)
Wouldn’t it be better if you could do a biomarker profile while you’re still in the ambulance?
German version/Deutsche Version
During his studies in Leipzig, Frank Biedermann had hardly ever heard of the subfield of his subject in which he is now one of the internationally recognized experts. This is because supramolecular chemistry leads a shadowy existence in Germany. There are no dedicated chairs for it, and professors of organic chemistry tend to cover it on the side, in contrast to France, for example, where Jean-Marie Lehn researched and taught in Strasbourg and was awarded the Nobel Prize in Chemistry in 1987 together with two American scientists for his work on the “development and use of molecules with structure-specific interactions of high selectivity”. Supramolecular chemistry is interested in interactions between different molecules that make no effort to bond permanently with each other – while traditional chemistry investigates under which conditions, how and with what result atoms and molecules can bond permanently with each other. Supramolecular chemistry thus has a special relationship to biology, since, for example, the main role between enzymes and their substrates is not played by solid bonds, but rather by their spatial fit, which is influenced and favored by electrostatic forces such as hydrogen bonds or hydrophobic conditions. It is therefore not surprising that the ever-deeper involvement with supramolecular chemistry has led Frank Biedermann further and further into biomedicine.
A momentous start gift
Biedermann’s interest in supramolecular complexes began by chance in Cambridge, England. In October 2007, he had arrived at the Melville Laboratory for Polymer Synthesis to write his master’s thesis. He was interested in how the properties of polymers change when molecules interact with each other. To his surprise, however, his supervisor Oren A. Scherman told him: “We’re no longer pursuing the polymer topic, because so many other groups are already working on it. Instead, investigate this barrel-shaped molecule here!” This molecule, called cucurbit[n]uril, had been discovered by Scherman in the literature. Friendly colleagues had given him samples when he started his own research group. His then master’s student and later doctoral student Frank Biedermann has not let go of this molecule since. He has contributed significantly to making it one of the most famous research and application objects in supramolecular chemistry.
Long forgotten molecules
Cucurbituriles are polymers of glycoluriles. The [n] in their middle stands for the number of their individual parts. The German chemist Robert Behrend synthesized them for the first time in 1905 but did not characterize them in detail. They were forgotten until their structure was elucidated at the University of Chicago in 1981. They resemble solid pumpkin-like cages that enclose a stable, inert cavity. Their ability to self-assemble makes them easy to produce. This makes them well suited as hosts for molecular guests and allows them to function as synthetic receptors. Host-guest relationships are a central field of research in supramolecular chemistry. Originally, it was thought that they followed the lock-and-key principle alone and that only structurally exactly matching guests would be let in by their hosts, explains Frank Biedermann. However, he says, this is only true for organic solutions, not for aqueous ones. “In organic solutions, you can build molecules that look like Olympic rings this way. But who does that help?” Host-guest chemistry would only become truly relevant if it also worked in aqueous solutions. A South Korean research group led by Kimoon Kim showed in 2000 that this is possible in principle with cucurbiturils. In his doctoral thesis, Biedermann brilliantly investigated how and why cucurbit[n]urils (CBn) can also be good hosts in aqueous environments.
The power of hydrogen bonds
His starting point was the question which guest molecules CBn accepts at all in aqueous solution. To answer this question, he measured the heat that developed when he titrated a possible guest molecule into a CBn solution. He benefited from instrumental advances that had made calorimetric methods sufficiently sensitive. “Cucurbituriles bind their guests so strongly that I had to measure at extremely low concentrations to even find an evaluable binding curve.” He booked the only instrument available for this in Cambridge in three-hour slots, sometimes at night. Systematically and without bias, he tested the host CBn‘s binding behavior with long lines of guests. Biedermann enjoyed the fact that his doctoral advisor gave him a lot of freedom in the process. To his surprise, he found that CBn are very good synthetic receptors for almost all possible guests, whether they are steroids or neurotransmitters, for example. The monopoly of the lock and key concept had run its course. “From my data, it became evident why this is so in aqueous solution,” Biedermann says. Water solvates more strongly than other solvents because of its dipolar structure. It forms hydrate shells around molecules that dissolve in it. The more a molecule is surrounded by such shells, the less a host such as CBn can incorporate it. But if instead water molecules fill the narrow, hydrophobic CBn cavity, they are themselves handicapped. They cannot form sufficient hydrogen bonds among themselves. This puts them in a high-energy state that is uncomfortable for them. “If I offer a ligand that can replace them, that water comes out and is satisfied to be able to crosslink better with other water molecules there,” Biedermann says. “And the energy released drives the ligand in.” This alternative binding model, he says, was the main finding from his doctoral work. From then on, it formed the basis for him to develop chemosensors for biological markers. [1]
Competition for the host molecule
One of the great hopes of supramolecular chemistry is that such chemosensors or suprasensors could become inexpensive and easy-to-use additions to standard medical diagnostics using antibody assays or chromatographic methods. To detect a specific substance in the blood, all that would be needed is a specific host molecule for that substance and the ability to derive and measure a signal from its binding that is proportional to the concentration of the substance. That’s easier said than done, as all the outstanding researchers who tackled this problem in the years following the rediscovery of cucurbituril learned, including Werner Nau at the University of Bremen, in whose lab Biedermann began his postdoctoral work in 2012 before continuing it with Luisa de Cola at the University of Strasbourg. That’s because blood, like other bodily fluids, contains many substances that compete for entry into the cavity of a host molecule, and CBn are not particularly choosy about who they let in. Only for substances with a relatively high blood level, such as glucose, are supramolecular chemosensors therefore already routinely useful. The substances that Frank Biedermann is interested in – neurotransmitters, individual amino acids, drugs and their metabolites – occur in the blood in concentrations thousands to millions of times lower. Armed with the knowledge and experience he had acquired in Cambridge, Bremen and Strasbourg, he began to track them down on his own in the fall of 2016 – as head of an Emmy Noether junior research group at the Karlsruhe Institute of Technology (KIT). 
