Biomolecular NMR Spectroscopy with Paramagnetic Spin Labels

Paramagnetic metal ions (e.g., Mn2+, Cu2+, Co2+, lanthanides) offer outstanding opportunities for the study of biomacromolecules by NMR spectroscopy. Paramagnetic NMR effects can be observed at large distances (up to 40Å) from the metal center and provide valuable restraints for the determination of the three-dimensional structure of proteins and their interactions with other biomolecules or small molecule ligands. The development of novel metal ion-binding tags and strategies for their covalent attachment to proteins have broadened the applicability of paramagnetic NMR, so that proteins without a natural metal binding site can be studied.

Previously, we used tagging with lanthanide ions to study the interaction of the cytokine interleukin-10 (IL-10) with glycosaminoglycans (GAGs), which are a class of highly sulfated carbohydrates found in the extracellular matrix of all animal cells. The pseudocontact shifts of the NMR signals of IL-10 and GAGs could be used to develop a structural model for the IL-10-GAG complex. The model showed GAG binding to a group of lysine and arginine residues located at the central crevice formed by the IL-10 dimer. Knowledge about the GAG-binding site of IL-10 is crucial to understand the function of GAGs in IL-10 biology and for the design of biomaterials composed of GAGs.

Three types of paramagnetic NMR data are commonly used because they provide long-range distance and/or orientational restraints useful for protein structure determination: PCS, RDC, and PRE. A computational framework was developed for the ROSETTA software that can use PCS, RDC, and PRE data and combine them with other local NMR data (chemical shifts, NOEs) for various applications such as de novo protein folding, protein-protein and protein-ligand docking, and modeling of symmetric complexes. A current project is to incorporate lanthanide-binding tags into the 4-helix membrane protein DsbB through conjugation to an unnatural amino which will allow the measurement of pseudocontact shifts as basis for ROSETTA structure calculations.

Paramagnetic NMR effects manifested in the NMR spectrum provide highly valuable restraints for structural modeling. (A) NMR spectrum of IL-10 in the presence of 1 eq. of dia- or paramagnetic lanthanides, respectively. PCS vectors of selected residues are indicated as black lines. (B) Left: Cartoon representation of IL-10. The PCSs induced by Tb3+ are depicted as isorsurfaces. Right: Model for the interaction of IL-10 with heparin obtained by PCS-guided docking. (C) Schema depicting the NMR data that are accessible through ROSETTA.

This approach holds great potential to study even larger membrane proteins because it only requires assignment of the NMR signals of the protein backbone.


  • Kuenze G, Bonneau R, Leman JK, Meiler J. Integrative Protein Modeling in RosettaNMR from Sparse Paramagnetic Restraints. Structure. 2019. 27(11):1721-1734.e5. doi: 10.1016/j.str.2019.08.012
  • Köhling S, Kuenze G, Lemmnitzer K, Bermudez M, Wolber G, Schiller J, Huster D, Rademann J. Chemoenzymatic Synthesis of Nonasulfated Tetrahyaluronan with a Paramagnetic Tag for Studying Its Complex with Interleukin-10. Chemistry. 2016. 22(16):5563-74. doi: 10.1002/chem.201504459
  • Kuenze G, Köhling S, Vogel A, Rademann J, Huster D. Identification of the Glycosaminoglycan Binding Site of Interleukin-10 by NMR Spectroscopy. J Biol Chem. 2016. 291(6):3100-13. doi: 10.1074/jbc.M115.681759
  • Kuenze G, Gehrcke JP, Pisabarro MT, Huster D. NMR characterization of the binding properties and conformation of glycosaminoglycans interacting with interleukin-10. Glycobiology. 2014. 24(11):1036-49. doi: 10.1093/glycob/cwu069

Design of Plastic-Degrading Enzymes

The PET-degrading enzyme Polyester Hydrolase Leipzig 7 (PHL7) bound to a PET polymer chain. PHL7 is promising candidate for an industrially used PETase enzyme and can efficiently degrade PET in post-consumer plastic. In our research, we use computer-aided protein design to engineer new variants of PHL7 with higher stability and activity.

