Skip to main content
NCBI home page
ACS AuthorChoice logoLink to ACS AuthorChoice
. 2026 Feb 24;98(9):6541–6546. doi: 10.1021/acs.analchem.5c04414

Quantification of Ligand–Membrane Interactions Using DNP-NMR Relaxometry

Chang Qi 1, Nirmalya Pradhan 1, Christian Hilty 1,*
PMCID: PMC12980485  PMID: 41732961

Abstract

The transverse relaxation rates (R 2) for 19F spins from a small molecule ligand are measured in the presence of phospholipid vesicles with varied compositions and concentrations. The sensitivity of detection is enhanced by hyperpolarization using dissolution dynamic nuclear polarization (D-DNP), enabling measurement of R 2 relaxation rates in single scans. The binding interaction is described as an equilibrium with a defined number of binding sites on the membrane. From the increase of R 2 as a function of lipid concentration, a parameter (f R 2,b)/K D is calculated, which depends on the fractional number of binding sites per lipid (f), the relaxation rate (R 2,b) of bound ligand, and the dissociation constant (K D) for each binding site. The relaxation rate of a bound ligand is modeled based on molecular motions, including rigid body tumbling, diffusion in the bilayer, wobble motions, and CF3 rotation. By estimation of R 2,b from chemical shift anisotropy and dipole–dipole relaxation, the dissociation constant K D/f is evaluated to greater than 10 mM for the ligand binding to 200 nm vesicles. The resulting values for the binding affinity were not significantly affected by the presence of cholesterol in the bilayer, but were reduced by vesicle aggregation. These data further demonstrate that R 2 relaxation measurements using DNP hyperpolarization provide a means to detect ligand–membrane binding and kinetic parameters. In addition to R 2, a set of other NMR derived parameters that are sensitive to molecular dynamics, such as R 1ρ, diffusion and cross-relaxation, and spectroscopic techniques, such as Laplace NMR, are compatible with the hyperpolarized method. The measurement of membrane–ligand binding can facilitate applications for drug discovery and biomedical studies.

graphic file with name ac5c04414_0005.jpg

graphic file with name ac5c04414_0004.jpg

Introduction

Cell membranes are increasingly recognized for their molecular diversity, as an environment where functional and pharmacological interactions occur not only with membrane proteins but also with lipid molecules and other components. The insertion of a small molecule drug modulates the function of the lipid membrane by affecting its physical properties such as thickness, fluidity, and charge, thus regulating the passive transport of materials across the lipid membrane. The membrane bilayers also play a role in the binding of small molecule inhibitors to membrane receptors, whereby molecules can initially bind to the membrane and subsequently diffuse toward the binding sites on the proteins. The interactions between ligand molecules and hormones, neurotransmitters, or ion channels that are distributed in the cell membrane but occupy only a minute fraction of its surface area are enhanced by the reduction in dimensionality to the two-dimensional membrane. Membrane binding also increases the formation of complexes bound to ligands with low water solubility. Recent studies reveal that ligand–membrane interactions are relevant to intercellular receptor–ligand binding as well. On this basis, understanding ligand–membrane binding provides essential guidance for designing new drugs.

In characterizing ligand–membrane interactions, NMR provides the possibility to preserve the conformation and stereostructure of the ligand–membrane bound complex under physiological conditions. Solid-state NMR spectroscopy of 1H, 31P, and other nuclei has been applied to characterize the conformation of the bound ligand and the kinetics of the binding process. , The interactions between a small molecule and the lipid membrane have also been investigated through the observation of NMR parameters including chemical shift difference, nuclear Overhauser effect, , paramagnetic relaxation rates, as well as residual dipolar coupling in solution.

Dynamic nuclear polarization (DNP), a technique that enhances the NMR signal by several thousandfold, can be used to decisively improve the detection limit in ligand binding studies. Hyperpolarization has been applied to facilitate the determination of kinetic parameters including binding affinity , and characterization of ligand binding epitopes. − Hyperpolarization-assisted ligand-observed NMR allows the screening of a ligand binding to immobilized proteins. The application of hyperpolarized 19F NMR enables background-free detection of a small molecule at a concentration of ∼μM in a single scan.

Here, we observe the relaxation rates of the hyperpolarized 19F spin on a small molecule ligand, which interacts with model membranes in small vesicles of different lipid compositions, to investigate the kinetic process of ligand–membrane binding. We then predict the contributions of the bound forms based on relaxation theory and use the results to estimate affinity ranges for the ligand binding to the membranes.

