Highly Informative Drug Screening by Overcoming NMR Restrictions
European Research Council Synergy
funded project N° 951459
2021-2026
Overview
The need for drug screening with increasingly higher throughput is dictated both by the increasing number of drug targets that are becoming available through genomics and by the increasing number of chemical molecules generated through combinatorial chemistry. Several high-throughput screening techniques exist that can scan large libraries of compounds, but the ever-increasing throughput has not translated into a significant increase in promising drug candidates. HiSCORE presents a synergistic approach to high-throughput, high-information drug screening that builds on the complementary skills of four laboratories supported by two external experts of drug screening:
(i) Research and design of innovative instrumentation for magnetic resonance that can provide small, hyperpolarized solid samples at intervals on the order of minutes, transfer and dissolve or liquefy these samples with minimum dilution, and acquire several high-resolution NMR spectra of the liquid samples in parallel, using complementary contrast-enhancement methods, in up to 1000 massively parallelized microfluidic detectors. The contrast between compounds that bind to targets and those that fail to bind can be boosted by exploiting long-lived states.
(ii) Applications of this instrumentation for binding assays and to measure the dissociation constants in the nano to micromolar range, and to determine kinetic rates of the association and dissociation for a large number of complexes of putative drug compounds and protein targets. (iii) Functional assays, in particular for systems that comprise multiple enzyme steps with intermediate products, to determine the efficacy of potential inhibitors, while fully exploiting the rich information that can be obtained by fluorine-19 NMR. (iv) Metabolomic assays to observe the metabolism of the compounds in cultures in cells in view of identifying potentially toxic side-products.
Partners
“HiSCORE” builds on the synergy of 3 institutions (beneficiairies): ENS (France), SKU (Netherlands), and KIT (Germany), the latter being represented by two different teams:
1. Geoffrey Bodenhausen ENS Paris
2. Benno Meier KIT Karlsruhe
3. Arno Kentgens SKU Nijmegen
4. Jan Korvink KIT Karlsruhe
The 4 groups contribute complementary expertise:
Geoffrey Bodenhausen and the pre-existing team at the Ecole Normale Supérieure in Paris (Abergel, Ferrage, Pelupessy, Baudin, Bouvignies, Birlirakis) develop NMR methodology, in particular by exciting long-lived states to extend the life-times of hyperpolarization, to improve the accurate determination of binding parameters, including kinetics, and to develop libraries of compounds.
| Surname | Last Name |
| Geoffrey | Bodenhausen |
| Philippe | Pelupessy |
| Mathieu | Baudin |
| Kirill | Sheberstov |
| Anna | Sonnefeld |
| Sebastiaan | Van Dyck |
| Razanahoera | Aiky |
| Wiame | Coline |
| Loutfi | Hadi |
Recruiting and appointing new staff: Anna Sonnefeld, Kirill Sheberstov, Sebastiaan van Dyck, Aiky Razanahoera, Coline Wiame, and Dr Hadi Loutfi. Kirill Sheberstov has been promoted to become a chargé de recherche du CNRS since January 2023 and is therefore no longer supported by the ERC, but continues to work in our laboratory. On the other hand, Stéphanie Toetsch no longer helps to manage the network. Her functions are now fulfilled by Alain Cimino.
Integrating pre-existing staff: Philippe Pelupessy, Guillaume Bouvignies, Mathieu Baudin. The first two carry responsibilities as directeurs de these, the latter has played a major role in ordering components for a rapid liquid-state DNP probe that is being assembled by the team of Prof. Sami Jannin in Lyon
Purchasing of commercially available instruments. Since the start of the ERC project, we have installed a 500 WB magnet, a 9.4 T superconducting magnet with its cryostat, probes, a microwave source, and a power supply to drive a magnetic tunnel. We have also installed a 60 MHz table-top NMR spectrometer and adapted its software to our needs. We have contributed to the design, development and evaluation of a four-channel liquid-state probe in collaboration with Voxalytics GmbH.
The system consists of a cryogen-free magnet with a helium recirculation system (white), a microwave source (yellow), an NMR console (green), a control unit (blue), and a vacuum manifold on top of a pump (purple). Courtesy of Prof Kong Oi Tan
Awards and recognitions
| Award type | Title of award | Recipient of award | Year |
| Prize for lecture | ENC prize | Anna Sonnefeld | 2023 |
| Travel grant | ENC | Anna Sonnefeld | 2023 |
Benno Meier and his pre-existing team (Kourilova, Kouril): Alternatives to dissolution DNP: ballistic transfer of solid hyperpolarized “bullets” with minimal dilution, improved understanding and acceleration of the DNP process, parallelization.
| Surname | Name | Institution / department |
| Benno | Meier | KIT/IBG |
| Michael | Jurkutat | KIT/IBG |
| Masoud | Minaei | KIT/IBG |
| Pooja | Pooja | KIT/IBG |
| Pooja | Singh | KIT/IBG |
Recruiting and appointing new staff: The research team of PI Meier has grown continuously over the initial period of the Synergy grant. Prior to the start of the grant, the team comprised Meier and three postdoctoral researchers (Hana Kourilova, Karel Kouril and Michael Jurkutat). In late 2021, we hired Pooja Pooja as PhD student to work on ligand-binding, and Dr Masoud Minaei as postdoctoral researcher to work on automated DNP. In May 2022, we hired Pooja Sing as PhD student to work on optimal polarization and polarimetry. In February 2023 we hired Dr Kajum Safiullin as postdoctoral researcher to work on CO2 and Xe based polarization matrices. Additionally, we hired Dmitrii Zasukhin as PhD student (funded by a DFG project). Dr Karel Kouril and Dr Hana Kourilova, both postdoctoral researchers, have benefitted from parental leaves..