Signals from the displaced guest
During his post-doc, Biedermann had learned to use dyes to detect the binding of an analyte to a host molecule. “Of course, it would be desirable for the receptor molecule to give a signal response directly when the analyte binds to it. Unfortunately, the CBn molecules are not optically active.” However, this circumstance can be circumvented by measuring the binding of an analyte not directly, but indirectly by the extent to which it displaces an indicator. This indicator is a reporter dye whose emission spectrum depends on the conditions in its environment. It is placed as a guest in the host space of CBn before the start of a measurement. During a measurement, it is displaced by the analyte. However, it is easy for other substances to displace the dye, giving false-positive signals. “The trick is to develop an indicator dye that on the one hand binds strongly enough to the receptor so as not to be susceptible to interfering substances, and on the other hand is weak enough to be displaced by the analyte that is to be measured,” says Frank Biedermann. With his research group, he succeeded in the first practical application of such an indicator displacement assay. He used it to determine blood levels of the Alzheimer’s drug memantine that are in the nanomolar range. Regular monitoring of this level in patients suffering from dementia is necessary to ensure their compliance.[2]
The valuable treasure of supramolecular data
Mathematical simulations of the binding behavior of host-guest complexes such as indicator dyes are an indispensable prerequisite for the development of applicable indicator displacement assays. Equally indispensable, Biedermann believes, will be machine learning to bring suprasensors into medical use more quickly. Google, he says, can hardly be used to search for the best sensor for a particular marker. That’s why he founded the SupraBank.org database. It is open to all supramolecular researchers. “We already have a lot of visitors looking at it, but still far too few putting their own research data in there.” But that’s important, he said, if the essence of knowledge is not to be lost. He said he hopes journals will do more to encourage their authors to contribute to databases like this one, which benefits the entire community. As part of the “SupraSense” project, for which the winner of an Aventis Foundation Life Sciences Bridge Award was recently awarded an ERC Consolidator Grant, SupraBank.org will be of great importance – both as a resource and as a repository for AI-assisted chemosensor development.
Suprasensors could save lives
“SupraSense,” Biedermann hopes, will also shed light on “how we need to build host molecules so that they bind strongly and selectively.” That’s where the old lock-and-key principle comes back into play. “We need the water molecules, but we also need to somehow wallpaper the inner cavity.” Restricting detection to small molecules made sense in this regard. That’s because antibody tests were far superior to chemosensors in detecting biological macromolecules. Chemosensors were probably much better at detecting neurotransmitters such as serotonin in urine, for example, and could thus also be used for the early detection of diseases. Abnormally high serotonin levels are indicative of carcinomas of the gastrointestinal tract. The amino acid tryptophan, on the other hand, may be a prognostic marker for sepsis, which is all too often detected only when all help comes too late. Of course, there is rapid diagnostic support in large hospitals, says Frank Biedermann, but not at the family doctor’s or in the ambulance. He experienced this himself a few years ago, when he was suddenly overcome by shortness of breath, fell over and shortly afterwards found himself in an ambulance with a suspected pulmonary embolism. In fact, he had only caught a kindergarten infection from his daughter and was dehydrated on a hot day. “Wouldn’t it be better if you could do a biomarker profile while you’re still in the ambulance?”
Author: Joachim Pietzsch, Wissenswort
Photos: © Uwe Dettmar
[1] F. Biedermann, V. D. Uzunova, O. A. Scherman, W. M. Nau, A. De Simone, Release of High-Energy Water as an Essential Driving Force for the High-Affinity Binding of Cucurbit[n]urils. J. Am. Chem. Soc. 2012, 134, 37, 15318-15323.
[2] S. Sinn, E. Spuling, S. Bräse, F. Biedermann, Rational design and implementation of a cucurbit[8]uril-based indicator-displacement assay for application in blood serum. Chem. Sci. 2019, 10, 6584-6593.