The widespread use of plastics has led to a significant accumulation of non-biodegradable plastic waste in the environment. Current recycling methods for synthetic plastics, like Polyethylene terephthalate (PET), a thermoplastic polymer that is widely used in the food packaging industry, are not sustainable and have a high energy consumption. Enzymatic catalysis is a sustainable recycling solution because it does not require harsh chemicals, high temperatures or pressures. However, there is still a gap preventing enzymes to scale to widespread industrial usage:

To address these gaps our research aims to design plastic degrading enzymes that have optimized properties for an industrial usage. To achieve this, our lab uses a combination of Rosetta protein design, machine learning directed evolution, molecular dynamics, and deep learning methods. Designed enzymes are produced and tested in the wet lab. Furthermore, X-ray crystallography is used to determine the structure of novel enzymes and obtain important insight into the enzyme reaction mechanism.


  • Richter PK, Blázquez-Sánchez P, Zhao Z, Engelberger F, Wiebeler C, Künze G, Frank R, Krinke D, Frezzotti E, Lihanova Y, Falkenstein P, Matysik J, Zimmermann W, Sträter N, Sonnendecker C. Structure and function of the metagenomic plastic-degrading polyester hydrolase PHL7 bound to its product. Nat Commun. 2023. 14(1):1905. doi: 10.1038/s41467-023-37415-x
  • Engelberger F, Zakary JD, Künze G. Guiding protein design choices by per-residue energy breakdown analysis with an interactive web application. Front Mol Biosci. 2023. 10:1178035. doi: 10.3389/fmolb.2023.1178035
  • Falkenstein P, Zhao Z, Di Pede-Mattatelli A, Künze G, Sommer S, Sonnendecker C, Zimmermann W, Colizzi F, Matysik J, Song C. On the Binding Mode and Molecular Mechanism of Enzymatic Polyethylene Terephthalate Degradation. ACS Catalysis. 2023. 13:6919-6933. doi: 10.1021/acscatal.3c00259

Integrative Modeling of Biomacromolecular Complexes

Essential biological processes are carried out by proteins and their complexes. Understanding the role that these complexes play in both health and disease requires knowledge of their 3-dimensional structure. For challenging targets such as complexes with multiple subunits that interact only transiently with each other or are very flexible, the major technologies in structural biology – X-ray crystallography, NMR spectroscopy, and cryo-electron microscopy – struggle to obtain experimental data that unambiguously define atomic detail. Increasingly, structural models of protein complexes are obtained by integrative approaches which combine measured experimental data with computational modeling.

Research in the lab employs tools and algorithms of the ROSETTA software suite to develop models of biologically important proteins guided by restraints obtained with different experimental methods (e.g., NMR, EPR, XL-MS). In the past, these models helped to elucidate the basis for ligand specificity of Bradykinin G-protein-coupled receptors, identified interdomain contacts in the nuclear receptor LRH-1 important for allosteric communication, and demonstrated that transmembrane signaling in the histidine kinase NasS involves a metastable coiled-coil linker. These integrative models are disseminated to the scientific community through deposition in the PDB-Dev database. In addition, we also develop new computational methods in ROSETTA for integrative modeling applications. Currently, ways to efficiently model spin labels and chemical probes, which are used by reporter group techniques like NMR, FRET or chemical crosslinking, are explored. This will allow faster and more accurate calculation of these structural restraints.


  • Seacrist CD, Kuenze G, Hoffmann RM, Moeller BE, Burke JE, Meiler J, Blind RD. Integrated Structural Modeling of Full-Length LRH-1 Reveals Inter-domain Interactions Contribute to Receptor Structure and Function. Structure. 2020. 28(7):830-846. doi: 10.1016/j.str.2020.04.020
  • Kuenze G, Bonneau R, Leman JK, Meiler J. Integrative Protein Modeling in RosettaNMR from Sparse Paramagnetic Restraints. Structure. 2019. 27(11):1721-1734.e5. doi: 10.1016/j.str.2019.08.012
Components of a data-driven modeling workflow. Experimental data obtained with NMR spin-labels, FRET probes, or MS crosslinks are translated into spatial restraints to bias the conformational search. Coarse-grained modeling efficiently explores the vast search space before the model or parts of it are switched to all-atom mode for high-resolution refinement. Integrative models are written in IHM format and stored in PDB-Dev. Modeling components are developed in Rosetta as C++ libraries, and connected with each other and with other applications (e.g. from IMP) through their common Python interface to build complete protocols. 