Experimental Methods

Vesicle Preparation

100% 1-Palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) (Avanti Polar Lipids, Alabaster, AL) vesicles as well as 70% POPC and 30% cholesterol (Avanti Polar Lipids, Alabaster, AL) vesicles were prepared for the experiments. An amount of 20 μmol of POPC lipids or 14 μmol of POPC lipids with 6 μmol of cholesterol were dissolved in 2 mL of CHCl3 and dried under N2 gas as a thin film. The lipids were further dried under vacuum overnight. After rehydrating with 5 mL of phosphate buffered saline (PBS; 137 mM sodium chloride, 2.7 mM potassium chloride, 10 mM sodium phosphate dibasic, and 1.8 mM potassium phosphate monobasic, pH 7.4) buffer, the mixture was thawed in a 55 °C water bath and frozen by liquid N2 10 times. Extrusion was carried out using an extruder kit (Avanti Polar Lipids, Alabaster, AL) with 0.2 or 1 μm membranes (Whatman, Maidstone, UK) 11 times. A nonextruded sample was used directly after the freeze–thaw steps.

DNP Experiments

A sample of 40 mM 2-methyl-3-(5-methylsulfanyl-[1,3,4]­oxadiazol-2-yl)-6-trifluoromethyl-pyridine (ICT5040; Aobious, MA) containing 15 mM 4-hydroxy-2,2,6,6-tetramethylpiperidine-1-oxyl (TEMPOL) radicals (Sigma-Aldrich, St. Louis, MO) was prepared in a mixture of DMSO-d 6 and D2O at 4:1 v/v (Cambridge Isotope Laboratories, Tewksbury, MA). The sample (10 μL for gas-driven experiment and 2 μL for liquid-driven experiments with flow cell) was loaded into a HyperSense DNP polarizer containing a 3.35 T magnet (Oxford Instruments, Abingdon, UK) and irradiated with 100 mW microwaves at a frequency of 94.005 GHz and temperature of 1.4 K for 40 min. The hyperpolarized sample was dissolved with heated PBS buffer at 10 bar and rapidly transferred to the NMR spectrometer. Gas-driven injection was used to determine the DNP signal enhancement, applying forward and back pressures of 262 and 150 psi, respectively. A stabilization period of 500 ms was applied prior the NMR experiment. , A 3D printed flow cell of 240 μL volume via liquid-driven injection was used in all of the other experiments. The flow cell was preloaded in an NMR instrument with a broad-band (BBO) probe (Bruker Biospin, Billerica, MA). In the liquid-driven injection, the dissolved DNP sample and vesicle samples were automatically loaded into sample loops. The volume of the nonhyperpolarized sample was determined by the 0.4 mL sample loop volume. Samples were subsequently injected into the flow cell by two high-pressure syringe pumps (Models 500D and 1000D, Teledyne ISCO, Lincoln, NE). The flow rates for the pumps were set to 80 and 200 mL/min, respectively. The corresponding injection times were 900 and 750 ms, applied with a simultaneous end point. The injection was followed by a 500 ms delay to stabilize the stopped fluid before NMR acquisition.

NMR Spectroscopy

19F signals were collected on the 400 MHz NMR spectrometer at 298 K. A single-scan Carr–Purcell–Meiboom–Gill (CPMG) pulse sequence p1 – [τ – p2 – τ] × n (Figure S1) with an echo time 2τ of 1721.9 μs and loop number n of 2875 was used to measure the R 2 relaxation rates. The p1 and p2 are hard pulses of π/2 and π flip angle applied with a pulse strength of γB 1 = 47.6 kHz. The raw data of each experiment was reshaped into a two-dimensional data set according to its loop number and echo time using Python (Python Software Foundation, https://www.python.org). A sine-shaped window function was applied to every echo, and zero filling was performed to both ends of each echo with the number of zeroes added equal to the size of each echo. Every set of data points from one echo was Fourier transformed to obtain a spectrum containing the single peak for the 19F spin. The integrals of the peaks from all echoes were fitted to a single exponential decay versus the acquisition time to determine the R 2 relaxation rate. The signal enhancement achieved by DNP was calculated by comparing the integrals of the 19F peaks obtained in the experiments with and without hyperpolarization.

Results and Discussion

2-Methyl-3-(5-methylsulfanyl-[1,3,4]­oxadiazol-2-yl)-6-trifluoromethyl-pyridine (ICT5040) is a small molecule inhibitor of the G-protein coupled CXCR4 cell membrane receptor (Figure a). The interaction of this prototype drug molecule with 1-palmitoyl-2-oleoyl-sn-glycero-3-phosphocholine (POPC) vesicles was characterized. Figure b shows the hyperpolarized 19F NMR signal of ICT5040. The 19F signal is distinct, even in samples that comprise mixtures of biological molecules, where this nucleus is not abundant. The DNP signal enhancement for 19F in this molecule was estimated by injecting an aliquot of the hyperpolarized sample into an NMR tube. A signal enhancement of 768 ± 21 fold was determined by comparison with the spectrum of the sample after decay of the spin hyperpolarization.

1.

1

Open in a new tab

(a) Structure of ICT5040. (b) 19F spectrum of 0.32 mM DNP hyperpolarized ICT5040 (HP; red) compared to the same sample measured in 64 scans after decay of hyperpolarization (Non-HP; blue). The vertical scale of the nonhyperpolarized spectrum is reduced by a factor of 641/2 = 8 to match noise levels.

