Utilising tissue-on-a-chip technology as an ex vivo model of breast cancer metastatic colonisation

Lead Research Organisation: University of Hull
Department Name: Biomedical Sciences


Breast cancer kills over 11,000 women each year in the UK. Virtually all of these women die because their breast cancer cells travel to other organs within the body such as the liver, lungs, bones and brain, where they grow into new tumours and stop the organs from working. Therefore, one way to stop women dying from breast cancer is to prevent breast cancer cells from growing in other organs. Currently, however, we do not understand how breast cancer cells move to and grow in other organs, so we are not able to prevent it with drugs.

Most studies trying to understand how breast cancer cells grow in different organs take place in mice. However this is not an ideal way to study the process, as it is very difficult to watch cancer cells growing inside a mouse. To get around this, large numbers of mice are used for each experiment with some being killed at different time points and the cancer cell spread to other organs then determined. This uses a lot of mice, and still does not allow us to see exactly what happens when cancer cells are growing in these organs.

Replicating the growth of breast cancer cells in other organs in a laboratory dish would be a way to better understand this process, whilst also reducing the number of mice used in research. Previous work has attempted to do this, however the laboratory systems which have been created lack the complexity of cancer cell growth seen in patients, so uptake amongst researchers has been low.

We have identified a model already in use which has the potential to be used to study breast cancer growth in other organs. This model has been developed at the University of Hull, and is called "tissue-on-a-chip". Tissue-on-a-chip involves taking a small amount of tissue from either a mouse or a human, and keeping it alive in a glass or polymer chip constantly supplied with flowing nutrients. These chips have previously been used to study both normal tissue and tumour tissue, and here we propose to study how breast cancer cells grow in other organs. Tissue from liver, a common site for breast cancer spread, will be placed inside the chips, and then cancer cells from breast tumours will be slowly flowed across and allowed to bind to and invade the liver. This will be monitored using powerful microscopes.

We believe that the successful development of this model could be of great use to scientists working to understand how cancer spreads. We have therefore designed this project firstly to adapt the technology for study breast cancer cells growing in other organs, and secondly to showcase its potential to other scientists. To adapt the tissue-on-a chip technology to study cancer spread we will transfer it to the University of Manchester, where we can use world-class facilities to watch cancer cells as they grow. We will also demonstrate how this technology can be used to test drugs to prevent cancer cells growing in other organs by adding different drugs, and seeing if we can prevent cancer cells from growing. As this technology is being established, we will begin discussions with other scientists across the UK to ensure that people know about this model, and will actively promote the benefits to their research alongside significantly reducing number of mice used in research.

Technical Summary

We aim to reduce the number of mice used in metastasis research by developing a novel ex vivo model to study metastatic colonisation. This tissue-on-a-chip model incorporates whole tissue pieces to model the metastatic microenvironment, combined with microfluidic channels to model the arrival of cancer cells at the metastatic organ, making it more physiologically relevant than in vitro models currently in existence. This technology has been extensively developed by Prof Greenman and Dr Green in a multi-disciplinary partnership at the University of Hull, and it is currently used for studies of tumour tissue determining response to chemo- and radiotherapy as a way of customising patient treatment.

In this project, we will transfer the technology to a new application: cancer metastasis. Initial experiments will take place in Hull to optimise conditions required for breast cancer cell colonisation. Colonisation will be monitored by immunohistochemical staining for human cytokeratin 19, in combination with an assessment of tissue viability by calcein-AM and PI staining and lactate dehydrogenase released from tissue effluent. Following optimisation of colonisation, the technology will be transferred to Dr Eyre at The University of Manchester. This new location will allow us to further develop the technology using advanced imaging and histological techniques available within the CRUK Manchester Institute. We will use confocal microscopy combined with second harmonic imaging to visualise cancer cells and surrounding tissue structures during colonisation, followed by automated multiplex immunofluorescence and multicolour slide scanning to determine metastatic niche cells associating with cancer cells during colonisation. Finally, we will demonstrate the application of this technology as robust method for screening anti-metastasis drugs, by perfusing candidate drugs into the microfluidics, and assessing the effect of these on cancer cell colonisation.

Planned Impact

We propose that tissue-on-a-chip technology can be utilised to reduce the number of mice used in metastasis studies. The vast majority of breast cancer metastatic colonisation studies currently take place in vivo, due to a lack of in vitro/ex vivo models which accurately reflect the metastatic process. A large number of mice are used in individual metastatic colonisation experiments, and mice are culled at specific timepoints throughout the experiments to assess the stage of cancer cell colonisation. We believe that this is an overuse of animals, and these experiments could be better performed ex vivo. Tissue-on-a-chip provides a 3Rs benefit whilst also providing a platform for studying metastatic colonisation which is more physiologically relevant than any in vitro/ex vivo model in existence. The development of this improved model will allow us to answer questions about metastatic colonisation which cannot be addressed in vivo, due to an inability to accurately image inside a living mouse.

In this project, tissue-on-a-chip will be used as a reduction model, as mouse liver will be used to model the metastatic microenvironment. However, each liver yields enough material for 48 chips to be set up in parallel, therefore a large amount of data can be gained from an individual mouse. Further, we plan to use "training mice" within the CRUK MI which would otherwise be culled, for the majority of the experiments. We have calculated the reduction potential of this model based on the numbers of mice used annually in the Manchester lab. In 2018, 76 mice were used for metastatic colonisation studies, over 3 experiments. Using tissue-on-a-chip would potentially reduce the number of mice from 76 to 3, a reduction of 73 mice (96%) locally. Nationally, we have identified 7 other groups undertaking breast cancer metastatic colonisation studies. Assuming they use a similar number of mice, this technology could give a reduction of over 500 mice in the UK each year. The accuracy of the scale of reduction is confirmed by Dr Penelope Ottewell (University of Sheffield), who confirms that by adopting this technology her lab will reduce mouse usage by over 60 mice/year (see letter of support).

To determine further worldwide reduction potential, we identified breast cancer metastatic colonisation studies published in 2018 from labs outside the UK. This revealed 14 publications. Each publication used an average of 23 mice across 3 experimental conditions, an average of 8 mice per experimental condition. All of these studies showed experiments which could be performed using tissue-on-a-chip rather than mice models. This has the potential to reduce the number of mice to 1 per publication, a reduction of about 300 mice used in published data/year. This literature search only included publications on breast cancer. In the longer term we plan to expand the model to also include other cancers, which will give a greater reduction potential.

These reduction estimates rely on uptake of the technology by other researchers. As our letters of support from world leaders in metastasis biology in the UK (Dr Gillian Farnie, Dr Penelope Ottewell), Europe (Dr Christina Scheel, Germany) and worldwide (Dr Tom Cox, Australia), there is an appetite for this technology throughout the metastasis community, and we have identified specific research groups that would be keen to adopt our model. To deliver this impact we will disseminate and promote our findings at scientific meetings, through the media, social media, and manuscripts including a JoVe paper providing video demonstrations of how to set up the technology. Once developed, we anticipate our tissue-on-a-chip models to initially serve as a pre-screening tool prior to in vivo experimentation. As confidence among the scientific community grows, and we continue to expand the technology to a fully human model, this technology could become a replacement for mice in metastatic research studies.


10 25 50