  • Kuenze G, Meiler J. Protein structure prediction using sparse NOE and RDC restraints with Rosetta in CASP13. Proteins. 2019. 87(12):1341-1350. doi: 10.1002/prot.25769
  • Joedicke L, Mao J, Kuenze G, Reinhart C, Kalavacherla T, Jonker HRA, Richter C, Schwalbe H, Meiler J, Preu J, Michel H, Glaubitz C. The molecular basis of subtype selectivity of human kinin G-protein-coupled receptors. Nat Chem Biol. 2018. 14(3):284-290. doi: 10.1038/nchembio.2551

Voltage-Gated Ion Channels

The KCNQ1 ion channel. (A) Side view of KCNQ1 in complex with Calmodulin (PDB: 6V00). Each protomer is shown in a different color, and Calmodulin is represented as cylinders. (B) Top: Snapshots from a molecular dynamics simulation of KCNQ1 showing permeation of potassium ions through the selectivity filter. Bottom: Number of ion permeation events through the KCNQ1 pore in molecular dynamics simulations in which a sustained transmembrane voltage was applied.
Variants of KCNQ1 with a change in the voltage of half-maximal channel activation (V1/2). The location of residues that are substituted in variants of KCNQ1 are shown as spheres and colored according to their V1/2 change.

Membrane transport is essential to many areas of cellular life. The movement of ions across the cell membrane is carried out by ion channels which play crucial roles in many physiological processes such as electrical signaling in the brain, muscular contraction, generation of the heartbeat, and hormone secretion. The importance of this class of membrane proteins is also reflected by the fact that many diseases are caused by inherited mutations in ion channel genes, so-called channelopathies, and that 15% of currently used drugs target ion channels.

Voltage-gated potassium (KV) channels are membrane proteins that selectively conduct K+ ions across the cell membrane, and open or close in response to changes in transmembrane voltage. Of the 40 known human KV channel subunits, the KCNQ1 (KV7.1) channel is special because of its wide range of physiological behaviors in both excitable cells such as cardiomyocytes and non-excitable cells such as epithelia. KCNQ1 commonly co-assembles with one of five tissue-specific KCNE auxiliary proteins (KCNE1-5) which profoundly modify the biophysical properties of KCNQ1 and have afforded KCNQ1 with two important physiological functions: (1) repolarization of the cardiac tissue following an action potential, and (2) water and salt transport in epithelia. Mutations in KCNQ1 are associated with different forms of heart arrhythmia such as long QT syndrome, atrial fibrillation, and Romano-Ward syndrome, as well as impairment of intestinal chloride ion secretion related to cystic fibrosis.

Research in the lab employs molecular modeling, molecular dynamics simulation, and machine learning methods to study the structure-dynamics-function relationships in the KCNQ1 channel. More specifically, we aim to answer questions of how KCNQ1 transitions between closed and open states, how KCNE proteins bind to KCNQ1 and modulate its function, and how mutations alter the structure, stability, and dynamics of KCNQ1. So far, the results of these investigations have identified molecular details of the interaction between KCNQ1 and KCNE1, which suggest an allosteric mode of KCNQ1 modulation. We have also quantified changes in protein thermodynamic stability for a large set of KCNQ1 mutations. Most of the loss-of-function mutations in KCNQ1 cause destabilizations of the protein structure consistent with these protein variants having folding and trafficking defects. To more completely understand how changes in KCNQ1 structure and dynamics affect its function on the cellular level, research on KCNQ1 is performed collaboratively with experimental biochemists and electrophysiologists.


  • Taylor KC, Kang PW, Hou P, Yang ND, Kuenze G, Smith JA, Shi J, Huang H, White KM, Peng D, George AL, Meiler J, McFeeters RL, Cui J, Sanders CR. Structure and physiological function of the human KCNQ1 channel voltage sensor intermediate state. Elife. 2020. doi: 10.7554/eLife.53901
  • Kuenze G, Duran AM, Woods H, Brewer KR, McDonald EF, Vanoye CG, George AL Jr, Sanders CR, Meiler J. Upgraded molecular models of the human KCNQ1 potassium channel. PLoS One. 2019. 14(9):e0220415.  doi: 10.1371/journal.pone.0220415
  • Huang H, Kuenze G, Smith JA, Taylor KC, Duran AM, Hadziselimovic A, Meiler J, Vanoye CG, George AL Jr, Sanders CR. Mechanisms of KCNQ1 channel dysfunction in long QT syndrome involving voltage sensor domain mutations. Sci Adv. 2018. 4(3):eaar2631. doi: 10.1126/sciadv.aar2631