Blog Overview

State-of-the-art bioinformatics for better drug discovery

29.3.17
Dr. Aurélien Rizk: "We are developing different assets for the discovery of functionally selective drugs with improved efficacy and reduced side effects."
Dr. Aurélien Rizk: "We are developing different assets for the discovery of functionally selective drugs with improved efficacy and reduced side effects."

Did you like this article? Please share it.

Tags

Innovation Areas

Downloads

The effect a drug has on the human body depends on how well its active substance find its way to the targeted cells in the body, and on the chemical reactions it activates in the maze of cell signaling pathways. To understand this process, InterAx Biotech AG, a spin-off from the Paul Scherrer Institute (PSI) and the ETH Zurich based at PARK INNOVAARE, has developed a unique solution. InterAx’s Biosensors Technology Platform enables a quantitative comparison of drug-induced GPCR (G protein-coupled receptors) signaling pathways, thus making it possible to identify better and safer selective lead molecules for this class of human receptors.


The unique approach of InterAx Biotech is to combine protein-based IP-protected biosensors and the mathematical modeling of signaling pathways that is fed with the data derived from the biosensors and cell image analysis. Dr. Aurélien Rizk, computer scientist, systems biologist and one of the co-founders of InterAx, explains the technology that lies behind the platform.

PARK INNOVAARE: The success of InterAx lies in combining cell engineering technologies (biosensors) and bioinformatics (system modeling). How are those two fields interconnected?

Dr. Aurélien Rizk: In 2014, during my postdoc at the Paul Scherrer Institute, I was working on the modeling of the signaling pathways of the serotonin receptor. I developed a mathematical model of the chemical reactions that take place in the cell after the stimulation of the receptor by a drug. It was possible to extend the model with additional data provided by Dr. Martin Ostermaier (CEO of InterAx Biotech), with whom we later founded the company. On the one hand, the biosensors he was working on were a perfect source of new data. On the other hand, the model helped interpret the data biosensors provide, make predictions and evaluate the tested compounds. Today, combining innovative IP-protected biosensors and the mathematical model, we are developing different assets for the discovery of functionally selective drugs with improved efficacy and reduced side effects for the largest class of human receptors, the G protein-coupled receptors (GPCRs).

What is your approach to the modeling of cell signaling pathways?

An important difference between existing approaches and ours is that at InterAx we also take into account the localization of receptors within the cell, i.e. we analyze the cell as a complex structure, a combination of “compartments”, and try to understand if and how the receptor’s location influences chemical reactions within the cell. By consequence, our approach requires new experimental methods to locate the receptors and advanced image analysis techniques to feed the variables into the mathematical model. For this, I have developed Squassh [1] (Segmentation and QUAntification of Subcellular SHapes) – an automatic image analysis technique making it possible to quantify the localization of receptors in the cell.

Squassh analysis example

Squassh analysis example: On the left-hand side, a florescent microscope image of a COS cell 30 mins after stimulation is represented; on the right-hand side – its squash analysis that will be later fed into the mathematical model.

What are the advantages of the Squassh technology?

Since every cell behaves differently, it is important to collect as much data as possible from different cells to build a coherent signaling pathway model. Using microscopy and different fluorescent markers to tag cell “compartments”, we can obtain enough images. However, it is nearly impossible to analyze all this information manually. Squassh is there to help process it quicker and with less bias. Moreover, compared to other image analysis software, Squassh is more precise as it can identify smaller vesicles (small structures within a cell, consisting of fluid enclosed by a lipid bilayer). The software can also be calibrated depending on the microscope characteristics, which makes it suitable for different microscopes, including high-throughput microscopes. Finally, in addition to the quantitative and qualitative advantages, Squassh is easy to use. It requires less tuning: any biologist can set up necessary parameters to get the best out of the software.

What have you already been able to achieve using Squassh?

I first used the Squassh technique to identify the location of a serotonin receptor in the cell during my postdoc at the PSI. Based on the data we obtained, we worked out a new hypothesis about the function of two variants of this receptor that are linked to addiction mechanisms in the human body. InterAx uses the software in combination with the biosensors to measure physiological outcome. For example, we have established the use of the biosensors for high-throughput microscopes for the rhodopsin receptor.

Where else can this technology be applied?

Naturally, the Squassh technology can be used in other fields of cell biology to obtain better data from fluorescence microscopy. For instance, the software can be used to identify mitochondria forms in the cell, whether they are round or elongated, which can give a new push to fundamental research and drug discovery for mitochondria-related diseases, such as cardiovascular or metabolic diseases. Besides that, the software can also be used to study the infection of cells by viruses by monitoring the internalization of virus particles in the cell. Finally, Squassh technology has a lot of potential to help us identify new insights for developing new or better medicine, but at the moment this is still the outlook for the future.


References

[1] Rizk, A., Paul, G., Incardona, P., Bugarski, M., Mansouri, M., Niemann, A., et al. (2014). Segmentation and quantification of subcellular structures in fluorescence microscopy images using Squassh. Nature Protocols, 9(3), 586–596.