Making processes in the body visible
An oxygen sensor in a cell culture test
Varying degrees of oxygen supply in a three-dimensional tumour model in cell culture (white = more oxygen, violet = less oxygen) / Confocal fluorescence microscopy
To observe processes in the body, we use and develop sensors that allow us to make specific molecules in cells and tissue visible. These sensors generate signals such as fluorescence that we can detect with the aid of a microscope. One of the sensor systems we have developed reveals to what extent cells are supplied with oxygen. This is of particular relevance when researching diseases such as heart attacks or cancer. To do so, we introduce the DNA of the genetically encoded sensors into the cells whose oxygen supply we want to study. The DNA integrates into the genetic code of these cells, enabling them to produce fluorescent proteins and light up differently according to their degree of oxygen supply.
One challenge is the fact that most fluorescent proteins require oxygen in order to fluoresce and can only indicate an oxygen deficiency (hypoxia) indirectly by the absence of fluorescence. We therefore developed an oxygen sensor for the first time using the protein UnaG, which fluoresces independently of oxygen and can now more precisely indicate a hypoxic state by producing a clear fluorescent signal. The tumour cells shown here produce two fluorescent proteins: while UnaG can always light up, the other protein fluoresces only in the presence of oxygen – and will fluoresce all the more strongly the more oxygen there is. In other words, the sensor shows not only whether hypoxia is present but also the degree of oxygen deficiency. Based on the ratio between the two signals, we were able to calculate clear gradations in the oxygen supply of the cells. We tested this in human cancer cells growing three-dimensionally, which in cell culture approximately represent the behaviour of a tumour. Better oxygen supply (white) could be seen in the outer regions, while oxygen was less well able to reach the centre (violet).
Nadine Bauer and Friedemann Kiefer
European Institute for Molecular Imaging, University of Münster
This study was conducted in cooperation with Cristian A. Strassert, Institute of Inorganic and Analytical Chemistry at the University of Münster.
Tracking down bacteria in the body
Lab insights
Symbolic representation of how our radiotracer functions: the tracer molecules (chemical structure) are transported with the blood through a blood vessel. Bacteria (green) that have colonised in the vascular wall absorb the tracer. The brightly shining spheres symbolise the radioactivity of the gallium-68 atoms we incorporate into the tracer molecules.© Philipp Bozdere, Andreas Faust. Chem Commun (Camb). 2024; 60(26): 3459–3596. DOI: 10.1039/d4cc00255e 
When chemically developing radiotracers, it is important for the tracer to become quickly distributed in the organism and for it to accumulate specifically at the structures we want to examine. We therefore design molecules “on paper” and then manufacture them in the lab.© Uni Münster - Michael Kuhlmann 
We use a cyclotron accelerator to produce the radioactive nuclides for our tracers. We have one such cyclotron in the basement of the Multiscale Imaging Centre. It is surrounded by a moveable shield for protection against radiation.© Uni Münster - Michael Kuhlmann 
Once we have synthesised a potential new radiotracer, we first check whether it will remain stable when in contact with blood in a test tube, and then examine how well the tracer will be absorbed by various bacteria in culture (picture).© Uni Münster - Michael Kuhlmann 
In targeted experiments with mice, we first assess how the tracer is distributed in a healthy organism (biodistribution). Then we check how the tracer accumulates at an artificially induced skin infection in the shoulder.© Uni Münster - Michael Kuhlmann
A newly developed tracer (yellow signal) has accumulated in bacteria in the right shoulder of a mouse. It also builds up in the bladder before being excreted. / Rendered 3D image of PET-CT data – positron emission tomography (PET, yellow signal) and computer tomography (CT, grey signal)
Bacterial infections pose a major challenge for modern medicine. Many bacteria can “hide” in a biofilm, making it impossible to find them in the blood. If they settle on injured tissue or prosthetics and spread around the body, the consequences can be very serious. We are developing radiotracers that could allow sources of infection to be detected at an early stage and their location pinpointed. Radiotracers are radioactively labelled substances that, once injected, are distributed around the body via the bloodstream and bind to defined structures or accumulate in cells. Even small quantities are sufficient for specific tomographs to make their radioactive signals visible.
To track down bacteria in the body we make use of their natural metabolic processes – in this example their iron metabolism. Bacteria produce specific molecules called siderophores that allow them to obtain iron from their surroundings. We create such siderophores synthetically and “load” them with the radioactive metal gallium-68. Because this metal behaves exactly like iron, chemically speaking, bacteria absorb our tracers via the same transport route. Once we have synthesised a new radiotracer candidate, we use a test tube and bacterial culture to check whether it will remain stable in the blood serum and ascertain how well various bacteria absorb it. Then we carry out targeted investigations with individual mice. The blood flow, nutrient concentrations and speed at which bacteria multiply during an infection are all factors that can influence how well a tracer will function in the organism. We have been able to demonstrate that the radiotracer shown is specifically absorbed by Escherichia coli bacteria in particular. We are now testing whether this will also happen with other types of bacteria that we know use this siderophore to absorb iron. At the same time, we are working on tailoring the tracer to clinical quality requirements so that it can begin to be used during patient examinations.