Different relaxation rates were observed for the molecules in the presence of lipid vesicles of varying sizes and compositions. In Figure a, the presence of POPC vesicles extruded through a 200 nm membrane resulted in a faster R 2 decay than for the free molecule, indicating that the molecule interacts with the lipid. The increase in R 2 for the 19F spins was quantified by varying the lipid concentration for each of the vesicle types (Figure b). A linear dependence of the observed relaxation rate R 2,obs on the lipid concentration was seen in all cases, with a vertical axis intercept in agreement with R 2,f . This behavior is explained by a simple model for the binding of the ligand molecule to available binding sites on the vesicles.

2.

2

Open in a new tab

(a) Data points and fitted curves for R 2 relaxation measurements of the 19F spin obtained in the absence (open diamond, R 2,f = 0.93 s–1) and presence (open triangle, R 2,obs = 3.37 s–1) of 0.015 mM 200 nm POPC vesicles where [L]t = 33 μM. The data was acquired with a CPMG pulse sequence with an echo time of 1.7 ms in a single scan. In the figure, only every 32nd data point is plotted for the sake of clarity. (b) R 2 relaxation rates of hyperpolarized 19F on ICT5040 measured with the vesicles indicated. The fitting equations of each line are R 2,obs = 179.84­[Lipids] + 0.84 (dotted line), R 2,obs = 173.75­[Lipids] + 0.90 (long dashed line), R 2,obs = 149.77­[Lipids] + 0.86 (solid line), and R 2,obs = 71.786­[Lipids] + 0.93 (short dashed line) with R 2 = 0.998, 0.964, 0.999, and 0.997, respectively. The lipid concentration was calculated according to the dilution factor in SI.

The model describing the interaction assumes that the ligand molecule can bind to a number of binding sites on the vesicle (eqs S1–S4). This number is proportional to the lipid concentration B t = f V t. Here, B t is the concentration of the assumed ligand binding sites and V t the concentration of vesicle forming lipids ([Lipids] in Figure b). The factor f indicates the number of binding sites per lipid molecule (whereby f can be larger or smaller than one). The fraction of the bound ligand is calculated from the chemical equilibrium between free and bound ligand molecules

Xb=KD+fVt+Lt−(KD+fVt+Lt)2−4fVtLt2Lt 1

In this equation, K D is the dissociation constant for a ligand molecule with a binding site and L t is the total ligand concentration. This equation corresponds to a well-known equation for binding found by solving the equations describing the equilibrium concentrations of free and bound species. Here, the factor f is added, which arises from the multisite model described above.

In the limit of fast exchange, the spin–spin relaxation rate is the average of the free and bound rates

R2,obs=XbR2,b+(1−Xb)R2,f 2

It is noted that fast exchange is an assumption that is not trivially fulfilled; however, it appears reasonable for the presumed nonspecific binding between the bilayer and the small molecule ligand.

The dependence of X b, and consequently of R 2,obs, on V t according to eqs and exhibits saturation behavior (Figure S2). Based on the observed linear dependence of R 2,obs on V t, it can be concluded that the lipid concentration is sufficiently low for an initial slope approximation to be valid. The slope of R 2,obs at V t = 0 under these conditions is

dR2,obsdVt=f(R2,b−R2,f)KD+Lt 3

with R 2,b and R 2,f denoting the spin–spin relaxation rates of bound and free ligands, respectively. When L t ≪ K D and R 2,f ≪ R 2,b, the slope becomes (f R 2,b)/K D.

Equation indicates that R 2,b and f/K D cannot be independently determined from the slope in Figure b. In the following, we compare this parameter under the assumption that R 2,b is independent of vesicle type and subsequently discuss an estimate for R 2,b based on prevalent relaxation mechanisms.

First, the interaction of ICT5040 in vesicles with or without cholesterol was investigated. Adding cholesterol to vesicle membranes is known to increase their viscosity and affect the transition temperature (T m). The T m for POPC is −2 °C. The presence of 30% cholesterol may slightly affect the T m and will make the phase transition less abrupt, but the measurements for vesicles with or without 30% cholesterol all take place in the liquid phase of the bilayer at room temperature. The addition of 30% (v/v) cholesterol also did not significantly affect the size or volume of vesicles as determined by dynamic light scattering (DLS) analysis (Figures S3 and S4, Tables S3 and S4).

Comparing the observed R 2 of ICT5040 at the same lipid concentration, we found that the relaxation rates were not significantly different for 200 nm POPC vesicles with or without 30% cholesterol. For example, the R 2,obs values were determined as 1.77 ± 0.03 s–1 and 1.83 ± 0.05 s–1 using 5 μM 100% POPC and 70% POPC with 30% cholesterol, respectively. Here, the errors were calculated from the individual fits of the relaxation rates.