Installation of new equipment. Since the start of the ERC project, we have installed a 400 WB magnet, and a 9.4 T superconducting magnet. In 2023 we installed new chillers to provide sufficient cooling power for the 9.4 T superconducting magnet. We have also purchased a flow cryostat which we are now using to develop DNP instrumentation, as well as various items for microwave generation: an arbitrary waveform generator, a power signal generator and a power supply for a klystron.
Purchasing of commercially available instruments: 400 MHz WB, flow cryostat, development of purpose-built instrumentation, a 9.4 T polarizer with its microwave source, an arbitrary waveform generator, and a signal generator
Development of a four-channel liquid-state probe in collaboration with Voxalytics GmbH.
Development of automated bullet-DNP experiments with automatic sample loading and ejection for unsupervised DNP experiments (in progress)
Installation of microwave sources (Meier, Janssen, Kentgens; WP2, Task 2.1). We have opted to install a 95 GHz klystron from the lab of Kentgens in the IBG lab. We are confident that the high peak power of this system (100 W) will enable the rapid polarization of protons using narrow-band radicals that allow one to achieve hyperpolarization by exploiting the solid effect. To do so, we required a dedicated power supply which we purchased from CPI. This power supply has been delivered only in June 2023, and we expect first results in summer 2023.
Sketch (a) and photograph (b) of the injection device. The sample is placed into a small Teflon bucket (the“bullet”, light blue), and polarized using the instrumentation (not shown). The NMR tube is prefilled with solvent and fixed inside the 3D-printed NMR tube adaptor using Teflon tape. After polarization has built up sufficiently, the tunnel between the polarizer and the injection device is energized to provide a field of approximately 60 mT, and the sample is shot through the tunnel into the injection device within 70 ms by using pressurized helium. Small holes in the venting union allow helium gas to escape. A steel tube is connected to the bottom of the venting union. A small brass stopper with a constricted diameter is brazed to the bottom of this steel tube. As visible in the magnified view on the bottom left, the bullet itself cannot pass the brass stopper, but the “naked” sample (red) travels by inertia through the constriction, is fragmented upon impact on the liquid surface, and dissolves upon its immersion in the ambient temperature solvent that is contained in the NMR tube. The NMR acquisition is started automatically 1.4 s after the ejection of the sample from the polarizer.
Arno Kentgens and his pre-existing team (Marco Tessari, Gerrit Janssen, Hans Janssen, Ruud Aspers): Novel miniaturized NMR detectors, microliter samples, microfluidics, quantitation of compounds in nM concentrations, rapid-melt DNP, supercritical CO2 as a solvent, para-hydrogen induced polarization (PHIP), iridium complexes, zero-quantum spectroscopy, Overhauser DNP.
| Surname | Name |
| Arno | Kentgens |
| Kawarpal | Singh |
| Bono | Jimmink |
| Marco | Tessari |
Appointment of new staff: Bono Jimmink, Mattia Negroni, Marie Harder
Integration of pre-existing staff: Marco Tessari, Gerrit Janssen, Hans Janssen, Ruud Aspers
MRI/MRS for parallel acquisition of NMR spectra on multiple samples using multi-channel support (Singh, Aspers, Janssen, Kentgens, Ayuso-Penna, Korvink; WP5, Task 5.2.) Parallel NMR detection using MRI protocols on a multi-channel sample was tested at 950 MHz using chemical shift imaging. A 3D-printed structure consisting of seven identical wells and fitting a 5 mm NMR tube was used for this experiment. Each well could be filled with up to 3.3 ml of water. FLASH images were measured, where the water signal from each individual well was properly resolved. Longitudinal relaxation was measured within a single well, indicating the possibility of parallel determination of T1 and T2 on different samples, which might be of interest for screening protein-ligand binding. The material used for the 3D-printed structure used in these tests is resistant to water, but it is not compatible with organic solvents. The current setup is incompatible with manual filling of the wells. Future developments: a 3D design to address the wells filling issues is under design. Parallel NMR detection on solutions (different pH or concentrations of relaxation agents) will be performed.
An optimized NMR stripline for sensitive hyperpolarized NMR of microliter sample volumes
(Singh, H. Janssen, G. Janssen, Kentgens; WP3, Task 3.4.) A stripline NMR detector has been developed which features a novel design allowing an improved lineshape. An external lock circuit provides stability over time to perform signal averaging or multidimensional experiments. This detector is advantageous for rapid-melt DNP. We demonstrated the stability of the setup by recording two-dimensional spectra. Future developments: Development of a “rapid-freeze” probe using molecules hyperpolarized by PHIP as a source of polarization. Testing if the design can be parallelized.
Awards and recognitions
| Award type | Title of award | Recipient of award | Year |
| Knight in the order of the | Lion of the Netherlands | Arno Kentgens | 2023 |
Jan Korvink and his pre-existing team (Brandner, Mager, Lehmkuhl, Becker): miniaturization based on microelectromechanical systems (MEMS), complementary metal-oxide-semiconductors (CMOS), massively parallel detection, automatic data analysis, and machine learning.
| Surname | Name | Institution / department |
| Jan G. | Korvink | KIT/IMT |
| Mazin | Jouda | KIT/IMT |
| Neil | MacKinnon | KIT/IMT |
| Jürgen | Brandner | KIT/IMT |
| Dilara | Faderl | KIT/IMT |
| Francisco | Ayuso P. | KIT/IMT |
| Nourhan Abdelraouf | Abouelell | KIT/IMT |
| Omar | Nassar | KIT/IMT |
Appointment of new staff: Dr. Omar Nassar from Cairo, Egypt, was appointed to conceptualise the system design of our parallel probe designs. After a short period, Dr. Nassar decided to go to industry. Mrs. Nourhan from Cairo, Egypt, joined the team as CMOS designer. She made progress in designing an 8-channel parallel acquisition chip, as the first signal processing stage of a parallel cluster of detectors, performing amplification, filtering, and digitalisation. She is currently on maternity leave and will actively rejoin the project in the last quarter of 2023. PACO. We recruited Mrs. Dilara Faderl in the HiSCORE team to work on parallel NMR spectroscopy. We have recently recruited Dr. Dominique Buyens to join the team. She is an NMR spectroscopist from Pretoria, South Africa with a strong interest in in vitro drug screening, and will focus on one of our final work packages of HiSCORE when she starts in September 2023.