Renato Margeta and Andreas Faust
European Institute for Molecular Imaging, University of Münster
Silke Niemann
Institute of Medical Microbiology, University Hospital Münster
This project is being carried out in collaboration with Sven Hermann and Sonja Schelhaas at the University of Münster’s European Institute for Molecular Imaging.
Molecular imaging in a breast cancer patient
Clinical imaging in a patient with a breast tumour and a metastasis in a lymph node in the armpit. We used a newly developed tracer to conduct the examination (light signal). The kidneys are also visible, as is the ureter via which the tracer is excreted. / Positron emission tomography (PET)
Images of the inside of the body provide information that is crucially important for patient care. In nuclear medicine we use low-dose radioactive tracers to generate images. This enables us to reveal molecules that are involved in specific transport and metabolic processes or active in tumour tissue. The new tracer shown here makes a molecule visible that occurs on the surface of certain cells in the microenvironment of tumours – on the cancer-associated fibroblasts. Colleagues in our team chemically developed the tracer together with cooperation partners. Having evaluated its benefits, first in studies with mice and then in patients with breast cancer, we now regularly use the tracer in the hospital.
Fibroblasts are connective tissue cells that are to be found in every tissue and organ. Tumours take advantage of these and many other cells, which enables them to grow and metastasise. In the environment around a tumour, especially breast cancer, fibroblasts change their form and function and produce something called the fibroblast activation protein (FAP). The new tracer binds to this protein. The image shown here is from one of the first examinations conducted with it. We judged the risk of metastasis to be very high in this patient because the tumour had grown considerably and we had already identified a lymph node metastasis. The new imaging technique was used to confirm that, fortunately, no further metastases had formed. The tracer has enabled us to detect new metastases in other patients, better assess the overall growth of the tumours and adapt treatment accordingly. In our research we are now exploring whether FAP images taken during the early stages of breast cancer could allow us to distinguish invasive from non-invasive forms and thus tailor early treatment to the patient’s precise needs. Colleagues in New York meanwhile are investigating whether FAP imaging could enable us to predict a patient’s response to chemotherapy with sufficient accuracy as to render breast or axillary surgery unnecessary in some cases.
Philipp Backhaus and Michael Schäfers
European Institute for Molecular Imaging, University of Münster & Department of Nuclear Medicine, University Hospital Münster
We conducted the clinical studies at Münster University Hospital.
Tracking immune cells in the body
The traces (red) of immune cell movement in the brain of a mouse. We are studying how these movement patterns change as a result of inflammation and cancer. We tested various imaging techniques in an attempt to increase the resolution. / Magnetic resonance imaging (MRI)
Video: Time-lapse images reveal the individual movements of immune cells (red arrowheads).
We can use MRI to make immune cells visible with the aid of iron-based contrast agents. Immune cells regard the iron particles as foreign invaders and annihilate them by absorbing them. The image shows the traces of immune cells patrolling slowly along short or longer sections of blood vessel walls. If an inflammation is present, they start moving more quickly and purposefully in a specific direction. To date it has not been possible to track this more rapid movement in MRI: it causes blurring in the image or exceeds the measurement boundaries entirely – meaning that fewer cells can be seen in the image. By counting the cells, we were able to conclude in experiments with mice that immune cells change their behaviour in response to inflammation and cancer even before any clinical symptoms occur, and that the pattern of their movements correlates with the severity of the disease.
To analyse these processes even more precisely, we tested various imaging techniques. MRI generates images using switchable magnetic fields and radio waves to which hydrogen atoms in the body tissue react. We can vary certain physical parameters (creating MRI sequences) in order to obtain different image contrasts and imaging times. To perform a quantitative comparison of different imaging techniques, we built a model of the cell movements: we placed iron particles in a tube with a gelatinous gel (agarose) and rotated the tube in a very slow and controlled fashion during the MRI scan. Using various imaging techniques, we were able to improve the resolution of the movement speed to a maximum level of 0.8 millimetres per minute. However, this is not yet sufficient to visualise fast-moving immune cells. Further improving the imaging method is thus an important part of our research. Ultimately, we want to discover exactly what role the cells play in the disease process and whether the changes in movements we observe can allow us to draw conclusions for diagnostic and therapy monitoring purposes.