The (f R 2,b)/K D values fitted from the experiments with POPC vesicles that contain or do not contain 30% cholesterol were 174 s–1·mM–1 and 180 s–1·mM–1, respectively (Figure ). Generally, the addition of cholesterol hinders diffusion of the lipids in bilayers. However, it does not appear to significantly affect the motions of the ligand molecule that cause the relaxation.

3.

3

Open in a new tab

Quantity (f R 2,b)/K D determined from the initial slope of the titrations of hyperpolarized ICT5040 with POPC vesicles. The vesicles were extruded as described in the legend. The error bars were calculated from a 5.6% error in lipid concentration (Figure S3) and fitting errors of the R 2,obs curves.

Second, we characterized the dependence of observed relaxation rates on the size of vesicles. Figure shows that POPC vesicles with 30% cholesterol extruded with a 200 nm membrane, a 1 μm membrane, and without extrusion exhibit progressively smaller (f R 2,b)/K D. Regardless of the extrusion process, the largest number of vesicles has a size of 100 nm (Figure S3 and Table S1). However, extrusion with membranes of larger pore size results in some larger vesicles that occupy more volume, as seen in Figure S4c. The observed trend in (f R 2,b)/K D is contrary to the increase in R 2,obs that would be expected if the molecule were rigidly bound to structures of increasing size. Based on the observed change in (f R 2,b)/K D, the vesicles extruded through the larger pores therefore exhibit weaker apparent binding or a smaller number of accessible binding sites evidenced by a smaller factor f. This behavior may be explained by larger vesicles containing a smaller accessible surface area per unit weight, which could, in part, be due to multilamellar features or aggregation behavior.

The initial slope of R 2,obs measured with increasing lipid concentration, (f R 2,b)/K D, is the ratio of R 2,b and the binding parameter K D/f. Because the titration does not reach a high lipid concentration, where R 2,obs reaches a plateau as shown in Figure S2, in the linear regime at low concentration, the binding parameters and relaxation rates of the bound ligand cannot be determined independently. In the following, we discuss estimates for R 2,b based on relaxation theory. Once the R 2,b range is known, the range of binding affinities of the molecule for the membrane can be predicted accordingly.

A vesicle as a rigid sphere exhibits a correlation time τc of several hundred microseconds according to the Stokes–Einstein equation (eq S13). The lateral diffusion over the curved vesicle surface introduces a dependence on a transverse diffusion coefficient. , The transverse diffusion for the small molecule is expected to be similar or faster than that of the lipid molecule, which was described as 1.9 × 10–11 m2 s–1 for POPC. For a vesicle of 200 nm diameter, this diffusion coefficient results in a reduced effective correlation time of τv = 43 μs (eq S15). , The dipolar interaction between the 19F and the closest averaged 1H at a distance of 3.01 Å and intrafluoromethyl relaxation with a distance of 2.25 Å between fluorine spins results in R 2,b(DD) = 1.4 × 105 s–1. The calculated chemical shift anisotropy for the molecule is Δσ = −56 ppm, which would contribute R 2,b(CSA) = 6.7 × 104 s–1.

Additional motions of the ligands reduce the relaxation rates. In the following, the motion of the entire ligand in the membrane and the fast rotation of the CF3 group are considered. The corresponding correlation times and order parameters are included in a model-free treatment of the relaxation. , The motion of the entire molecule may be described as a wobble motion with a typical diffusion coefficient D w = 1/(6τw) ≈ 107–109 s–1. The order parameter for a DPPC lipid molecule in a 100 nm vesicle was reported as ∼0.6 in the liquid phase. The bilayer of the POPC vesicles is also in the liquid phase at the experimental temperature of 298 K. For fluorescent probes in DPPC or POPC lipid membranes, the order parameter varied from 0.8 to 0.4 among the gel and liquid phases. The order parameter for the small molecule should be lower than that for a lipid. On the other hand, a model-free treatment of CF3 linked to a protein side-chain resulted in an order parameter and correlation time as low as 0.1 and 10–9 s, respectively. On the basis of these values, we consider S w of 0.1–0.4 and τw of 10–9–10–8 s for ICT5040 in the membrane.

For the CF3 rotation, the order parameter was estimated by using a computer simulation. The unlabeled carbon atom does not carry a nuclear spin, and the dominant interaction is with an adjacent hydrogen atom. The autocorrelation function was calculated using a time and ensemble average of random traces generated from jumps between three sites, resulting in S r = 0.54 for an F–H pair and S r = 0.24 for F–F (Figure S5). The latter matches a theoretically expected value of 1/4. In analogy to CH3, a correlation time τr = 10–11 s is assumed for the CF3 group. With these parameters, dipole–dipole relaxation results in an R 2,b(DD) of 3.6 × 103 s–1 to 1.5 × 104 s–1 (eqs S17–S19). The R 2,b(CSA) is 6.7 × 103 s–1 to 2.7 × 104 s–1, whereby the symmetrized chemical shift tensor implicitly includes fast CF3 rotation.

