Subprojects

Cardiovascular, neurological and metabolic research on campus make increasing use of cell sequencing, which entails a transition to very large and complex data sets. In order to take future cross-disciplinary coordinated steps towards translational-focused research based on such datasets, reliable and scalable analysis pipelines will be established as a standardized platform and corresponding internal and external datasets will be employed for continuous validation and further exploration. The focus here is on the integration of single cell analyses on transcript or chromatin level with further data sets (e.g. results from experimental models, functional assays on in vitro isolated cell populations, as well as investigations by means of proteomics and metabolomics), especially on the level of pathways and biological networks.

Principal Investigators:

• Prof. Dr. Maria Grandoch
• Prof. Dr. Gunnar Klau

Co-Principal Investigators

• Prof. Dr. Jens W. Fischer
• Prof. Dr. Norbert Gerdes
• Prof. Dr. Axel Gödecke
• Prof. Dr. Jürgen Schrader
• Prof. Dr. Dr. Dr. Sven G. Meuth

Postdoc

• Dr. Alexander Lang

Cardiovascular pathological processes (e.g. atherosclerosis, aneurysm, myocardial infarction, stroke) are often accompanied by infiltration of lesions by different immune cell populations (e.g. T cells, macrophages), which thereby change their functional phenotype and metabolic activity. Such changes will be analyzed via already collected and future single-cell RNA sequencing data (related to P1) from experimental models of cardiovascular disease. In particular, approaches from research on the metabolism of plant and bacterial cells will be transferred to mammalian and ultimately human cells, starting from an integration of transcriptome data with metabolome data from functional metabolic models The aim of this transdisciplinary approach is to elucidate links between disease-associated metabolic phenotypes of specific immune cells, their gene expression, epigenetic accessibility, and localization in the corresponding diseased tissue (spatial transcriptomics).

Principal Investigators:

• Prof. Dr. Norbert Gerdes
• Prof. Dr. Martin Lercher

Co-Principal Investigators

• Prof. Dr. Jürgen Schrader
• Prof. Dr. Björn Usadel
• Prof. Dr. Dr. Dr. Sven G. Meuth

Postdoc

• Dr. Tin Yau Pang

The effect of the metabolome on the cells and cell assemblies (morphome) of cardiovascular tissues, and vice versa of the morphome on the metabolome, leaves traces and patterns that can be detected by sensitive noninvasive imaging techniques (spectroscopy, parametry, attenuation indices). The detection of quantifiable patterns by multiparametric MRI in the pathophysiological chain of atherosclerotic vascular disease (coronary vascular inflammation, atherothrombosis, and ultimately end-organ damage) or myocarditis (myocarditis or graft rejection) is significant for risk stratification of patients*. Neural networks for segmentation and classification will be trained on the image datasets of the clinical and small animal experimental (CRC 1116, TR 259) cohorts and subsequently validated on second datasets. The algorithm will then be evaluated for robustness and extended to additional cohorts. In this way, the metabolic analyses of P2 will be continued in this project and then investigated in clinical cohorts in collaboration with P4 (see below).

Principal Investigators:

• Prof. Dr. Ulrich Flögel
• Prof. Dr. Stefan Harmeling

Co-Principal Investigators

• Prof. Dr. Malte Kelm
• PD Dr. Florian Bönner

Postdoc

• Dr. Iman Marivani

Cardiogenic shock is a life-threatening multisystem disease with high lethality caused by heart failure. It involves severe metabolic changes that ultimately leads to multiorgan failure. The influence and interaction of comorbidities such as diabetes, renal insufficiency, liver disorders, and physical frailty are still poorly understood. Multimodal analyses of metabolism with an interplay of unsupervised (to identify subgroups and characterize their trajectories) and supervised (to assign them to the appropriate groups) algorithms are needed. Cluster procedures will be used to identify robust subgroups and characterize them in terms of outcome. Growing time series will then be used to characterize the course of predictive performance and to identify the earliest possible time of assignment to outcome-relevant groups. Ultimately, both approaches will be combined by performing subtype assignment for new, individual patients and integrating it into the prediction.

