Wissenschaftliches Bild

Start-up support for translating ideas into practice

Three MDC teams have now been awarded grants from the BMBF for their ideas. The grants aim to make laboratory discoveries available to patients as quickly as possible. The focus is on creating new approaches for diagnosing and treating cancer and multiple sclerosis.

Three-dimensional maps of the gene activity of tumor cells and the surrounding tissue, new immunotherapies against cancer that unlike previous approaches work without using viruses, and the use of artificial intelligence to better detect and understand changes in the brains of multiple sclerosis (MS) patients – this is what the research projects at the Berlin-based Max Delbrück Center for Molecular Medicine in the Helmholtz Association (MDC) are all about that are now being funded by the German Federal Ministry of Education and Research (BMBF) for either one and two years. The financial support is intended to help the teams speed up the translation of their basic scientific findings into clinical practice.

3D maps of gene activity

Special software assembles wafer-thin tissue slices, in each of which the gene activity of individual cells has been analyzed, into three-dimensional images. In this spatial image of a mouse brain different colors characterize a small number of exemplarily genes.

The 3DGenes (Three-Dimensional Molecular Pathology for Personalized Medicine) project is being supported by the BMBF through its “GO-Bio initial” program for one year to the tune of €100,000. A three-member team from the MDC’s Systems Biology of Gene Regulatory Elements Lab led by Professor Nikolaus Rajewsky is exploring starting a company that will do things like help develop better treatment strategies for cancer patients. Along with PhD student Jonathan Alles, project leader Dr. Nikolaos Karaiskos and postdoc Dr. Marvin Jens are involved in the project. The researchers are working together with other members of the lab to create three-dimensional maps of gene activity in tumors and surrounding tissue.

“Diseases arise as a result of complex changes in cells and tissues,” explains Jens. “To understand them better, we’re not only analyzing individual cells, but also examining how cells interact at the boundary between healthy and diseased tissue – in the case of cancer, the interaction between the tumor and the immune system.” In Rajewsky’s lab, the team is developing computer models that help them study how cells communicate and identify molecular changes that can only be found in a tissue context. “The insights this provides into the development and progression of cancer have far-reaching potential for improving diagnostics and creating personalized therapies where, for example, the patient’s own immune system fights the tumor,” says Jens.

Creating immunotherapies without viruses

Another project aimed at developing new and improved immunotherapies for treating cancer is syncDNA (Innovative DNA Technologies for Non-Viral Immunotherapies), which is also receiving €100,000 in funding for one year through “GO-Bio initial.” Led by Dr. Jan Pille, head of the MDCell – Helmholtz Innovation Lab, the project’s team is devising strategies to equip the T cells of the immune system with tailored molecular structures that fight tumors more effectively. The team is not using viruses for this purpose. Instead, the researchers are searching for approaches that employ enzymes alone to provide the immune cells with the necessary information in the form of DNA for attacking the cancer tissue in a targeted way.

“We are working on producing the requisite DNA fully synthetically, that is, without the help of bacteria,” explains Pille, adding that genetic material will also be introduced into the immune cells as a “taxi cab” with no viruses on board and will contain the blueprints for generating tumor-specific surface receptors on T cells that are capable of latching onto the cancer cells. “To do this we are using an enzyme system called ‘Sleeping Beauty,’” explains Pille. “This recognizes the artificial DNA and incorporates it into the genetic material of the T cells, causing them to start producing the receptor they need to fight the cancer cells.” Pille also points out that eliminating the need for bacteria and viruses makes personalized therapies, which are tailored completely to the respective tumor, safer, faster, and more cost effective.

Smart tools for radiology

The third BMBF-funded project at the MDC, SyReal (Synthesis of Realistic Data for Applicable Artificial Intelligence in Medicine), will initially focus on changes in the brain and other vital organs. “We will replicate variations of these changes in synthetic MRI data to train algorithms to unlock the secrets of diseases like multiple sclerosis faster and earlier,” says the project leader at the MDC, Dr. Sonia Waiczies of Professor Thoralf Niendorf’s Experimental Ultrahigh-Field MR Lab. Under the BMBF guideline “Generation of Synthetic Data for Artificial Intelligence,” SyReal is receiving nearly €1.7 million in support over two years, of which around €250,000 will go to the MDC. The coordinator of the project, which also involves industrial partners, is Professor Christoph Lippert of the Hasso Plattner Institute for Digital Engineering in Potsdam.

“We’re using AI techniques to develop special algorithms that will help doctors more easily detect disease-related changes in tissue,” explains Waiczies. “Such algorithms must be able to recognize the features of rare diseases as well as be robust to external confounding factors.” Last year, Waiczies and her team demonstrated that a new pathological finding in an animal model of multiple sclerosis could also be observed in patients. “Our results indicated that pathologically relevant macroscopic changes in brain ventricles require brain MRI data to be analyzed more thoroughly,” notes Waiczies. In the future, such an analysis will be performed using realistic data generated by machine learning techniques to compare synthetic images of the organ.

“Multiple sclerosis is a very complex disease, and complex methods are required to identify features of the disease in the brains of patients that the human eye can’t detect,” says Waiczies. “With the algorithms we’ve developed, we hope to give radiologists and pathologists a tool to help them make an exact diagnosis.”

Text: Anke Brodmerkel