With eq and L t ≪ K D, the resulting K D/f value is between 57 mM and 230 mM. These values are assuming that the exchange between free and bound ligands is fast. An exchange contribution to R 2,obs due to slower exchange would lead to an underestimation of the K D value. For other vesicles that were extruded through a 1 μm membrane or nonextruded, 92% or 67% of the vesicles are of the size 100 nm, while larger vesicles are present (Table S2). Nevertheless, the experimental (f R 2,b)/K D value became smaller. This may be due to the aggregation of large vesicles, as discussed above, which would reduce the number of accessible binding sites, f.

The change in the R 2 relaxation of ICT5040 in different vesicles can be compared with line widths of lipid molecules. Figure S6 shows the 1H NMR spectra for different types of vesicles used in the experiments. The 1H line width was obtained for peaks from three regions, −N­(CH3)3 from the headgroup, −CH2 from the long chain, and −CH3. Some narrow signals that are not likely part of the bilayer are not considered in the analysis. The 1H line width measured for the−CH2 peak in the 100% POPC sample is about 78 Hz (Table S5). The presence of 30% cholesterol further broadens the 1H line width. This behavior is consistent with a reduction in the mobility of lipid molecules in the vesicle bilayer due to interactions with cholesterol. By contrast, the observed relaxation rate for the ligand that binds to vesicles with or without cholesterol does not show a significant change, suggesting that the mobility of the ligand in the vesicle membrane is only slightly affected by interactions with cholesterol.

The above differences illustrate the necessity of directly measuring ligand relaxation rates. The R 2 parameter is interesting for the measurement of weak interactions at low concentrations because it is strongly influenced by binding, and the observed signal is averaged between free and bound forms. Even a small fraction of the total ligand in the bound form can produce a detectable change in the signal. The observation of the total hyperpolarized ligand signal distinguishes this method from other biophysical techniques that are based on measurement of the less abundant bound ligand complex, which may render the detection of weak interactions with dissociation constants in the mM range challenging.

The DNP-assisted relaxometry method allows the detection of ligand–membrane interactions for low concentrations of the ligand. A signal enhancement of 1000-fold results in a submicromolar detection limit, which mimics a physiological concentration. The single scan experiment is measured with a signal-to-noise ratio that would require approximately 2 weeks of averaging in conventional NMR. DNP simplifies the observation of a small amount of ligand by making hyperpolarized signals equivalent to or larger than the background signals. Although the 19F signal measured here is not affected by background signals, the polarization enhancement can be important for resolving signals from abundant spins such as 1H or 31P. In conventional NMR, signal overlap can be severe in the presence of large amounts of membrane lipid.

NMR relaxation is among the most sensitive observables that depend on molecular dynamics. The R 2 parameter is strongly influenced by microsecond to millisecond exchange events. The existence of an extensive body of relaxation theory developed since the advent of NMR spectroscopy facilitates the interpretation of relaxation data. Spin hyperpolarization is becoming more widely available for biological molecules. New applications of relaxation theory extend the utility of hyperpolarization for dynamic measurements in biomolecular NMR spectroscopy.

The same measurement of R 2 relaxation may be further applied to study the binding of ligands to membrane proteins in vesicles or possibly in cells. Since the structure of membrane proteins is more rigid compared to phospholipids in the membrane bilayer, the R 2,b for the ligands that bind to membrane proteins should be much larger compared to that for a ligand–membrane bound complex, but the number of binding sites would be lower.

Other NMR observables beyond R 2 are compatible with the described method. In the presence of slow exchange dynamics, R 1ρ, R 2 dispersion, or self-diffusion, the latter of which can be measured isotropically or anisotropically, can show an association with lipid structures or molecular complexes. Cross-relaxation is also highly sensitive to binding to rigid complexes, which is exploited in transfer nuclear Overhauser effect (tr-NOE) experiments. , The method is compatible with Laplace-NMR (LNMR), which analyzes the dynamic parameters by the time domain decay of signals and resolves multiple decay rates. Applications of ultrafast LNMR, where two-dimensional spectroscopy is capable of correlating these parameters, are emerging. , A combination of these measurements may provide dynamic information on interactions in heterogeneous systems comprising membranes and membrane constituents beyond the binding affinity.

Conclusions

The ligand–membrane interaction was studied by observing the transverse relaxation rates of a small molecule ligand that binds to vesicle bilayers at varied lipid concentrations. The observed relaxation rates can be interpreted in the context of relaxation theory to obtain information on affinity and the number of binding sites in the ligand–vesicle binding equilibrium. The determination of model parameters assist in the determination of ligand–membrane protein interactions for drug discovery by quantifying the off-target contribution to R 2 relaxation that is due to lipid binding. They may also help in understanding the mechanism of drug delivery and action, where the association of a drug molecule in the membrane bilayer lowers the barrier of its binding to membrane proteins and provides an environment that could mediate the protein–ligand interactions.

Supplementary Material

ac5c04414_si_001.pdf (1.9MB, pdf)

Acknowledgments

Financial support from the National Institutes of Health (Grant R01GM132655) and the Welch Foundation (Grant A-1658) is gratefully acknowledged.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acs.analchem.5c04414.