Principal Investigators:

• Prof. Dr. Dr. Christian Jung
• Prof. Dr. Simon B. Eickhoff

Co-Principal Investigators

• Prof. Dr. Artur Lichtenberg
• Prof. Dr. Malte Kelm

Postdoc

• Dr. Nicolás Nieto

Macrophages differentiated in vitro from bone marrow monocytes of hyperglycemic, hyperinsulinemic mice exhibit altered polarization behavior. Equivalently, in the context of neuroinflammation, reprogramming of intracellular metabolism of microglia occurs, stimulating their immunological functions such as cytokine production and phagocytosis, which in turn further enhance the process of inflammation. This suggests the formation of a "metabolic memory" through epigenetic mechanisms. The aim of this research is to identify metabolically induced chromatin and transcriptional changes, initially in a mouse model, and to model them by incorporating pathway information. Machine learning methods (e.g. multi-task learning) will then be used to transfer this "metabolic signature" to humans. Thus, the conservation of this signature in monocytes from patients with hyperglycemia and subclinical inflammation will be investigated, its influence on circulating monocyte populations after acute myocardial infarction will be determined and thus tested for its prognostic utility.

Principal Investigators:

• Prof. Dr. Axel Gödecke
• Prof. Dr. Gunnar Klau

Co-Principal Investigators

• Dr. Peter Ebert
• Prof. Dr. Tobias Marschall
• Prof. Dr. Michael Roden
• Prof. Dr. Dr. Dr. Sven G. Meuth

Postdoc

• Dr. Mugdha Srivastava

Diabetes mellitus is a heterogeneous disease with differences in the risk of comorbidities and complications. Using simple clinical parameters, the main types of diabetes (type 1 and type 2 diabetes) could already be divided into subtypes (clusters). The cluster analyses allow precise identification of risk factors of rapid progression of complications. By setting up specific data analyses, (integrative) prediction models will be performed within the prospective longitudinal patient cohort of the German Diabetes Study. Here, we will use AI to integrate typical clinical parameters, complex metabolic parameters, as well as metabolomics and genomics data using metabolic maps for accurate identification of at-risk patient groups to enable tailored prevention, diagnosis and therapy of diabetes and its complications.

Principal Investigators:

• Dr. Oana Patricia Zaharia
• Prof. Dr. Björn Usadel

Co-Principal Investigators

• Prof. Dr. Michael Roden
• Prof. Dr. Dr. Svenja Caspers

Postdoc

• Dr. Christian Binsch

The cellular mechanisms of manifestation of diabetes and associated organ damage are still largely unclear. The analysis of the dynamic adaptation to metabolic conditions, as well as the individual self-regulation of metabolic processes in the light of genetic predisposition and multidimensional environmental conditions, enables the identification of metabolic switches that go beyond classical biomarkers to describe concepts of the pathophysiology of diabetes. Transformation and model-based integration of data generated from multi-omics approaches, visualization of tissue-specific dynamic flow rates at the cellular level using multinuclear magnetic resonance spectroscopy, respirometry and tracer dilution techniques are combined with intensive metabolic phenotyping of (pre)clinical models and parameters of interorgan communication. The aim is to develop therapeutic concepts for prevention of diabetes and its complications.

Principal Investigators:

• Dr. Sofiya Gancheva
• Prof. Dr. Tobias Marschall

Co-Principal Investigators

• Dr. Maria Apostolopoulou
• Dr. Daniel Dörr
• Prof. Dr. Martin Lercher
• Prof. Dr. Björn Usadel

Postdoc

• Dr. Polina Lipaeva

Skeletal muscle has a critical function in insulin-regulated glucose uptake, dysfunction of which precedes the development of type 2 diabetes mellitus. Exercise intervention improves insulin action in many patients ("responders") through mechanisms that are poorly understood. In this project, muscle cells from metabolically healthy and insulin resistant donors will be studied to investigate the influence of genetic predisposition and exercise intervention in the development of diabetes. To this end, genetic, genomic and epigenetic datasets (DDZ, BMFZ, WGGC) will be linked with mitochondrial energy metabolism, fluorescence microscopy and phosphoproteomics data from preclinical (cell/animal models, UKD) and clinical biospecimens (muscle biopsies, DDZ) to understand dysfunctional insulin action at the patient level in silico, allowing more precise therapy assignment in diabetes and novel pharmacological approaches.

