Introduction to Solid State NMRReview of Solid Polymer Characterization, Selected Articles, Book list
The following is an abstract of a presentation made in Angra dos Reis, RJ, Brazil on 13th May, 1997 at the 6th (semiannual) Encontro de Usários of the Associação de Usários de REssonância Magnética Nuclear (AUREMN), the Brazilian NMR Association, and published in that association's Annals. (Reproduced by permission.)
CHARACTERIZATION OF SOLID POLYMERS BY CP/MAS NMR
The application of NMR to the study of polymers has a long history. These studies have been directed to the examinationof both structure and molecular motion. Many of the earliest studies were performed on solid polymers and used low resolution techniques. However, many more studies have been made on polymer solutions and on rubbery polymers, using high resolution techniques.[1,2] More recently, with the development of techniques to obtain high resolution NMR spectra of rigid solids, these methods have been applied to glassy and crystalline polymers as well.[3,4]
There are two basic reasons to choose solid state NMR over solution NMR when studying rigid polymers. First there is the class of systems which cannot be put into solution with out changing them chemically. Cross linked polymers often fall into this class. Other polymers may be so fragile chemically that they decompose upon exposure to solvent. The second class of system that is better studied by solid state methods is not difficult to dissolve nor is it chemically fragile, but it would be changed physically if put into solution. A study designed to examine the physical properties of a solid polymer had better be done on the solid itself if it is to be relevant. There are many examples of relaxation studies made on solids polymers which succeed in examining molecular motion in the solid state.
Special techniques are required to obtain high resolution NMR spectra in the solid state. This is because rapid, isotropic molecular motion, which is responsible for simplifying so many nuclear interactions in the liquid state, is not present in sufficient amount in rigid solids. The special methods either supply the missing motion or remove the troublesome interactions. Depending on whether the spin system being studied is an abundant spin system with abundant internuclear couplings to be dealt with, or a rare spin system coupled primarily to a second spin system of abundant spins, different methods must be used to obtain high resolution NMR spectra. This overview will be principally concerned with the rare spin case, although some abundant spin methods and results will be discussed. We shall also be concerned primarily with natural abundance studies. Finally, unless stated otherwise, only spin one-half nuclei will be considered, most commonly 1H and 13C.
Solid State NMR Methods
The most important solid state interaction to be dealt with, if we are to obtain high resolution spectra, is direct dipolar broadening of the object spin by its neighboring spins. In a study of rare spins, this broadening arises from the nearby abundant spins, frequently protons. It is easily removed by irradiating the abundant spins at their resonant frequency at sufficient rf power to overwhelm the abundant spin local fields. This example of heteronuclear decoupling is simple and straightforward and only needs sufficient rf power. Homonuclear decoupling, as is required to remove dipolar broadening in a system of coupled abundant spins is not so straightforward. It requires one of a number of irradiation schemes, either a train of pulses of carefully chosen length and phase, or suitably off-resonance cw irradiation. The observation of rare spins during heteronuclear decoupling is uncomplicated. Much greater instrumental demands result from the necessity to irradiate and observe the spin system simultaneously, as is required by the multiple-pulse homonuclear decoupling schemes.
Another barrier to high resolution in solids arises as a result of the absence of rapid, isotropic molecular tumbling: chemical shift anisotropy (CSA). The chemical shift associated with each different chemical environment is orientation dependent. For some spins in particularly anisotropic environments, the range of chemical shifts due to anisotropy is almost as large as the full range of isotropic chemical shifts normally seen for that spin. Overlapping CSA patterns obscure the isotropic shift information which is the usual object of an NMR experiment. The CSA is a problem for both rare spins and abundant spins. The solution is to substitute rapid sample rotation about a single axis for the missing molecular tumbling. The angle this axis makes relative to the laboratory magnetic field is called the magic angle, and the whole operation is usually called magic angle sample spinning (MAS). The magic angle happens to turn out to be half the tetrahedral angle or, in other words, the angle the diagonal of a cube makes with any of its edges. A single rotation axis would not work if the CSA did not have a particularly simple angular dependence. For more complicated line broadening functions, such as quadrupolar interactions, techniques requiring rotation about two or more axes are required. Simultaneous high-speed rotation about two different axes has been achieved.
