CTA: Are there any limitations or downsides to these methods?

The biggest limitation is that to make these networks you need a lot of high quality data, which are largely absent right now.

CTA: What are the implications of this research? How might the finding that cell shape can regulate signaling improve researchers’ understanding of cancer?

Dr Bakal: For researchers, it is important to assume that whatever behavior we are studying, whether it is signaling or any cellular behavior downstream of signaling, such as division, proliferation, apoptosis, or differentiation, that the shape of a cell is playing a major part in deciding the outcomes of those decisions. The shape of the cell is deciding cell fate. When we study cancer cells, we often forget that. People study using different biochemical techniques or genetic techniques, but very rarely are they aware of the shape of the cell as they are doing these studies. By understanding the shape of the cell you can understand how different biochemical pathways might be affected by it.

From a clinical or treatment perspective, it is important to always be thinking whether we can manipulate the shape of cell as a better way to treat tumors rather than, for example, targeting a rogue enzyme. A lot of currently-used chemotherapeutics that are effective target cytoskeletal components. It is thought the main reason the drugs work is because they halt cell division, but it is possible that they could also be helping treatment because they are affecting cell shape in other ways.

CTA: What are the planned next steps for your research?

Dr Bakal: Our next steps are to build more data and put more data into our model. Right now, our models based on shape data from 2-dimensional tissue culture systems, gene expression data, and mRNA expression data from those systems in breast cancer. There are lots more data out there and I think we can generate some of these datasets ourselves.

We linked shape to genes, but we could also link shape to protein expression, protein phosphorylation, or microRNA expression. We can make these models better and make them more realistic by generating data for cells that have been, for example, cultured in more physiologically relevant models.

One thing that I would like to do is repeat this study but measure shape and expression in cells in the 3-dimensional environment that cells are more likely to see in body. We want to make the models better, but also make them more physiologically relevant and try to develop it for other cancer types.

Reference

  1. Sailem HZ, Bakal C. Identification of clinically predictive metagenes that encode components of a network coupling cell shape to transcription by image-omics. Genome Res. 2017;27(2):196-207. doi: 10.1101/gr.202028