Integration of pre-existing staff: Senior scientist Dr Neil MacKinnon, an NMR spectroscopist, is actively involved in supervising the NMR activities of the HiSCORE group. Senior scientist Prof. Jürgen Brandner has joined the HiSCORE group to supervise the microfluidic activities. Senior scientist Dr. Mazin Jouda has joined the HiSCORE group to supervise activities involving NMR hardware, such as resonator design, CMOS design, MR imaging, and deep learning algorithm development. Senior scientist Dr. Dario Mager has joined the HiSCORE group to supervise the automation activities. Senior scientist Dr. Sören Lehmkuhl is supporting the team with his insight in spin dynamics, and background in hyperpolarisation techniques based on parahydrogen as a source of spin order. Mr. Moritz Becker, who is funded from another source, is contributing to the HiSCORE project, by developing a deep learning approach to parallel shimming.
Purchasing of commercially available instruments: There was no need to purchase an NMR spectrometer for the ERC project, as a suitable instrument was already available in the laboratory. Instead, the grant was used to purchase a state-of-the-art 3D nanolithography system, a Nanoscribe Quantum X, which is capable of structuring photopolymers with isotropic voxels of as small as 100 nm. We also ordered a double resonant four-sample parallel acquisition liquid state probe from Voxalytic GmbH.
Development of purpose-built instrumentation: One of the team members, Mr Achim Voigt, developed a digitally controlled 24 channel stable shim current source (6 currents per detector) for adjusting the field homogeneity of four-sample parallel acquisition liquid state probes, such as the Voxalytic probe and the team’s novel four channel probe. Two copies of the current design have been delivered to the Bodenhausen and Meier groups. PhD student Yen-Tse Chen has developed a four-channel parallel miniaturized stripline probe, equipped with built-in spherical harmonic shim coils.
Arrayed resonators and electronics (IMT + IMM, Task 5.1): We have raised the number of parallel detectors to 2 and then to 4, thereby exposing some of the challenges of parallelization, namely the isolation of RF channels to leakage from other nearby resonators, as well as the isolation of other magnetic field systems across measurement sites, such as shim coils, and pulsed field gradients, both of which pose a significant technical challenge. New concepts are currently emerging which aim to address these issues. We are also developing the necessary electronic front ends for the resonators, which will convert signals rapidly after reception to the digital domain, to protect them from further noise contamination. The next generation of parallel detectors will aim for 8 resp. 16 measurement sites, implying 16 resp. 32 radiofrequency channels. Indeed, given the space inside a conventional wide bore magnet, this will reach the limit of conventional approaches of signal handling, as well as the limit of RF isolation, so that novel approaches are now required.
Optimize parallel microfluidic sample preparation (IMT + IMM + Dalvit, Task 5.2): Microfluidic parallelization is in principle possible, but the major challenge that remains here is the reduction of flow friction, so as to achieve a very rapid transfer of sample to the NMR resonator. Our initial efforts are currently on hold as we search for a suitable team member.
Scaling-up to massively parallel detector array (IMT + IBG + IMM + ENS): Since we are currently considering imaging approaches, this array might become redundant in the future. As a result, we are currently studying the potential gains of each approach, before committing to a future concept.
Orchestrating parallel experiments (IMT + ENS + IMM + IBG): We will soon be in a position to perform our first parallel experiments, based on the parallel hardware from vendors (Bruker for the parallel-detection console, Voxalytic GmbH for the parallel probehead).
In addition to the consortium comprising these 4 principal investigators, two external advisors contribute to the project:
Alvar Gossert and related pre-existing teams at ETH (Roland Riek, Felix Torres): ): Alvar Gossert has worked for 10 years in the drug development/screening laboratories of Novartis in Basel, and in 2017 he returned to academia to ETH Zürich where he continues to develop NMR methods for drug discovery. An amendment to change his role from consultant to a partner of the consortium was signed on April, 24th 2023..
| Surname | Name |
| Alvar | Gossert |
| Niels | Lorz |
Claudio Dalvit has worked for 9 years in drug development/screening at Novartis in Basel, 9 years at Pharmacia & Upjohn in Milan, then at the Italian Institute of Technology (IIT) in Genova, and finally at the University of Neuchatel. Now emeritus professor.
Project objectives
(A) Development of novel instrumentation to produce hyperpolarized solid samples at intervals of 1 minute or less, and to parallelize 4, 10, or up to 100 microfluidic micro-coils or strip-line detectors to meet two objectives: (i) Study different sample compositions, in particular different concentration ratios, e.g. by varying the ratio [ligand] : [protein] = [L] : [P] by carrying out titrations. (ii) Functional screening by varying the ratio [inhibitor]/[enzyme]. (iii) Complementary experiments (T1, T2, CPMG, CEST, TLLS, TLLC…), with variable pulse intervals and relaxation delays. Observation of nuclei such as 1H, 2H, 13C, 15N, 19F, 31P, etc.