Sensitivity is always a problem for rare spin NMR at natural abundance. It is even more a problem in the solid state where spin-lattice relaxation times, especially for spins without directly bonded abundant spins, can be exceptionally long. The usual approach to this problem is the third important technique used to obtain carbon, and other rare spin, NMR spectra in the solid state. Cross polarization (CP) is used to generate rare spin polarization by causing the rare spins to come to thermal equilibrium with the abundant spin system, which usually relaxes much more rapidly than the individual rare spins.[9,10] This process usually takes the form of a double rotating frame experiment and gives rise to a number of possible rotating frame relaxation experiments, which, as we shall see in the next section, are rich in information.
The combination of experiments described above for rare spins is usually called cross polarization/magic angle spinning (CP/MAS). Abundant spins are observed with combined rotation and multiple pulse spectroscopy (CRAMPS). Resolution is usually not as good in the solid state as that achieved in solution, even in crystalline systems. This is because there are still other sources of line broadening and limits to polarization lifetimes not seen in low viscosity liquids. In the glassy state, even less resolution is to be expected, because of the inherent disorder present in the glass. The isotropic chemical shift is sensitive to intermolecular as well as intramolecular interactions. Each slightly broadened line seen in the solid state spectrum is actually a band of isotropic chemical shifts reflecting the range of intramolecular interactions. Nevertheless, simple high resolution solid state spectra have been used with great effectiveness to answer basic questions about the primary chemical and physical structure in the solid state. The process of cross linking to produce an insoluble polymer can be readily followed as a solid state reaction. Non-equivalence of chemically equivalent atoms caused by restricted rotation in the solid state is also straightforward to detect.
The cross polarization experiment has been described as a double rotating frame experiment. Both the abundant spin and the rare spin need to be spin-locked in an appropriate radiofrequency field if they are to communicate readily. The time constant for the loss of spin-locked polarization, usually characterized by the relaxation time T1rho can be measured in both spin systems. It behaves differently, however, in a rare spin system than in an abundant spin system. In an abundant spin system, T1rho is a spin-lattice relaxation time sensitive to motion at frequencies double the precession frequency around the spin-locking radiofrequency field. Furthermore, it is subject to spin diffusion, that is, spins within a certain distance of each other tend to relax cooperatively. Since this relaxation time can be measured through the rare spin system, it is possible to detect sample heterogeneity if different rare spins detect different groups of abundant spins that are not in good communication. Phase separation in polyblends and block copolymers can be detected in this experiment.
Spin diffusion does not operate effectively between well separated rare spins. Thus each rare spin relaxes at a rate indicative of its own special environment. Furthermore, rare spins may relax directly to the lattice and be sensitive to molecular motion, or relax to the abundant spin system and sense the strength of the coupling to that system. Rare spin T1rho spin lattice contributions are sensitive to motional frequencies near the precession frequency about the radiofrequency field. To use rare spin T1rho as a detector of molecular motion, however, it is first necessary to determine the dominant relaxation mechanism.
Line Shape Analysis
The techniques by which high resolution spectra are obtained in the solid state require the suppression of the dipolar coupling and the averaging away of the CSA. The result is the loss of considerable information. There are ways to recover this lost lineshape information. The CSA information can be recovered by spinning slowly enough to leave spinning sidebands in the spectrum from which the lost CSA lineshape can be inferred by a suitable mathematical analysis. If overlap between spinning sidebands and centerbands proves too complicated to interpret, two-dimensional methods exist which separate the spinning sidebands from the isotropic shifts. The most straightforward way to recover the lost dipolar lineshape information in a rare spin NMR experiment is to perform a two-dimensional NMR experiment in which the dipolar coupling is restored during the t1 evolution period. The data analysis is much clarified if the abundant spins are decoupled from one another while they are allowed to couple to the rare spins. There are a variety of these separated local field (SLF) experiments in current use.[18,19] The lineshape information obtained from these experiments can be combined with relaxation time measurements to characterize both rates and amplitudes of molecular motion. Particularly, molecular motion will leave its mark on a detected lineshape by the extent to which it averages the lineshape expected for the interaction in the absence of motion.