Principal Investigators:

• Prof. Dr. Hadi Al-Hasani
• Prof. Dr. Tobias Marschall

Co-Principal Investigators

• Prof. Dr. Dagmar Wieczorek
• Prof. Dr. Axel Gödecke
• Prof. Dr. Andreas Reichert

Postdoc

• Dr. Stephan Majda

Metabolic liver disease (NAFLD) can progress to cirrhosis and eventually hepatocellular carcinoma (HCC) in some patients. Currently, it is unclear which patients are at particular risk of progression, making targeted surveillance and prevention measures difficult. In this project, deep-learning algorithms for risk prediction will be trained on histological biopsies from NAFLD patients and validated against histological sections from livers of healthy body donors, building on previous work in gastroenterological oncology. By means of "reverse engineering", histological areas in the sections will be characterized that have a particularly high predictive power in order to obtain conclusions about pathophysiological relationships ("explainable AI").

Principal Investigators:

• Prof. Dr. Tom Lüdde
• Prof. Dr. Dr. Svenja Caspers

Co-Principal Investigators

• Prof. Dr. Stefan Harmeling
• Dr. Jakob Nikolas Kather

Postdoc

• Dr. James Alex Brooks

The progression of NAFLD to NASH, fibrosis, cirrhosis, and HCC described in Project 9 is influenced by a number of risk factors, including genetic variants, and represents a potential target for future therapies. Building on the cohort available in the Department of Gastroenterology, a risk stratification will be developed based on genetic (PNPLA3, TM6SF2, GCKR, MBOAT7, HSD17B13, MARC1, etc) and metabolic factors, which will allow classification of patients into subgroups with low or high risk of progression or HCC. These risk scores will then be applied to a prospective cohort that is currently being established and systematically combined with the results from Project 9.

Principal Investigators:

• Prof. Dr. Verena Keitel-Anselmino
• Prof. Dr. Tobias Marschall

Co-Principal Investigators

• Prof. Dr. Tom Lüdde
• Prof. Dr. Dagmar Wieczorek

Postdoc

• Dr. Arda Söylev

In various models of experimental autoimmune encephalomyelitis and ischemic stroke, inflammatory responses lead to network damage in the central nervous system (CNS). In this process, the various CNS-resident cells but also infiltrating immune cells play a major role. At the same time, we know that inflammation is associated with metabolic shifts, which in turn can influence and epigenetically modify the immune system. In addition to immunological, histological and electrophysiological methods, interactions between cell compartments will be investigated by using transcriptome, proteome and metabolome analyses of the individual, isolated cell types. Through this combination of methods, a temporally and spatially highly resolved description of the pathophysiological processes will be achieved in order to derive appropriate molecular targets for diagnostics and therapy in the sense of translation. As a further translation into the clinic, proteome and metabolome signatures of multiple sclerosis and stroke patients with metabolic syndrome will be identified, sorted by BMI quartiles and correlated with the clinical course of the disease.

Principal Investigators:

• Prof. Dr. Dr. Dr. Sven G. Meuth
• Prof. Dr. Alexander Dilthey

Co-Principal Investigators

• Prof. Dr. Tobias Marschall
• Prof. Dr. Björn Usadel

One challenge in monitoring and predicting the course of chronic progressive or relapsing diseases is the relatively long intervals between physician contacts. However, if patients present themselves only when deterioration has occurred, preventive possibilities are limited. The aim of this project is to develop, implement and evaluate a new approach based on regular detailed pathophysiological characterization with continuous telemedical tracking via the JuTrack platform to develop predictive models for early detection of clinical deterioration. To this end, a new cohort will be established that will undergo regular clinical and immunological characterization (stratification of baseline risk) and then continuous follow-up (adjustment of prediction based on current condition). The integration of multi-omics data (transcriptomics, proteomics, metabolomics) may serve here the molecular understanding of underlying immunometabolic mechanisms, with the goal of identifying specific biomarkers (e.g., immune cell signatures or complement signatures) as clinical predictors as well as therapeutic target structures.

Principal Investigators:

• Prof. Dr. Dr. Dr. Sven G. Meuth
• Prof. Dr. Simon B. Eickhoff

Co-Principal Investigators

• Prof. Dr. Dr. Svenja Caspers
• PD Dr. Jürgen Dukart
• Prof. Dr. Stefan Harmeling

Postdoc

• Dr. Somayeh Maleki Balajoo