(B) Development of novel chemistry: develop binding assays to measure dissociation constants in the range nM < KD < µM that is important for fragment-based drug design, to determine kinetic rates kon and koff of the association and dissociation of ligand-protein complexes, to develop functional assays to determine IC50 of inhibitors of multiple-step enzyme cascades, and to explore metabolomic assays in cultures of cells.
Main results to date
Team of Geoffrey Bodenhausen
Evaluating and ordering new instrumentation. Purchasing a 500 MHz wide-bore liquid-state spectrometer equipped with 5 and 10 mm probes for 1H, 13C and 19F NMR, with a console capable of running 4 parallel experiments. Purchasing a 60 MHz ‘table-top’ liquid-state spectrometer. Development of MRI methods at 800 MHz using two samples in a concentric arrangement. Development and evaluation of a protype of a liquid-state NMR probe with four parallel samples. Purchasing of 9.4 T polarizer with cryogenics, probes, microwave sources, etc. Installation of a console provided by Jan Korvink to observe NMR spectra under cryogenic conditions. Installation of a home-built probe for “bullet DNP” provided by Benno Meier. We have agreed with the team of Prof. Sami Jannin in Lyon that we shall purchase a rapid liquid-state DNP probe that he has designed and developed.
Development of new experimental methods: Excitation and observation of long-lived states (LLSs) and of long-lived coherences (LLCs). Investigation of the spontaneous propagation of localized LLCs throughout aliphatic chains. Effects of paramagnetic relaxation agents on the lifetimes of LLLs and LLCs. Effects of protein-ligand binding on the lifetimes of LLSs and on the linewidths of LLC spectra. Contrast between lifetimes of LLSs and LLCs in pure ligands and in protein-ligand mixtures. Study of contrast in localized and delocalized LLS.
Discovery and detailed investigation with the team of prof. Kong Oi Tan (Bodenhausen’s successor at ENS), of so-called “hypershifted spins”, i.e., of protons that are in the immediate vicinity of paramagnetic centers, and that were hitherto believed to be unobservable and hence called “hidden spins”. It turns out that these protons are perfectly observable, albeit shifted by about 1.6 MHz from the center of the proton spectrum. Their relaxation properties have been characterized at temperatures below 4 K. These investigations have been carried out on the Cryomagnetic system that was purchased with ERC funds, with contributions by Kirill Sheberstov and Geoffrey Bodenhausen.
Team of Benno Meier
Evaluating and ordering new instrumentation: Purchasing of a 400 MHz wide-bore liquid-state spectrometer. Purchasing of a 9.4 T polarizer with cryogenics, probes, microwave sources, etc. Acquiring a power supply for a 100 W microwave klystron provided by Arno Kentgens. Acquiring a flow cryostat.
Developing new experimental methods: enhancing proton polarization by transfer from carbon-13 polarization by reverse INEPT (Insensitive Nuclei Enhanced by Polarization Transfer.)
Study of relaxation in hyperpolarized solids at low temperatures and low fields, and improvement of the theoretical understanding of DNP processes.
Development and testing of a probe for “bullet DNP” that reliably provides high polarization and rapid sample transfer. Development of injection systems. Evaluation of a home-built flow-through probe provided by Arno Kentgens. Applications to ligand-protein mixtures in association with Alvar Gossert at ETHZ. Applications to prenucleation of calcium carbonate solutions in association with Dennis Kurzbach (University of Vienna). Applications to long-lived states that are spontaneously populated at very low spin temperatures because of a violation of the high-temperature approximatin (deviation of linearity of Boltzmann’s law), in association with Bodenhausen’s team at ENS in Paris. The team of Benno Meier continues to develop an automated DNP system with automated loading and disposal in view of carrying out unsupervised screening experiments.
Team of Arno Kentgens
Evaluating and ordering new instrumentation: microwave source;
Developing new experimental methods: hyperpolarisation by hydrogenative or non-hydrogenative Parahydrogen Induced Polarization (PHIP). Transfer of hyperpolarization to target molecules in the solid state using a “rapid freeze” probe. Exploration of long-lived states using iridium complexes that can bind para-hydrogen, as suggested by the team at ENS in Paris. Transfer of such LLSs to ligands that are temporarily docked to the iridium complexes using selective polychromatic or non-selective cross-polarization. Propagation of LLS along aliphatic chains.
Team of Jan Korvink
Evaluating and ordering new instrumentation: NMR of miniaturized samples;
Developing new experimental methods: MRI of parallel samples;
Exploration of T2 contrast for variable ligand-protein ratios using MRI;
Development of novel methods to discriminate between enantiomers using a combination of electric and magnetic fields.This was not anticipated in the original proposal but could have considerable impact on the evaluation of drugs, for example by distinguishing between the R and S forms of the infamous drug thalidomide.
Major achievements
Research and technological achievements
Developments by the team of Geoffrey Bodenhausen
Improved contrast (Sonnefeld, Sheberstov, Razanahoera, van Dyck, Bodenhausen.) Long-lived states (LLSs) and long-lived coherences (LLCs) hold considerable promise for improving the contrast between free and protein-bound ligands. Aliphatic chains (R-CH2-CH2-CH2-R’) are ubiquitous in drugs, but it was hitherto believed that their protons could not be used to support LLSs or LLCs. We have developed methods that can readily achieve this. The contrast between relaxation rates of LLSs in aliphatic chains in free and bound drugs turned out to be much more dramatic (and thus more favourable for drug screening) than expected.
