression; TIMP1, collagen, aSMA and TGFb1 by qRT-PCR.
2. Hepatotoxicity: measure transaminase release from "leaky" damaged hepatocytes into culture media.
3. Monitor hepatocyte function (cyp2e expression and cytochrome p450 activity).

Objective 1-2

1. Optimise seeding protocol to recreate a fibrotic liver. Isolate hepatocytes and hepatic stellate cells (HSC) from rat liver. Co-culture hepatocytes with increasing numbers of HSC. Assess culture viability and ask if co-culture causes HSC to HM activation or fibrosis using histological and biochemical readouts above.
2. Co-culture hepatocytes with activated HM and monitor fibrogenesis.
3. Once optimised in rat the protocol will be translated into primary human cultures.

Objective 3
Normal or fibrotic LiverChip will be treated +/- proven anti-fibrotics (angiotensin inhibitors or 5HT2BR antagonist) to validate the system. Anti-fibrotic effects and compound toxicity will be assessed.

Achieving our objectives will provide pilot "proof-of-concept" data which can be taken forward in future grant proposals to develop a novel system to model progression and regression of human liver disease and co-monitor toxicity. In the longer term a fully optimised and validated system would offer an attractive research tool which permits testing of novel therapeutics and provides further understanding of the complex cell:cell interactions in liver disease. The "controllability" of the system will permit accurate modelling of the liver's response to environmental cues, interaction with immune modulators and damage stimuli in a controlled setting to aid target validation. This would be of direct interest and relevance to other academic researchers and industry and ultimately lead to a direct reduction in animal numbers to generate cells. The use of LiverChips for drug discovery could lead to more effective in vitro experiments and minimise the number of in vivo models of toxic or surgical liver injury in mice and rats, which carry moderate and severe banding.

An indirect measure of the use of mice and rats in liver research can be gained from a simple Pubmed search. Entering "liver fibrosis and mouse" generates 3106 publications and "liver fibrosis and rat" generates 3630 publications. If you limit the year range to the previous decade the number of hits for rat are 1820 over half of the total publications, suggesting that this is a growing area of research. The close collaboration between our academic research team and Zyoxel will provide training in important skills for those employed on the project. Success in this pilot project would strengthen these collaborative links and could lead to future project grant applications to NC3Rs or MRC to develop the technology and transfer the skills to other organ systems in collaboration with colleagues within our institute.

The plan is ambitious and whilst it will not completely replace disease models it will be a refinement and reduction. Investment to build upon, develop and further refine LiverChip systems over the following years have the potential to achieve the common goal to help reduce experimental animal numbers and procedures carrying both a moderate and substantial bands. Typically these are the carbon tetrachloride chemical injury model (moderate) and the surgical bile duct ligation model (substantial) which despite in house refinements still carries a mortality of approximately 30%.

Once robust and validated systems have been established to recapitulate the complex cell:cell interactions to model liver disease, the experience and technology developed could be exploited to model other organ based fibrotic diseases such as kidney, skin or lung. Over the previous decade 1080 papers using mice have been published in the field of renal fibrosis and searching lung fibrosis and mice over the same period generates 2243 hits, indicating the potential long term benefits of developing this technology.