As normally analyzed, the CSA is assumed to arise from a crystal powder average of molecular orientations. Just as an anisotropic motion can average the CSA to an unusual lineshape, an oriented sample will yield an atypical lineshape. Several different techniques have been developed to characterize oriented samples and to calculate lineshapes.
Solid polymers have been characterized extensively by Solid-State NMR methods. Spectral information, linewidths, and relaxation studies give direct information about structure and motion in the solid-state. However, even the most ordered polymers possess a certain amount of heterogeneity, which often affects the properties of the material. Blends of compatible or incompatible polymers, block copolymers, and multiphase systems present even greater tendencies to develop domains of unknown size, which need to be understood to understand the mechanical properties of the solid.[22,23]
One class of NMR experiments which has the potential to determine domain sizes and differences in properties between domains involves the determination of spin diffusion. The time it takes for spin information to cross from one domain to another will tell the size of the domain if the rate of spin diffusion is known. Diffusion is usually measured by setting up some kind of non-uniform concentration gradient and then watching it dissipate. In this case, the concentration gradient is spin polarization or other spin property. In the classical Goldman-Shen experiment, suppression of the signal from one of two communicating systems with a well-enough resolved chemical shift difference was based on differences in precession rate alone. Restoration of equilibrium between the two systems gave information about the rate of communication between them. (This original experiment was all the more remarkable for having been done on a phase-incoherent spectrometer!)
Some spatial information can also be gotten from multi-phase systems which develop their own gradients by means of unequal relaxation rates within the phases. However, in most common, less cooperative cases, it is necessary to enhance the resolution between the parts of the system to be separated before undertaking to prepare gradients by perturbing the population (selective excitation or suppression) of part of the system.[25-27] 2D methods are often useful in the preparation of the system and resolution of the data in these cases.[4,28,29] Spin diffusion studies usually focus on the abundant spin system because spin diffusion is too slow in the rare spin system if labels are not used. For that reason multiple pulse decoupling (often combined with magic angle sample spinning as CRAMPS) is usually the preferred method of enhancing the resolution between parts of the system. It also allows control of communication between abundant spins. Whether the abundant spin or a rare spin ultimately is observed depends on which high resolution spectrum is most sensitive to the effects of spin diffusion.
As in the early days of low resolution solid state NMR, many of the applications of high resolution solid state NMR have been to polymers. It has been used to study solid state reactions, chemical and physical structure, phase separation, and extensively to characterize the frequency, amplitude and particular axes of molecular motion. The basic NMR techniques responsible for high resolution: decoupling, magic angle sample spinning, and cross polarization were early developments in the science of NMR. New approaches to amplify these methods and make them more useful are being developed every day.
The Solid State NMR reading list is in pdf format
NMR of RNA
NMR has moved a long way...
Biomolecular nuclear magnetic resonance (NMR) has advanced rapidly (Methods in Enzymology Vol 239 and 261 ) since the publication in 1989 of the Methods in Enzymology volumes 176 and 177. Much progress has resulted from generalization of heteronuclear two-dimensional (2D) NMR experiments with 13C and15N labeled biomolecules to higher dimensions. Three- and four- dimensional techniques, although straightforward conceptual elaborations of two- dimensional NMR, have proved to be powerful tools to the structural biochemist. The development of sophisticated pulse sequences has driven the development of more elaborate instrumentation ( and vice versa).