Simultaneous observation of LLSs in a mixture. The mixture contained ca. 10 mM ethanolamine, taurine, ß-alanine, GABA, and DSS in a 50 mM phosphate buffer in D2O. (A) Conventional 1H NMR spectrum of the mixture acquired with eight scans. (B) Spectrum acquired after a poly-SLIC sequence with five RFs indicated by wavy arrows in (A), with eight scans and an LLS relaxation interval 𝛕rel = 3 s. The spectrum is scaled by a factor of 50 with respect to (A). (C) LLS decays of four molecules measured simultaneously. The LLS lifetimes were determined from monoexponential fits, ignoring weak initial oscillations that can be neglected after 𝛕rel > 1.3 s: TLLS (taurine) = 20.6 ± 2.3 s, TLLS (ethanolamine) = 14.1 ± 0.7 s, TLLS (GABA) = 8.9 ± 0.3 s, and TLLS (DSS) = 5.9 ± 0.5 s.
Developments by the team of Benno Meier
UV polarizing agents (Meier, Singh, Zasukhin, Kouril: WP2, Task 2.2.) We benefit from a fruitful collaboration with Andrea Capozzi (EPFL) who is a leading expert on the use of UV-excited radicals. We are now using these routinely in our lab, for example for the creation of radicals in carbonate matrices (collaboration with Aharon Blank, Technion Haifa).
Cross-polarization and thermal mixing (Meier, Kouril, Jurkutat: WP2, Task 2.4.) We have studied in detail thermal mixing in pyruvic acid doped with trityl. For protons we were able to propose a parameter-free model that neatly describes the dynamics of thermal mixing. This requires only a single calibration of the size of the Non-Zeeman electron reservoir. In the case of 13C, the dynamics are constrained by slow nuclear spin diffusion. Experimental work on the combination of thermal mixing with DNP and cross-polarization is in progress.
Optimizing sample throughput (Meier, Minaei, Safiullin: WP2, Task 2.5) We have demonstrated automatic loading of up to 18 samples into the polarizer, as well as automatic cleaning of the injection device. Work on a hyphenation of these steps is ongoing.
Transfer through magnetic tunnels (Meier, Pooja, Kouril, Bodenhausen, Sheberstov: WP4, Task 4.1) Power supplies to drive solenoids have been purchased by both ENS and IBG. At IBG, we have also designed a magnetic tunnel using permanent magnets, in collaboration with Jonas Milani. This tunnel provides a field of 0.5 T for the straight sections of the bullet transfer.
Developments by the team of Arno Kentgens
Parahydrogenation of triple-bonds in quasi-symmetric molecules to generate hyperpolarized long-living singlet order in solution (Jimmink, Aspers, Tessari, Kentgens; WP3, Task 3.1). Homogeneous parahydrogenation of a symmetric alkyne (dimethyl acetylenedicarboxylate) has been performed to produce hyperpolarized singlet order in the product molecule (dimethyl maleate). Performing the reaction at high fields results in the rapid loss of singlet order and an enhanced emission signal of the vinyl protons originating from parahydrogen. This failure is attributed to the formation of a reaction intermediate in which the symmetry between the parahydrogen protons is broken, resulting in a conversion of singlet order to longitudinal spin order. Conversion of longitudinal spin order to hyperpolarized (negative) magnetization takes place via relaxation driven by cross-terms (interference) between the chemical shift anisotropy and the dipolar interactions of the vinyl protons. A kinetic model has been developed that quantitatively accounts for the formation of hyperpolarization on the vinyl protons originating from parahydrogen. Future developments: these results suggest that para-hydrogenation might be an efficient route to generate hyperpolarized singlet states in low magnetic fields to prevent losses of singlet order, and in sufficiently high magnetic fields to prevent avoided level-crossings that result in the conversion of singlet order into magnetization.
Developments by the team of Jan Korvink
Ligand binding revealed by T2-based MRI contrast of 19F (Faderl, Ayuso-Penna, Jouda, MacKinnon, Brandner, Korvink.) Enhanced T2 relaxation caused by ligand association with target proteins was used to determine binding characteristics by MRI. A ligand that is not bound to any protein in solution generates bright 19F MR images, which will lose intensity as the protein:ligand ratio increases. By measuring a series of solutions with varying the protein:ligand ratios, binding affinities can be extracted. The benefit of the MRI approach is that multiple samples, each with a different protein:ligand ratio, can be measured simultaneously in a single experiment. To date, the imaging protocols enabling this measurement have been optimized for up to 5 samples measured in parallel. In future developments, the number of samples measured in parallel can be increased by a factor 2. Work will continue on the optimization of sample holders enabling a high degree of measurement parallelization. MRI acquisition acceleration could be addressed, for example, by exploring alternative data sampling schemes.
Establishing a deep learning platform (IMT + ENS, Task 7.2): We have started to implement deep learning approaches, by first focusing on the use of deep learning for hardware optimization, in the present case, shimming to achieve a higher field homogeneity. Three publications have been prepared, two of which are already published. Our initial successes indicate that deep learning is a powerful approach that requires extensive data collection before it can be successfully applied.
Novel methodologies
One of the key aspects of drug screening by NMR is that one should be able to exploit a contrast between ligands that have an affinity for protein targets and ligands that cannot bind. It is well known in the field of drug screening by NMR that such a contrast can arise when observing linewidths (T2, best observed by Carr-Purcell-Meiboom-Gill or CMPG spin echo sequences.) We have found that a much better contrast can arise by observing linewidths of long-lived zero-quantum coherences (TLLC), or by observing the lifetimes of long-lived population imbalances (TLLS.) We have now demonstrated that one can hyperpolarize long-lived states directly using our bullet-DNP apparatus to achieve very low spin temperatures.
The excitation and reconversion of LLSs and LLCs by polychromatic spin-lock induced crossing (poly SLIC) was not planned in the proposal. The resulting methods provide very good contrast between free and bound drugs, but suffer from low quantum yields. It is therefore of paramount importance to boost the sensitivity by “hyphenating” SLIC with hyperpolarization. The success of this combination has been described in a recent publication.
