X-ray solution scattering studies of protein folding

Protein folding is a reaction in which an extended polypeptide chain acquires maximal packing through formation of secondary and tertiary structures. Compactness and shape are, therefore, critical properties characterizing the process of protein folding. Because the stability of the native state is determined by the subtle free energy balance between the native and denatured states, the characterization of the denatured state is also essential to understand the conformational stability of the native state. We show that solution X-ray scattering is the best technique available today to address these problems. Although the structural resolution of the unfolded or compact denatured states elucidated from solution X-ray scattering is low, it provides a variety of information complementary to that obtained by NMR or X-ray crystallography.     

Reduced-denatured ribonuclease A is not in a compact state

Dynamic light scattering and circular dichroism experiments were performed to determine the compactness and residual secondary structure of reduced and by 6 M guanidine hydrochloride denatured ribonuclease A. We find that reduction of the four disulphide bonds by dithiothreitol at 20 degrees C leads to total unfolding and that a temperature increase has no further effect on the dimension. The Stokes' radius of ribonuclease A at 20 degrees C is R(s) = (1.90 +/- 0.04) nm (native) and R(s) = (3.14 +/- 0.06) nm (reduced-denatured). Furthermore, circular dichroism spectra do not indicate any residual secondary structure. We suggest that reduced-denatured Ribonuclease A has a random coil-like conformation and is not in a compact denatured state.     

Compact thermally-denatured state of a staphylococcal nuclease mutant from resonance energy transfer measurements

Thermal denaturation of a staphylococcal nuclease mutant K78C, where lysine 78 is replaced by cysteine, was studied by circular dichroism (CD) and resonance energy transfer. CD spectra suggest that residual structures remain in the denatured state. Steady-state energy transfer from intrinsic tyrosines to a single and intrinsic tryptophan was measured at different temperatures. In the thermally-denatured state of K78C, there is still a substantial degree of energy transfer from tyrosine(s) to tryptophan, indicating residual structures in the denatured state. The cysteine residue in mutant K78C was labeled with a cysteine specific probe IAEDANS. Fluorescence decays of the tryptophan were measured to estimate distance distributions between Trp 140 and IAEDANS at position 78. Measurements were done as a function of temperature from 4 degrees C (native) to 65 degrees C (denatured) both with and without Ca2+ and inhibitor pdTp. Below 30 degrees C, the apparent distance distribution of both the ligand-free nuclease and the enzyme with bound pdTp can be adequately described by a Gaussian model. Above 40 degrees C, where the ligand-free nuclease but not the ternary complex begins to denature, two different populations are required to fit the data both with and without pdTp. One population has a compact structure and the other has an expanded structure. As temperature rises, the population of the expanded structure increases. At the highest temperature, the non-native compact structure is still the major form (60 to 70%). The overall thermally-denatured states of staphylococcal nuclease mutant K78C in the absence and presence of ligands are thus compact and heterogeneous.     

Cortical interneurons upregulate neurotrophins in vivo in response to targeted apoptotic degeneration of neighboring pyramidal neurons

The pressure-induced unfolding of wild-type staphylococcal nuclease (Snase WT) was studied using synchrotron X-ray small-angle scattering (SAXS) and Fourier-transform infrared (FT-IR) spectroscopy, which monitor changes in the tertiary and secondary structural properties of the protein upon pressurization. The experimental results reveal that application of high-pressure up to 3 kbar leads to an approximate twofold increase of the radius of gyration Rg of the native protein (Rg approximately 17 A) and a large broadening of the pair-distance-distribution function, indicating a transition from a globular to an ellipsoidal or extended chain structure. Analysis of the FT-IR amide I' spectral components reveals that the pressure-induced denaturation process sets in at 1.5 kbar at 25 degrees C and is accompanied by an increase in disordered and turn structures while the content of beta-sheets and alpha-helices drastically decreases. The pressure-induced denatured state above 3 kbar retains nonetheless some degree of beta-like secondary structure and the molecule cannot be described as a fully extended random coil. Temperature-induced denaturation involves a further unfolding of the protein molecule which is indicated by a larger Rg value and significantly lower fractional intensities of IR-bands associated with secondary-structure elements. In addition, we have carried out pressure-jump kinetics studies of the secondary-structural evolution and the degree of compactness in the folding/unfolding reactions of Snase. The effect of pressure on the kinetics arises from a larger positive activation volume for folding than for unfolding, and leads to a significant slowing down of the folding rate with increasing pressure. Moreover, the system becomes two-state under pressure. These properties make it ideal for probing multiple order parameters in order to compare the kinetics of changes in secondary structure by pressure-jump FT-IR and chain collapse by pressure-jump SAXS. After a pressure jump from 1 bar to 2.4 kbar at 20 degrees C, the radius of gyration increases in a first-order manner from 17 A to 22.4 A over a timescale of approximately 30 minutes. The increase in Rg value is caused by the formation of an extended (ellipsoidal) structure as indicated by the corresponding pair-distance-distribution function. Pressure-jump FT-IR studies reveal that the reversible first order changes in beta-sheet, alpha-helical and random structure occur on the same slow timescale as that observed for the scattering curves and for fluorescence. These studies indicate that the changes in secondary structure and chain compactness in the folding/unfolding reactions of Snase are probably dependent upon the same rate-limiting step as changes in tertiary structure.     