Structure determination via NMR is distinctly different from the situation with X-ray crystallographic determination of structure, yielding a single most favorable structure. In contrast with the NMR an effort is made to generate an array of structures which are consistent with the available NMR data in order to define the limits of the structure. It is desirable to make the array of structures large enough to map out the conformational space which will accommodate the available experimental data.
NMR spectroscopy is currently the only method available for the high-resolution solution structure determination of the biomolecules (23-25).
Structure determination by NMR
Reviews of the steps involved in Protocol 1 are given by Wuthrich.
Prior to 1982, assignment of resonances in the complex NMR spectra of proteins was achieved using 1D NMR, and was based, for the most part, on the assumption that the structure of the protein in solution was the same as in the available X-ray structure. The introduction of 2D NMR techniques such as COSY and NOESY in the early 1980s dramatically increased the resolution of protein NMR spectra and as a result the amount of information which could be obtained from the spectra also increased. In 1982 Wuthrich and co-workers introduced a systematic method for the full assignment of the 2D NMR spectra of proteins which relied only on information about the amino acid sequence of the protein; this method is the sequential assignment procedure (1,2). This approach is now used by most researchers who apply NMR spectroscopy to the study of proteins.
NMR Studies of RNA. NMR spectroscopy is the method of choice to study details of the solution structure of nucleic acids (23-25). Known motifs observed in RNA (30) include: double helical stems, hairpin loops (31-33a-c), internal loops, bulges (34a-c), mismatched base pairs (35), multi-stem junctions, base triples (36-38), and pseudoknots (39a,b). The specific shape as well as the hydrogen-bonding properties of RNA will contribute to the recognition between RNA and proteins (22). Understanding the structure, the dynamics and the thermodynamic stability of variety of structural motifs in the IRE will add to understanding of the principles of RNA folding, the specificity of RNA protein interactions, and functional features of mRNA.
Nucleic acids are perhaps the most intensively studied molecules found in nature, which is not surprising in view of their role as carriers of genetic information. It has been a deep-rooted belief among molecular biologist that structure dictates function and, therefore, these studies have often involved the elucidation of structural aspects of these molecules. In a similar spirit it has been a deep-rooted belief among NMR spectroscopists that "their" field of spectroscopy should be able to contribute to the elucidation of the structure and dynamics of molecules, in particular, biomolecules, in a way unparalleled by other forms of spectroscopy. Over the last few years this goal has come within reach thanks to the development of multidimensional NMR techniques which provide the necessary spectral resolution for the successful interpretation of the complicated spectra of biomolecules. Triggered off by the development of two-dimensional NMR, an avalanche of NMR studies on nucleic acids has appeared (1). Methods for spectrum interpretationand structure elucidation have been developed.
1. Helical structures, Hairpin Loops, Internal Loops and Single Base Bulges. Hairpin loops are important structural elements providing sites for interactions with proteins and other nucleic acids and serving as nucleation sites for RNA folding (31,39). Despite this fact, relatively little detailed information is available regarding such RNA structures even in the most recent NMR studies of 24-nucleotide variant of the RNA binding sequence for the coat protein of bacteriophage. R17 structure has not been precisely defined. Most of what is known has been derived from X-ray structure of tRNAs and small duplex oligonucleotides (39).
Recently, NMR spectroscopy has been applied to a variety of small RNA molecules; solution conformation of several RNA hairpins have been reported (31). The 3D structure of individual RNA hairpins is highly variable. Although stems are generally A form, local perturbations due to sequence or structural context are common. Loop conformations vary over a wide range.
Thermodynamic studies on RNA hairpin loops showed that stability depends on the formation of hydrogen bonds between base pairs in the loop (30-34,37). Most recently the three dimensional conformation of a 24-nucleotide variant of the RNA binding sequence for the coat protein of bacteriophage R17 has been analyzed using NMR, molecular dynamics and energy minimization (31). Although the loop structure is still tentative, the known interactions sites with coat protein are easily accessible from the major groove side of the loop. Base pairing in the hairpin loop appears to occur in the IRE deduced from the poor reactivity of G18 and C14 to RNAase T1 and dimethyl sulfate, respectively (15-18; Fig. 1) and the decrease in temperature stability of both the main helix (Tm=10°) and specific features of the imino environment in the 1H spectrum of the G18A mutant (Fig. 6; 1).