Enrica Wilken, Max Masthoff, Anne Helfen and Cornelius Faber
Clinic for Radiology, University of Münster
- Wilken E, Havlas A, Masthoff M, Moussavi A, Boretius S, Faber C. Radial compressed sensing imaging improves the velocity detection limit of single cell tracking time-lapse MRI. Magn Reson Med. 2024; 91(4): 1449-1463. DOI: 10.1002/mrm.29946
- Masthoff M, Freppon FN, Zondler L, Wilken E, Wachsmuth L, Niemann S, Schwarz C, Fredrich I, Havlas A, Block H, Gerwing M, Helfen A, Heindel W, Zarbock A, Wildgruber M, Faber C. Resolving immune cells with patrolling behaviour by magnetic resonance time-lapse single cell tracking. EBioMedicine. 2021; 73: 103670. DOI: 10.1016/j.ebiom.2021.103670
Simulating the signals of moving cells in the body
A digital representation of the human body. The coloured dots symbolise moving cells or molecules that emit radioactive radiation. We use simulations of this kind to develop mathematical models for biomedical imaging with positron emission tomography (PET). / 3D computer simulation
Video: In our simulation, 100 radioactive molecules are introduced into the bloodstream. They migrate via the blood to the heart, then to the lung and end up being distributed throughout the body.
Biomedical images are not a direct visual representation of the interior of the body; rather they are created by artificially generating signals such as light, sound or radiation and using various technologies to measure them and convert them into images. The images are thus a reconstruction of the internal view of the body. PET creates biomedical images using signals from radioactive tracers. With the aid of this technology we want to make it possible to track individual cells and molecules in the organism. This could be relevant, for example, when studying inflammation, in which immune cells migrate from the blood into tissue.
Our colleagues are developing complex mathematical models that can process the signal data from moving cells. We are working on ways to experimentally verify how well these models work. These include computer simulations with which we digitally replicate biological scenarios. We visualise the blood vessels with their diameters and typical branching patterns, as well as the speed and direction of the blood flow. We calculate on the basis of the laws of physics how cells or molecules will move in the blood. In the example here we have simulated radioactive sugar molecules (18F-FDG). As 18F-FDG is used as a PET tracer for clinical applications, we know precisely how these molecules behave in the body. The red molecules circulate in the blood, the yellow molecules are in the capillary bed in which substances are exchanged between blood and tissue, and the green molecules are metabolised in a cell. We calculate probabilities for these movement patterns and input the data into the simulation accordingly. The next step is to assign radioactivity to the cells. We also simulate the measurement of radioactive decay on the computer using a digital PET scanner. We compare the image reconstructed with the new algorithms and the simulated image, which allows us to assess how accurately we have identified the position of the cells.
Nils Marquardt and Klaus Schäfers
European Institute for Molecular Imaging, University of Münster
This is a joint project undertaken together with Marco Mauritz and Benedikt Wirth from Applied Mathematics Münster: Institute for Analysis and Numerics, University of Münster.
Moving cell in a tube phantom
Mathematical reconstruction of the signals emitted by an artificial cell moving through a tube during a tomography scan / Positron emission tomography (PET)
Video: Reconstruction of the cell movement over time
Biomedical images generated by positron emission tomography show the spatial distribution of radioactive tracers used to label specific cells or molecules in the body. The tracers produce signals because radioactive atoms are unstable. They decay over time and emit particles in the process – positrons in the case of PET. When a positron interacts with an electron, two gamma rays are emitted in opposite directions. A ring of detector crystals captures the gamma rays – whenever two gamma rays are captured by two detectors simultaneously, the original site of positron decay can be assumed to lie on a line connecting the two detectors. If the signals are not emitted from a fixed point but by cells moving around the body it is much more difficult to calculate images on this basis. For this purpose we are developing mathematical models.
We use the optimal transport theory to incorporate knowledge about how a cell moves from one place to another into our mathematical models. Of all possible cell movements, we assume that the one requiring the least kinetic energy will be the most realistic. A new mathematical model was tested here in an initial experiment. For this purpose we developed a phantom – that is to say an experimental laboratory setup that mimics a biological scenario. To simulate moving cells, we added radioactivity to alginate, which is used for example by dentists to take impressions of a patient’s teeth, and cut small cubes (1 x 1 x 1 millimetre) from it. We placed a silicon tube with several coils into the PET and carried out a scan while our alginate cells were pumped through the tube in oil. Because the artificial cells contained significantly more radioactivity than would be realistic in an organism, we used only around five percent of the detected radioactive decay for the image reconstruction. We are now working on optimising the computation speed of our algorithms and plan in future to use them together with biologists and doctors to answer various questions about cell behaviour in the body.
Marco Mauritz and Benedikt Wirth
Institute for Applied Mathematics: Analysis and Numerics, University of Münster
Nils Marquardt and Klaus Schäfers
European Institute for Molecular Imaging, University of Münster