Prior to the demonstration of bullet-DNP, it was believed that radical-induced relaxation would be prohibitively fast during the transfer of hyperpolarized solids. Within the ERC project, we have studied the relaxation of 13C-labelled pyruvic acid as a function of field and temperature. An innovative model based on the concept of spin temperature and triple-spin-flips accurately describes the relaxation over 2.5 orders of magnitude in magnetic field strength, and for temperatures up to 40 K. Our studies (J Phys Chem and Phys Chem Chem Phys) indicate that relaxation should be mostly limited to protons, provided one uses radicals with narrow EPR spectra. The ongoing installation of a powerful microwave source will serve to exploit this novel insight.
After dissolution, radicals such as Tempol or Fremy’s salt can be quenched by adding a solution of ascorbate (vitamin C) to the bullet prior to freezing.
For the determination of kinetic parameters, we have opted to run the first hyperpolarized ligand-binding experiments using pyruvate. In order to be sensitive to binding, one requires hyperpolarized protons. Therefore, we use a reverse INEPT sequence to transfer polarization from hyperpolarized 13C to 1H. This now enables us to detect pyruvate at sub-micromolar concentrations. We are now extending this work to detect binding of pyruvate to prolyl hydroxylase domain proteins which is connected to oxygen homeostasis.
Non-hydrogenative Parahydrogen Induced Polarization (nhPHIP) is an efficient tool to enhance the signals of compounds that can transiently associate to an iridium complex in solution. Aminoethanol was employed as a model system for nhPHIP hyperpolarization. In the presence of an iridium-Imes catalyst, an excess of 1-methyl-1,2,3-triazole and parahydrogen in solution, a transient complex is observed in which aminoethanol associates to the metal in the equatorial plane via the amino group. Transfer of hyperpolarization from the hydrides to the amino protons was achieved using a TACSY sequence (ca. 250 ms). The hyperpolarization of the amino protons was further transferred to the directly attached methylene protons via a DIPSI-2 mixing sequence. The first results indicate signal enhancements of 13% resp. 6.5% relative to the hyperpolarized hydrides for the amino- and methylene protons. Future developments: optimization of the experimental conditions (i.e., co-ligands, temperature, solvent) to increase hyperpolarization levels on the methylene protons. Optimization of the dissociation rate of the complex. Hyperpolarization of methylene protons in primary amines via non-hydrogenative parahydrogen-induced-polarization.
Hyperpolarization can also be achieved by Chemically Induced DNP (CINDNP). A collaboration with the team at ETH of G. R. Stadler, M. Bütikofer, F. Torres, R. Riek, A. Sonnefeld, K. Sheberstov, A. Razanoera, and G. Bodenhausen has shown promising results.
Significant achievements and breakthroughs
Aliphatic chains (like R-CH2-CH2-CH2-R’) are ubiquitous in drugs, but it was hitherto believed that their protons could not be used to support LLSs. The contrast between relaxation rates of LLSs in aliphatic chains in free and bound drugs turned out to be much more dramatic (and thus more favourable for drug screening) than expected, although these features are not yet completely understood and therefore deserve further investigation.
A selection of common molecules with methylene groups where LLS can be excited. The CH2 groups that were experimentally found to be accessible for excitation of LLSs by SLIC are emphasized in pink. In the case of (chiral) lysine (VII), the diastereotopic ß and 𝛄 CH2 groups feature distinct chemical shifts, so that LLSs can be excited by a variety of methods, while the protons of the ∂ and 𝛆 CH2 groups have nearly degenerate chemical shifts, where LLSs can be best excited by SLIC. With the exception of pentanol (XIII) and trimethoxy(propyl)silane (XIV), all compounds were measured in D2O.
While we planned to study field- and temperature-dependent relaxation in solids containing radicals, it came as a surprise to us that the proton relaxation in these materials can be modelled quantitatively and without any free parameters, using only an experimental calibration of the heat capacity of the Non-Zeeman electron reservoir.
Completely unexpected was the development by our external advisor Dr Claudio Dalvit of new theoretical methods that provide an analytical description of concentrations during the titration of ternary mixtures in intermediate exchange regimes.
Other important outputs that have arisen from this project
Software for simulations and for managing experiments has been exchanged between partners and can be made available to other users. Data bases for drugs and protein targets have been developed for internal consultations.
Dissemination efforts
In addition to publications listed above, the work of the HiSCORE network has been presented in lectures by B. Meier at HypMix (June 2022), by G. Bodenhausen at EUROMAR in Utrecht (July 2022), by M. Jurkutat at FGMR (September 2022), by A. Sonnefeld and by G. Bodenhausen in two separate lectures at the Experimental NMR Conference (ENC) in Asilomar (April 2023). Anna Sonnefeld won a prize for having submitted one of the best abstracts to ENC. Kirill Sheberstov and Geoffrey Bodenhausen have given lectures at ETH in Zürich, by Zoom to a Hyperpolarization meeting that was held in Leipzig, and on the international Zoom series “Konstantin Ivanov Intercontinental Seminar”.
These international meetings are supported by various instrument companies, and are attended by their top management. They are very keen to hear about our new methods, and we have developed strong ties with them over the years. We had fruitful exchanges with pharmaceutical industry (Sanofi in Paris, Novartis in Basel, …)
Meetings
We have held regular on-line meetings of the entire network that have allowed our objectives to converge.
Six physical meetings of all members of the entire network have been held so far: in Karlsruhe (21-23 March 2022), Nijmegen (26-29 October 2022), Paris (27-28 February 2023), Schloss Beuggen in South Germany (8-10 October 2023), Soeterbeek near Nijmegen (24-26 April 1924) and Karlsruhe in B Meier premises (22-23 October 2024). These meetings greatly contributed to consolidating the network, but we do not believe that this report should give a detailed record of our personal exchanges.
