Periodic Report Summary - CANCERSYS (Mathematical modelling of beta-catenin and ras signalling in liver and its impact on proliferation, tissue organization and formation of hepatocellular carcinomas)
The goal of the CANCERSYS project is to establish a three-dimensional integrative mathematical multi-level model capable of explaining fundamental processes in the formation of hepatocellular cancer.
On the tissue level the development of a tumour is modelled from the precursor cell until the tumour extends over several lobules. The three-dimensional architecture of each liver lobule during tumour development is resolved in time and space. Within each lobule sinusoids, the portal triads and central veins, the hepatocytes and hepatocyte-derived tumour cells, and other important cell types are explicitly modelled. The individual cells are parameterised within single-cell based models by cell-biophysical and cell-biological parameters, such as the cell proliferation status and rate, the death rate, the cell micro-motility, the cell polarity etc. (Hoehme et al., PNAS, 2010).They are tracked experimentally during the tumour growth process together with the size of the tumours, the distribution of vessels, etc.
This set of parameters - which we name 'process parameters' - are used to quantitatively characterise the tumour development process on the cell and tissue scale. In a later step dynamic models of the beta-catenin and ras core modules and their interactions are integrated into each model cell in order to gain insights into processes controlling single cell decisions. Signalling core modules of the -catenin and the ras networks were established based on literature data and on quantitative immunoblotting.
Modelling with whole genome expression profiling from hepatocytes with activated beta-catenin or activated ras signalling identified negative interactions between both pathways. Signalling through -catenin leads to attenuation of expression of genes that are positively regulated by the ras signalling module and vice versa. The model predicted several DUSP proteins as probable candidates for the negative -catenin/ras interactions.
Currently, validation experiments with siRNA knockdown of the candidate DUSP proteins are performed. Predictions obtained by the mathematical model will be validated by inducible transgenic mice, in which beta-catenin and ras signalling can be manipulated in hepatocytes. In an iterative process the model will be validated and adjusted to the in vivo situation. For model validation of -catenin signalling we used a tamoxifen inducible transgenic mouse, where knockout of exon 14 of the APC gene causes overexpression of -catenin. When -catenin overexpression is induced in virtually all hepatocytes this leads to proliferation and hepatomegaly within seven days.
A spatial-temporal model of this scenario has been established (see http://www.ifado.de/cancersys/publications/index.html online). When, in contrast, the APC knockout is induced in a small fraction of hepatocytes this leads to hepatocellular carcinomas within two to six months. We observed two histologically distinct types of hepatocellular carcinomas, namely poorly and well-differentiated tumours, which are presently under investigation.
On the tissue level the development of a tumour is modelled from the precursor cell until the tumour extends over several lobules. The three-dimensional architecture of each liver lobule during tumour development is resolved in time and space. Within each lobule sinusoids, the portal triads and central veins, the hepatocytes and hepatocyte-derived tumour cells, and other important cell types are explicitly modelled. The individual cells are parameterised within single-cell based models by cell-biophysical and cell-biological parameters, such as the cell proliferation status and rate, the death rate, the cell micro-motility, the cell polarity etc. (Hoehme et al., PNAS, 2010).They are tracked experimentally during the tumour growth process together with the size of the tumours, the distribution of vessels, etc.
This set of parameters - which we name 'process parameters' - are used to quantitatively characterise the tumour development process on the cell and tissue scale. In a later step dynamic models of the beta-catenin and ras core modules and their interactions are integrated into each model cell in order to gain insights into processes controlling single cell decisions. Signalling core modules of the -catenin and the ras networks were established based on literature data and on quantitative immunoblotting.
Modelling with whole genome expression profiling from hepatocytes with activated beta-catenin or activated ras signalling identified negative interactions between both pathways. Signalling through -catenin leads to attenuation of expression of genes that are positively regulated by the ras signalling module and vice versa. The model predicted several DUSP proteins as probable candidates for the negative -catenin/ras interactions.
Currently, validation experiments with siRNA knockdown of the candidate DUSP proteins are performed. Predictions obtained by the mathematical model will be validated by inducible transgenic mice, in which beta-catenin and ras signalling can be manipulated in hepatocytes. In an iterative process the model will be validated and adjusted to the in vivo situation. For model validation of -catenin signalling we used a tamoxifen inducible transgenic mouse, where knockout of exon 14 of the APC gene causes overexpression of -catenin. When -catenin overexpression is induced in virtually all hepatocytes this leads to proliferation and hepatomegaly within seven days.
A spatial-temporal model of this scenario has been established (see http://www.ifado.de/cancersys/publications/index.html online). When, in contrast, the APC knockout is induced in a small fraction of hepatocytes this leads to hepatocellular carcinomas within two to six months. We observed two histologically distinct types of hepatocellular carcinomas, namely poorly and well-differentiated tumours, which are presently under investigation.