The stem of the IRE hairpin is interrupted by a single-base bulge on the 5' side for the five transferrin receptor IREs (a-e) and the eALAS (erythroid aminolevulinate synthase) IRE. A four-nucleotide internal loop characterizes the stem of the ferritin IRE with three nucleotides, UGC, on the 5' side and C on the 3' side. Internal loops and bulges have been previously examined by NMR models for 5S ribosomal RNA (40a-b,41) and the TAR element in HIV viruses (27a-d,42). For example, 2D NMR analysis of the shorter sequence of the E loop from 5S ribosomal RNA of X. laevis (n=27), followed by restrained molecular dynamics showed that the highly conserved internal loop closes to form a G/A base pair and a reverse Hoogsteen A/U base pair; the resulting conformation suggested the possibility of a base triple (36,37). The conformation and dynamics of an RNA oligonucleotide (n=25 nucleotides), which is a model of the tetrahymena group I intron, has also been investigated by one- and two-dimensional NMR spectroscopy (37,38); the exchangeable proton spectrum indicated that two helices stack coaxially with adjacent single-stranded nucleotides forming base triple.
NMR spectroscopy has also been used to analyze the effect of single base bulges on intercalation (34). In these hairpins, single-base bulges seem to strongly facilitate changes to alternate conformations. It seems likely that the functional role for single-base bulges and internal loops in RNAs may be to permit the RNA to switch between different conformations (27d). In this context it is useful to recognize that the internal loop region of the ferritin IRE, which interacts with both initiation factors plus the IRP, has alternate conformations detected by probing with 1,10-phenanthroline-Cu (21).
2. Folded helical structures: Pseudoknots and Other Higher Order Interactions. Higher order interactions of the IRE are likely to be responsible for the folding and unusual reactivity observed with protein nucleases, transition metal complexes (3,4,15,16) and the NMR spectrum of the wild-type compared to the hairpin loop mutant (G18A) IRE sequences (Figs. 5 and 6). Base-triples (36-38), pseudoknots (39a-c) and coaxially stacked helices (38,43) are among essential tertiary interactions in RNA three-dimensional structure.
RNA pseudoknots are found in virtually all classes of RNA and play key roles in a variety of biological processes. Pseudoknots formation folds the 3’ ends of many plant viral genomic RNAs into structures that resemble transfer RNA in global folding and in their reactivity to transfer RNA-specific proteins. The solution structure of the 44- nucleotide pseudoknotted T arm and acceptor arm of the transfer RNA-like structure of turnip yellow mosaic virus (TYMV) was determined at an atomic level in all regions by nuclear magnetic resonance (NMR ) spectroscopy. The structure resulting from this interaction between the minor groove and single -stranded RNA at helical junction display internal mobility, which may be a general feature of RNA pseudoknots that regulates their interaction with proteins or other RNA molecules.
The hammerhead, a RNA oligomer of 40 nucleotides with tertiary interactions, is fully active in the self-cleavage reaction (44). Complete assignment of low-field oligomer ring imino protons between adjacent pairs has led to NOEs indicating a pseudoknot (39a-c). Formation of a pseudoknot in the corresponding oligomer is shown by the existence of two helical stem regions deduced from NOEs between the imino protons. The two-stem regions stack to form a continuous helix with a possible minor distortion in helical stacking at the junctions of stems and loops (39a-c,43). tRNA, the RNA paradigm for 20 years (28a-c), also illustrates how the deduction of stacked helices and junction distortions (16) is made from NMR spectroscopy (38-40). In addition, the results of NMR spectroscopy of tRNA in solution confirmed the L-shape previously observed in crystals (28c).