  • CPMG pulse sequence, quantification of sample concentration in flow cell, binding equilibrium, dynamic light scattering, relaxation equations, 1H NMR signals of lipids (PDF)

The authors declare no competing financial interest.

References

  1. Martins P. T., Velazquez-Campoy A., Vaz W. L. C., Cardoso R. M. S., Valério J., Moreno M. J.. Kinetics and Thermodynamics of Chlorpromazine Interaction with Lipid Bilayers: Effect of Charge and Cholesterol. J. Am. Chem. Soc. 2012;134(9):4184–4195. doi: 10.1021/ja209917q. [DOI] [PubMed] [Google Scholar]
  2. Vauquelin G., Packeu A.. Ligands, Their Receptors and··· Plasma Membranes. Mol. Cell. Endocrinol. 2009;311(1–2):1–10. doi: 10.1016/j.mce.2009.07.022. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Payandeh J., Volgraf M.. Ligand Binding at the Protein–Lipid Interface: Strategic Considerations for Drug Design. Nat. Rev. Drug Discovery. 2021;20(9):710–722. doi: 10.1038/s41573-021-00240-2. [DOI] [PubMed] [Google Scholar]
  4. Medeiros-Silva J., Jekhmane S., Paioni A. L., Gawarecka K., Baldus M., Swiezewska E., Breukink E., Weingarth M.. High-Resolution NMR Studies of Antibiotics in Cellular Membranes. Nat. Commun. 2018;9(1):3963. doi: 10.1038/s41467-018-06314-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Matsumori N., Murata M.. 3D Structures of Membrane-Associated Small Molecules as Determined in Isotropic Bicelles. Nat. Prod. Rep. 2010;27(10):1480–1492. doi: 10.1039/c0np00002g. [DOI] [PubMed] [Google Scholar]
  6. Hsu S.-T., Breukink E., de Kruijff B., Kaptein R., Bonvin A. M. J. J., van Nuland N. A. J.. Mapping the Targeted Membrane Pore Formation Mechanism by Solution NMR: The Nisin Z and Lipid II Interaction in SDS Micelles. Biochemistry. 2002;41(24):7670–7676. doi: 10.1021/bi025679t. [DOI] [PubMed] [Google Scholar]
  7. Chandrasekhar I., van Gunsteren W. F., Zandomeneghi G., Williamson P. T. F., Meier B. H.. Orientation and Conformational Preference of Leucine-Enkephalin at the Surface of a Hydrated Dimyristoylphosphatidylcholine Bilayer: NMR and MD Simulation. J. Am. Chem. Soc. 2006;128(1):159–170. doi: 10.1021/ja054785q. [DOI] [PubMed] [Google Scholar]
  8. Sadet A., Stavarache C., Bacalum M., Radu M., Bodenhausen G., Kurzbach D., Vasos P. R.. Hyperpolarized Water Enhances Two-Dimensional Proton NMR Correlations: A New Approach for Molecular Interactions. J. Am. Chem. Soc. 2019;141(32):12448–12452. doi: 10.1021/jacs.9b03651. [DOI] [PubMed] [Google Scholar]
  9. Evanics F., Prosser R. S.. Discriminating Binding and Positioning of Amphiphiles to Lipid Bilayers by 1H NMR. Anal. Chim. Acta. 2005;534(1):21–29. doi: 10.1016/j.aca.2004.06.061. [DOI] [Google Scholar]
  10. Ardenkjær-Larsen J. H., Fridlund B., Gram A., Hansson G., Hansson L., Lerche M. H., Servin R., Thaning M., Golman K.. Increase in Signal-to-Noise Ratio of > 10,000 Times in Liquid-State NMR. Proc. Natl. Acad. Sci. U. S. A. 2003;100(18):10158–10163. doi: 10.1073/pnas.1733835100. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Kim Y., Hilty C.. Affinity Screening Using Competitive Binding with Fluorine-19 Hyperpolarized Ligands. Angew. Chem., Int. Ed. 2015;54(16):4941–4944. doi: 10.1002/anie.201411424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Liu M., Kim Y., Hilty C.. Characterization of Chemical Exchange Using Relaxation Dispersion of Hyperpolarized Nuclear Spins. Anal. Chem. 2017;89(17):9154–9158. doi: 10.1021/acs.analchem.7b01896. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Lee Y., Zeng H., Mazur A., Wegstroth M., Carlomagno T., Reese M., Lee D., Becker S., Griesinger C., Hilty C.. Hyperpolarized Binding Pocket Nuclear Overhauser Effect for Determination of Competitive Ligand Binding. Angew. Chem., Int. Ed. 2012;51(21):5179–5182. doi: 10.1002/anie.201201003. [DOI] [PubMed] [Google Scholar]