High populations of non-native structures in the denatured state are compatible with the formation of the native folded state

The structures of the denatured states of the spectrin SH3 domain and a mutant designed to have a non-native helical tendency at the N terminus have been analyzed under mild acidic denaturing conditions by nuclear magnetic resonance methods with improved resolution. The wild-type denatured state has little residual structure. However, the denatured state of the mutant has an approximately 50% populated helical structure from residues 2 to 14, a region that forms part of the beta-sheet structure in the folded state. Comparison with a peptide corresponding to the same sequence shows that the helix is stabilized in the whole domain, likely by non-local interactions with other parts of the protein as suggested by changes in a region far from the mutated sequence. These results demonstrate that high populations of non-native secondary structure elements in the denatured state are compatible with the formation of the native folded structure.     

Thermodynamic characterization of the coupled folding and association of heterodimeric coiled coils (leucine zippers)

Folding thermodynamics of nine heterodimeric, parallel coiled coils were studied by isothermal titration calorimetry (ITC) and thermal unfolding circular dichroism measurements. The heterodimers were composed of an acidic and a basic 30-residue peptide, which when in isolation were monomeric and essentially unstructured. The reaction followed a two-state mechanism indicating that folding and association were coupled. delta Hfold, delta Sfold and delta Cp normalized per mol of residue were of the same magnitude as for monomeric globular proteins, hence the energetics of folding and association of the heterodimeric coiled coils was balanced similarly to the folding of a single polypeptide chain. Cavity creating Leu/Ala substitutions revealed strong and position-dependent energetic coupling between leucine residues in the hydrophobic core of the coiled coil. delta Gunfold (equivalent to -delta Gfold in the two-state reaction) was determined from thermal unfolding. Global stability curves were calculated according to the Gibbs-Helmholtz equation and using the combined free energy data from ITC and thermal unfolding. Maximum stabilities were between 15 and 37 degrees C and cold denaturation could be demonstrated by direct calorimetry. The stability curves were based on free energies of folding measured between 10 and 85 degrees C and under identical solvent conditions. This represents a novel experimental approach which circumvents the use of varying solvent conditions as is typically required to measure protein stability curves. Discrepancies were noticed between van't Hoff enthalpies deduced from thermal unfolding and measured by direct calorimetry. The discrepancies are thought to be due to residual ordered structure in the denatured single chains around room temperature but not near the transition midpoint temperature Tm. This demonstrates that over an extended temperature range the assumption of a common denatured state implicit in the van't Hoff analysis may not always be valid.     

Denatured states of yeast phosphoglycerate kinase

Structures of proteins in unfolded states have important implications for the protein folding problem and for the translocation of polypeptide chains. Acid-denatured, cold-denatured, and 6 M guanidine hydrochloride (GuHCl) denatured yeast phosphoglycerate kinase (PGK) are ensembles of flexible unfolded molecules with rapidly interconverting structures of the individual polypeptide chains. They differ, however, in their physical properties, such as in coil size and in stiffness over a short distance along the chain. These properties of polypeptide chains can be described well by persistence statistics. A solution containing 0.7 M GuHCl at 4.5 degrees C is nearly a Theta-solvent for PGK. By contrast, 6 M GuHCl is a good solvent for PGK. Acid-denatured PGK at low ionic strength has the most expanded and stiffest chains. The conformation of heat-denatured PGK should be more compact than that of random walk chains at the Theta-point, as can be inferred from measurements on other proteins. Investigations of heat-denatured PGK by scattering methods are unfeasible due to aggregation of the protein. The persistence length as a measure of chain stiffness varies between a = 1.74 nm for cold-denatured PGK and a = 3.0 nm for acid-denatured PGK. The distribution functions of the gyration radii were calculated from the X-ray scattering data for all unfolded states and compared with the radius of gyration of the natively folded molecule.     

Conformations and folding of lysozyme ions in vacuo

Proton transfer reactivity of isolated charge states of the protein hen egg-white lysozyme shows that multiple distinct conformations of this protein are stable in the gas phase. The reactivities of the 9+ and 10+ charge state ions, formed by electrospray ionization of "native" (disulfide-intact) and "denatured" (disulfide-reduced) solutions, are consistent with values calculated for ions in their crystal structure and fully denatured conformations, respectively. Charge states below 8+ of both forms, formed by proton stripping, have similar or indistinguishable reactivities, indicating that the disulfide-reduced ions fold in the gas phase to a more compact conformation.     