3. NMR analysis of tRNA
4. RNA/Protein Interactions. RNA/protein interactions (22,50) studied by NMR (51) spectroscopy include Tat/TAR) (27a-c), polycistronic mRNA/viral coat proteins (46-48), 16S RNA/sarcin ribosomal protein S5 (49) and S17 ( ) and ricin ribosomal protein ( ).
The interaction of the TAR stem/loop in HIV RNA with the regulatory protein called Tat activates the transcription of the RNA. The stem in TAR is interrupted by a three-residue bulge in the middle of the 5'-side of the stem. The bulged residue -U23, C24 and U-25 changed conformation when Tat protein binds or when a small basic peptide from Tat binds, or even when free arginamide binds, thus, permitting examination of the RNA structure in the simple complex with arginamide (27a-e). In the absence of arginamide, the three residues in the bulge stack on top of each other, continuing the strand to which they belong in a way that accounts for the observation that the stem as a whole is bent. Arginamide hydrogen bonds to the phosphate groups that link A22 to U23 and U23 to C24 as well as to hydrogen-bond acceptors belonging to the guanine of the GC immediately above the bulge. This appears to position U23 so that it can form a base-triple with the AU immediately above the GC pair adjacent to the bulge which, in turn, causes C24 and U25 to turn outward into solution. Among other things, this conformational change straightens the stem. The importance of the AUU base-triple, which is the most distinctive feature of the model, has been verified by site-directed mutagenesis (27e). The advantage of introducing selectively enriched nucleotides was illustrated first for a TAR sequence (42) by using the 13C chemical shift to provide an alternative method to determine ribose puckers in large RNAs (52,53).
5. Affinity and Specificity
Understanding the molecular determinants of the binding energy of an RNA-protein interaction is a very important but insufficient goal: specificity is equally important. Many RNA-binding proteins bind any RNA weakly, regardless of its sequence or structure. However, biological function requires discrimination of cognate RNAs (the correct, relevant targets) from noncognate RNAs, which are present in very large excess in the cell. The difference in binding energy between cognate and noncognate RNAs defines how specific an RNA-protein interaction is. Understanding this specificity adds a further, intriguing dimension to the characterization of these RNA molecules play a central role in all the main functions of living molecules: storage of genetic information, propagation of the genetic material, and enzymatic activity. RNA molecules do not perform those functions alone, but in tight associations with RNA-binding proteins. Thus, RNA-protein recognition is central to understanding a wide range of biological processes.
RNA-protein recognition cannot be understood with out recognizing that the diversity and complexity of RNA structures are comparable to those of proteins. This structural diversity defines an enormous variety of sites and shapes for intermolecular recognition.
6. Molecular modeling as a tool to find 3D structures of RNA: The analysis of RNA 3-D structures is crucial for understanding the structure-function relationship and evolution of RNA. Molecular modeling aims to find the 3-D structures of RNA that satisfy the structural information obtained by various methods such as electron microscopy, neutron scattering, low resolution X-ray and NMR analysis, site-directed mutagenesis, crosslinking, chemical and biochemical probing, phylogenetic comparison and secondary structure prediction by free energy minimization (27c-d,31,53-55). A current example of the application of NMR analysis and modeling is the RNA binding sequence IRE.
The IRE and IRE/IRP complex. The IRE family is the best characterized group of eukaryotic mRNA regulatory elements and is one of the longest conserved RNA sequences known (n=28-30). A hairpin loop with higher order interactions has been indicated for the IRE by comparing wild-type ferritin mRNA to RNA with altered IRE sequences, using classical protein nucleases, an alkylating reagent and shape selective transition metal complexes as probes (15-18,21); interactions as far away as 20 nucleotides upstream from the IRE have been observed using site-directed mutation and Fe-bleomycin as a probe (17). The results have been related to mRNA function measured as ferritin synthesis in vitro (17,20).
References … coming soon on separate page