  14. Wang Y., Kim J., Hilty C.. Determination of Protein–Ligand Binding Modes Using Fast Multi-Dimensional NMR with Hyperpolarization. Chem. Sci. 2020;11(23):5935–5943. doi: 10.1039/D0SC00266F. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Qi C., Wang Y., Hilty C.. Application of Relaxation Dispersion of Hyperpolarized 13C Spins to Protein–Ligand Binding. Angew. Chem., Int. Ed. 2021;60(45):24018–24021. doi: 10.1002/anie.202109430. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Wang Y., Hilty C.. Amplification of Nuclear Overhauser Effect Signals by Hyperpolarization for Screening of Ligand Binding to Immobilized Target Proteins. Anal. Chem. 2020;92(20):13718–13723. doi: 10.1021/acs.analchem.0c01071. [DOI] [PubMed] [Google Scholar]
  17. Chen H.-Y., Hilty C.. Implementation and Characterization of Flow Injection in Dissolution Dynamic Nuclear Polarization NMR Spectroscopy. ChemPhysChem. 2015;16(12):2646–2652. doi: 10.1002/cphc.201500292. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Bowen S., Hilty C.. Rapid Sample Injection for Hyperpolarized NMR Spectroscopy. Phys. Chem. Chem. Phys. 2010;12(22):5766–5770. doi: 10.1039/c002316g. [DOI] [PubMed] [Google Scholar]
  19. Vinader V., Ahmet D. S., Ahmed M. S., Patterson L. H., Afarinkia K.. Discovery and Computer Aided Potency Optimization of a Novel Class of Small Molecule CXCR4 Antagonists. PLoS One. 2013;8(10):e78744. doi: 10.1371/journal.pone.0078744. [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. Kim Y., Hilty C.. Affinity Screening Using Competitive Binding with Fluorine-19 Hyperpolarized Ligands. Angew. Chem., Int. Ed. Engl. 2015;54(16):4941–4944. doi: 10.1002/anie.201411424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Gossert A. D., Jahnke W.. NMR in Drug Discovery: A Practical Guide to Identification and Validation of Ligands Interacting with Biological Macromolecules. Prog. Nucl. Magn. Reson. Spectrosc. 2016;97:82–125. doi: 10.1016/j.pnmrs.2016.09.001. [DOI] [PubMed] [Google Scholar]
  22. Leekumjorn S., Sum A. K.. Molecular Characterization of Gel and Liquid-Crystalline Structures of Fully Hydrated POPC and POPE Bilayers. J. Phys. Chem. B. 2007;111(21):6026–6033. doi: 10.1021/jp0686339. [DOI] [PubMed] [Google Scholar]
  23. Redondo-Morata L., Giannotti M. I., Sanz F.. Influence of Cholesterol on the Phase Transition of Lipid Bilayers: A Temperature-Controlled Force Spectroscopy Study. Langmuir. 2012;28(35):12851–12860. doi: 10.1021/la302620t. [DOI] [PubMed] [Google Scholar]
  24. Heaton N. J., Althoff G., Kothe G.. Observation of Lateral Diffusion in Biomembranes by Excitation Transfer 31P NMR: Estimation of Vesicle Size Distributions. J. Phys. Chem. 1996;100(12):4944–4953. doi: 10.1021/jp952301e. [DOI] [Google Scholar]
  25. Althoff G., Stauch O., Vilfan M., Frezzato D., Moro G. J., Hauser P., Schubert R., Kothe G.. Transverse Nuclear Spin Relaxation Studies of Viscoelastic Properties of Membrane Vesicles. II. Experimental Results. J. Phys. Chem. B. 2002;106(21):5517–5526. doi: 10.1021/jp012829l. [DOI] [Google Scholar]
  26. Dolainsky C., Unger M., Bloom M., Bayerl T. M.. Two-Dimensional Exchange 2H NMR Experiments of Phospholipid Bilayers on a Spherical Solid Support. Phys. Rev. E Stat. Phys. Plasmas Fluids Relat. Interdiscip. Top. 1995;51(5):4743–4750. doi: 10.1103/PhysRevE.51.4743. [DOI] [PubMed] [Google Scholar]
  27. Roberts M. F., Redfield A. G.. Phospholipid Bilayer Surface Configuration Probed Quantitatively by 31P Field-Cycling NMR. Proc. Natl. Acad. Sci. U. S. A. 2004;101(49):17066–17071. doi: 10.1073/pnas.0407565101. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Klauda J. B., Roberts M. F., Redfield A. G., Brooks B. R., Pastor R. W.. Rotation of Lipids in Membranes: Molecular Dynamics Simulation, 31P Spin-Lattice Relaxation, and Rigid-Body Dynamics. Biophys. J. 2008;94(8):3074–3083. doi: 10.1529/biophysj.107.121806. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Rüdisser S. H., Goldberg N., Ebert M.-O., Kovacs H., Gossert A. D.. Efficient Affinity Ranking of Fluorinated Ligands by 19F NMR: CSAR and FastCSAR. J. Biomol. NMR. 2020;74(10–11):579–594. doi: 10.1007/s10858-020-00325-x. [DOI] [PubMed] [Google Scholar]