Three bilateral visits have been organized:
(1) ETH-IBG: applications of bullet DNP to drug/target interactions; enhancing proton polarization by transfer from carbon-13 polarization (reverse INEPT).
(2) ENS-IBG: use of very low spin temperature to create population imbalances and long-lived states.
(3) ENS-ETH: use of photo CIDNP to hyperpolarize various ligands that can support LLSs.
We have transferred a console from KIT to ENS, and a klystron from SKU to KIT
Much has yet to be done to integrate various aspects, but highly motivated staff, competent support, sophisticated instrumentation, and adequate lab space are all in place. Joint on-line and physical meetings and pairwise encounters have proven to be very effective.
| Activity | Speaker | Title | Date | Place | Audience | Attendance | Countries |
|---|---|---|---|---|---|---|---|
| lecture | Bodenhausen | LLSs | July 2022 | Utrecht NL | Major meeting | 200 | World-wide |
| lecture | Sonnefeld | Drug screen | April 2023 | Asilomar USA | Major meeting | 100 | World-wide |
| lecture | Bodenhausen | LLSs | April 2023 | Asilomar USA | Major meeting | 100 | World-wide |
| lecture | Meier | Bullet-DNP | June 2022 | Nantes FR | Topical Meeting | 50 | World-wide |
| lecture | Jurkutat | Low-field relaxa-tion | Sep. 2022 | Karlsruhe DE | Major meeting | 100 | World-wide |
| lecture | Sheberstov | LLS | March 2023 | ETH | Departe-mental meeting | 50 | Local |
| lecture | Sheberstov | LLC | 2023 | Ivanov Sem. | Zoom | 40 | World-wide |
| lecture | Bodenhausen | LLS/ LLC | 2024 | Ivanov Sem. | Zoom | 40 | World-wide |
| lecture | Bodenhausen | LLS/ LLC | June 2024 | Leipzig Spin Resonance Colloquium | Zoom | 200 | World-wide |
Publications
- Kouřil, K.; Gramberg, M.; Jurkutat, M.; Kouřilová, H.; Meier, B. A Cryogen-Free, Semi-Automated Apparatus for Bullet-Dynamic Nuclear Polarization with Improved Resolution. Magn. Reson. 2021, 2 (2), 815–825. https://doi.org/10.5194/mr-2-815-2021.
- He, M.; Faderl, D.; MacKinnon, N.; Cheng, Y.-T.; Buyens, D.; Jouda, M.; Luy, B.; Korvink, J. G. A Digital Twin for Parallel Liquid-State Nuclear Magnetic Resonance Spectroscopy. Commun Eng 2024, 3 (1), 90. https://doi.org/10.1038/s44172-024-00233-0.
- Faderl, D.; Chenakkara, A.; Jouda, M.; MacKinnon, N.; Gossert, A. D.; Korvink, J. G. Accelerated Screening of Protein–Ligand Interactions via Parallel T 2 -Weighted 19 F-MRI. Anal. Chem. 2024, 96 (24), 9859–9865. https://doi.org/10.1021/acs.analchem.4c00333.
- Becker, M.; Lehmkuhl, S.; Kesselheim, S.; Korvink, J. G.; Jouda, M. Acquisitions with Random Shim Values Enhance AI-Driven NMR Shimming. Journal of Magnetic Resonance 2022, 345, 107323. https://doi.org/10.1016/j.jmr.2022.107323.
- Turhan, E.; Pötzl, C.; Keil, W.; Negroni, M.; Kouřil, K.; Meier, B.; Romero, J. A.; Kazimierczuk, K.; Goldberga, I.; Azaïs, T.; Kurzbach, D. Biphasic NMR of Hyperpolarized Suspensions─Real-Time Monitoring of Solute-to-Solid Conversion to Watch Materials Grow. J. Phys. Chem. C 2023, 127 (39), 19591–19598. https://doi.org/10.1021/acs.jpcc.3c04198.
- Liang, J.; Davoodi, H.; Wadhwa, S.; Badilita, V.; Korvink, J. G. Broadband Stripline Lenz Lens Achieves 11 × NMR Signal Enhancement. Sci Rep 2024, 14 (1), 1645. https://doi.org/10.1038/s41598-023-50616-0.
- Sheberstov, K. F.; Sonnefeld, A.; Bodenhausen, G. Collective Long-Lived Zero-Quantum Coherences in Aliphatic Chains. The Journal of Chemical Physics 2024, 160 (14), 144308. https://doi.org/10.1063/5.0196808.
- Becker, M.; Jouda, M.; Kolchinskaya, A.; Korvink, J. G. Deep Regression with Ensembles Enables Fast, First-Order Shimming in Low-Field NMR. Journal of Magnetic Resonance 2022, 336, 107151. https://doi.org/10.1016/j.jmr.2022.107151.
- Wadhwa, S.; Buyens, D.; Korvink, J. G. Direct Chiral Discrimination with NMR. Advanced Materials 2024, 2408547. https://doi.org/10.1002/adma.202408547.
- Razanahoera, A.; Sonnefeld, A.; Sheberstov, K.; Narwal, P.; Minaei, M.; Kouřil, K.; Bodenhausen, G.; Meier, B. Hyperpolarization of Long-Lived States of Protons in Aliphatic Chains by Bullet Dynamic Nuclear Polarization, Revealed on the Fly by Spin-Lock-Induced Crossing. J. Phys. Chem. Lett. 2024, 15 (35), 9024–9029. https://doi.org/10.1021/acs.jpclett.4c01457.