  30. Lipari G., Szabo A.. Model-Free Approach to the Interpretation of Nuclear Magnetic Resonance Relaxation in Macromolecules. 1. Theory and Range of Validity. J. Am. Chem. Soc. 1982;104(17):4546–4559. doi: 10.1021/ja00381a009. [DOI] [Google Scholar]
  31. Clore G. M., Driscoll P. C., Wingfield P. T., Gronenborn A. M.. Analysis of the Backbone Dynamics of Interleukin-1.Beta. Using Two-Dimensional Inverse Detected Heteronuclear Nitrogen-15-Proton NMR Spectroscopy. Biochemistry. 1990;29(32):7387–7401. doi: 10.1021/bi00484a006. [DOI] [PubMed] [Google Scholar]
  32. Lipari G., Szabo A.. Effect of Librational Motion on Fluorescence Depolarization and Nuclear Magnetic Resonance Relaxation in Macromolecules and Membranes. Biophys. J. 1980;30(3):489–506. doi: 10.1016/S0006-3495(80)85109-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Steele H. B. B., Sydor M. J., Anderson D. S., Holian A., Ross J. B. A.. Using Time-Resolved Fluorescence Anisotropy of Di-4-ANEPPDHQ and F2N12S to Analyze Lipid Packing Dynamics in Model Systems. J. Fluoresc. 2019;29(2):347–352. doi: 10.1007/s10895-019-02363-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Rashid S., Lee B. L., Wajda B., Spyracopoulos L.. Side-Chain Dynamics of the Trifluoroacetone Cysteine Derivative Characterized by19 F NMR Relaxation and Molecular Dynamics Simulations. J. Phys. Chem. B. 2019;123(17):3665–3671. doi: 10.1021/acs.jpcb.9b01741. [DOI] [PubMed] [Google Scholar]
  35. Singer P. M., Asthagiri D., Chapman W. G., Hirasaki G. J.. Molecular Dynamics Simulations of NMR Relaxation and Diffusion of Bulk Hydrocarbons and Water. J. Magn. Reson. 2017;277:15–24. doi: 10.1016/j.jmr.2017.02.001. [DOI] [PubMed] [Google Scholar]
  36. Tropp J.. Dipolar Relaxation and Nuclear Overhauser Effects in Nonrigid Molecules: The Effect of Fluctuating Internuclear Distances. J. Chem. Phys. 1980;72(11):6035–6043. doi: 10.1063/1.439059. [DOI] [Google Scholar]
  37. Werbelow L. G., Marshall A. G.. Internal Rotation and Methyl Proton Magnetic Relaxation for Macromolecules. J. Am. Chem. Soc. 1973;95(16):5132–5134. doi: 10.1021/ja00797a006. [DOI] [Google Scholar]
  38. Richarz R., Nagayama K., Wüthrich K.. Carbon-13 Nuclear Magnetic Resonance Relaxation Studies of Internal Mobility of the Polypeptide Chain in Basic Pancreatic Trypsin Inhibitor and a Selectively Reduced Analogue. Biochemistry. 1980;19(23):5189–5196. doi: 10.1021/bi00564a006. [DOI] [PubMed] [Google Scholar]
  39. Ardenkjaer-Larsen J.-H., Boebinger G. S., Comment A., Duckett S., Edison A. S., Engelke F., Griesinger C., Griffin R. G., Hilty C., Maeda H., Parigi G., Prisner T., Ravera E., van Bentum J., Vega S., Webb A., Luchinat C., Schwalbe H., Frydman L.. Facing and Overcoming Sensitivity Challenges in Biomolecular NMR Spectroscopy. Angew. Chem., Int. Ed. 2015;54(32):9162–9185. doi: 10.1002/anie.201410653. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Ni F.. Recent Developments in Transferred NOE Methods. Prog. Nucl. Magn. Reson. Spectrosc. 1994;26:517–606. doi: 10.1016/0079-6565(94)90000-0. [DOI] [Google Scholar]
  41. Zhang G., Ahola S., Lerche M. H., Telkki V.-V., Hilty C.. Identification of Intracellular and Extracellular Metabolites in Cancer Cells Using 13C Hyperpolarized Ultrafast Laplace NMR. Anal. Chem. 2018;90(18):11131–11137. doi: 10.1021/acs.analchem.8b03096. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Ahola S., Zhivonitko V. V., Mankinen O., Zhang G., Kantola A. M., Chen H.-Y., Hilty C., Koptyug I. V., Telkki V.-V.. Ultrafast Multidimensional Laplace NMR for a Rapid and Sensitive Chemical Analysis. Nat. Commun. 2015;6(1):8363. doi: 10.1038/ncomms9363. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

ac5c04414_si_001.pdf (1.9MB, pdf)

Articles from Analytical Chemistry are provided here courtesy of American Chemical Society

RESOURCES