- Bastawrous, M.; Ghosh Biswas, R.; Soong, R.; Jouda, M.; MacKinnon, N.; Mager, D.; Korvink, J. G.; Simpson, A. J. Lenz Lenses in a Cryoprobe: Boosting NMR Sensitivity Toward Environmental Monitoring of Mass-Limited Samples. Anal. Chem. 2023, 95 (2), 1327–1334. https://doi.org/10.1021/acs.analchem.2c04203.
- Sonnefeld, A.; Razanahoera, A.; Pelupessy, P.; Bodenhausen, G.; Sheberstov, K. Long-Lived States of Methylene Protons in Achiral Molecules. Sci. Adv. 2022, 8 (48), eade2113. https://doi.org/10.1126/sciadv.ade2113.
- Dreisewerd, L.; Aspers, R. L. E. G.; Feiters, M. C.; Rutjes, F. P. J. T.; Tessari, M. NMR Discrimination of D – and L -α-Amino Acids at Submicromolar Concentration via Parahydrogen-Induced Hyperpolarization. J. Am. Chem. Soc. 2023, 145 (3), 1518–1523. https://doi.org/10.1021/jacs.2c11285.
- Wu, B.; Aspers, R. L. E. G.; Kentgens, A. P. M.; Zhao, E. W. Operando Benchtop NMR Reveals Reaction Intermediates and Crossover in Redox Flow Batteries. Journal of Magnetic Resonance 2023, 351, 107448. https://doi.org/10.1016/j.jmr.2023.107448.
- Alinaghian Jouzdani, M.; Jouda, M.; Korvink, J. G. Optimal Control Flow Encoding for Time-Efficient Magnetic Resonance Velocimetry. Journal of Magnetic Resonance 2023, 352, 107461. https://doi.org/10.1016/j.jmr.2023.107461.
- Razanahoera, A.; Sonnefeld, A.; Bodenhausen, G.; Sheberstov, K. Paramagnetic Relaxivity of Delocalized Long-Lived States of Protons in Chains of CH 2 Groups. Magn. Reson. 2023, 4 (1), 47–56. https://doi.org/10.5194/mr-4-47-2023.
- Sonnefeld, A.; Bodenhausen, G.; Sheberstov, K. PolychromaticExcitation of Delocalized Long-Lived Proton Spin States in Aliphatic Chains. Phys. Rev. Lett. 2022, 129 (18), 183203. https://doi.org/10.1103/PhysRevLett.129.183203.
- Kouřilová, H.; Jurkutat, M.; Peat, D.; Kouřil, K.; Khan, A. S.; Horsewill, A. J.; MacDonald, J. F.; Owers-Bradley, J.; Meier, B. Radical-Induced Hetero-Nuclear Mixing and Low-Field 13 C Relaxation in Solid Pyruvic Acid. Phys. Chem. Chem. Phys. 2022, 24 (46), 28242–28249. https://doi.org/10.1039/D2CP04535D.
- Jurkutat, M.; Kouřilová, H.; Peat, D.; Kouřil, K.; Khan, A. S.; Horsewill, A. J.; MacDonald, J. F.; Owers-Bradley, J.; Meier, B. Radical-Induced Low-Field 1 H Relaxation in Solid Pyruvic Acid Doped with Trityl-OX063. J. Phys. Chem. Lett. 2022, 13 (44), 10370–10376. https://doi.org/10.1021/acs.jpclett.2c02357.
- Wong, Y. T. A.; Aspers, R. L. E. G.; Uusi-Penttilä, M.; Kentgens, A. P. M. Rapid Quantification of Pharmaceuticals via 1 H Solid-State NMR Spectroscopy. Anal. Chem. 2022, 94 (48), 16667–16674. https://doi.org/10.1021/acs.analchem.2c02905.
- Hoffmann, F.; Kouřil, K.; Berger, S. T.; Meier, B.; Luy, B. Rheo-NMR at the Phase Transition of Liquid Crystalline Poly-γ-Benzyl- L -Glutamate: Phase Kinetics and a Valuable Tool for the Measurement of Residual Dipolar Couplings. Macromolecules 2023, 56 (19), 7782–7794. https://doi.org/10.1021/acs.macromol.3c00488.
- Cheng, Y.-T.; Jouda, M.; Korvink, J. Sample-Centred Shimming Enables Independent Parallel NMR Detection. Sci Rep 2022, 12 (1), 14149. https://doi.org/10.1038/s41598-022-17694-y.
- Schmidt, D.; Gartner, P.; Berezkin, I.; Rudat, J.; Bilger, M.; Grünert, T.; Zimmerer, N.; Quarz, P.; Scharfer, P.; Brückel, J.; Jung, A. P.; Singh, P.; Pooja, P.; Meier, B.; Stahlberger, M.; Schabel, W.; Bräse, S.; Lanza, G.; Nesterov‐Mueller, A. Selective Peptide Binders to the Perfluorinated Sulfonic Acid Ionomer Nafion. Adv Funct Materials 2024, 34 (20), 2214932. https://doi.org/10.1002/adfm.202214932.
- Stern, Q.; Sheberstov, K. Simulation of NMR Spectra at Zero and Ultralow Fields from A to Z – a Tribute to Prof. Konstantin L’vovich Ivanov. Magn. Reson. 2023, 4 (1), 87–109. https://doi.org/10.5194/mr-4-87-2023.
- Yang, J.; Wang, P.; Korvink, J. G.; J. Brandner, J.; Lehmkuhl, S. The Steady‐State ALTADENA RASER Generates Continuous NMR Signals**. ChemPhysChem 2023, 24 (14), e202300204. https://doi.org/10.1002/cphc.202300204.
Disclaimer
| Funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or European Research Council. Neither the European Union nor the ERC can be held responsible for them. |