Structural changes of beta-lactoglobulin B induced by urea. Evidence of residual structure

Several spectroscopic methods have been used to study the structure of beta-lactoglobulin B at pH 2.1 in the presence of 8M urea. Fluorescence and polarization of fluorescence spectroscopy measurements indicate that the two tryptophanyl residues of the protein are exposed to the solvent in the denatured state. CD in the far-UV indicates that the amount of secondary structure in the denatured state is comparable to that found in the native state, whereas the CD spectrum in the near-UV shows that the tertiary structure is not completely disordered. The results of one-dimensional 1H NMR spectroscopy show that some local non-random structure is maintained in the denatured state, but most of the polypeptide chain has an extended non-globular conformation under the conditions of the present experiments. This conclusion is reinforced by the results of two-dimensional 1H NMR conducted on denatured samples of beta-lactoglobulin B. The study of states with intermediate levels of order will aid the understanding of how the native structure of beta-lactoglobulin B is organised during the refolding pathways.     

Monitoring the sizes of denatured ensembles of staphylococcal nuclease proteins: implications regarding m values, intermediates, and thermodynamics

Fluorescence and size-exclusion chromatography (SEC) are used to monitor urea denaturation of wild-type staphylococcal nuclease (SN) as well as the m+ and m- mutants A69T and V66W, respectively. It is found that the SEC partition coefficient, 1/Kd, is directly proportional to the Stokes radii of proteins. From the Stokes radii, the denatured ensembles of the three proteins are found to be highly compact in the limit of low urea concentration and expand significantly with increasing urea concentration. The m values from fluorescence-detected denaturation of the SN proteins are generally considered to reflect the relative sizes of denatured ensembles. However, the rank order of m values of the SN proteins studied do not correspond to the rank order of denatured ensemble sizes detected by 1/Kd, suggesting that m values reflect more than just surface area increases on denaturation. SEC provides two complementary ways to demonstrate the existence of intermediates in urea denaturation and illustrates that V66W undergoes a three-state transition. Fluorescence-detected urea denaturations of A69T and wt SN do not correspond with 1/Kd-detected denaturation profiles, a result that would ordinarily mean that the transitions are non-two-state. However, this interpretation fails to recognize the rapidly changing size and thermodynamic character of the denatured ensembles of these proteins both within and outside of the transition zone. The implications of the changing sizes and thermodynamic character of the denatured ensembles for SN proteins are manifold, requiring a reconsideration of the thermodynamics of proteins whose denatured ensembles behave as those of SN proteins.     

The determinants of fat intake in a multi-ethnic New Zealand population. Fletcher Challenge--University of Auckland Heart and Health Study Management Committee

The problem of a variety of denatured forms of the protein molecule under equilibrium conditions is considered. The experimental conditions are described at which the protein molecule can exist in various non-native states. The history of the discovery of a universal intermediate molten globule state and the current status of research in this field are briefly outlined. Particular emphasis is placed on the fact that the molten globule state is a thermodynamic state of the protein molecule that is separated from both the native and the completely unfolded state by "all-or-none" transitions, i.e., intramolecular analogs of the 1st-order phase transitions. It is also shown that the molten globule state is not the only intermediate state observed for a particular protein under equilibrium conditions. The main structural features of the protein molecule in various denatured conformations are described. How many molten globule states there exist? A molten globule, a precursor of the molten globule, a highly structured molten globule: are these particular conformational states or different forms of the unique intermediate state? Or different forms of the native protein molecule with different degrees of disorder? Or differently structured forms of the unfolded polypeptide chain? This review is an attempt to answer these questions.     

Association-induced folding of globular proteins

It has generally been assumed that the aggregation of partially folded intermediates during protein refolding results in the termination of further protein folding. We show here, however, that under some conditions the association of partially folded intermediates can induce additional structure leading to soluble aggregates with many native-like properties. The amount of secondary structure in a monomeric, partially folded intermediate of staphylococcal nuclease was found to double on formation of soluble aggregates at high protein or salt concentrations. In addition, more globularity, as determined from Kratky plots of small-angle x-ray scattering data, was also noted in the associated states.     

Calmodulin binding to myosin light chain kinase begins at substoichiometric Ca2+ concentrations: a small-angle scattering study of binding and conformational transitions

We have used small-angle scattering to study the calcium dependence of the interactions between calmodulin (CaM) and skeletal muscle myosin light chain kinase (MLCK), as well as the conformations of the complexes that form. Scattering data were measured from equimolar mixtures of a functional MLCK and CaM or a mutated CaM (B12QCaM) incompetent to bind Ca2+ in its N-terminal domain, with increasing Ca2+ concentrations. To evaluate differences between CaM-enzyme versus CaM-peptide interactions, similar Ca2+ titration experiments were performed using synthetic peptides based on the CaM-binding sequence from MLCK (MLCK-I). Our data show there are different determinants for CaM binding the isolated peptide sequence compared to CaM binding to the same sequences within the enzyme. For example, binding of either CaM or B12QCaM to the MLCK-I peptide is observed even in the presence of EGTA, whereas binding of CaM to the enzyme requires Ca2+. The peptide studies also show that the conformational collapse of CaM requires both the N and C domains of CaM to be competent for Ca2+ binding as well as interactions with each end of MLCK-I, and it occurs at approximately 2 mol of Ca2+/mol of CaM. We show that CaM binding to the MLCK enzyme begins at substoichiometric concentrations of Ca2+ (< or = 2 mol of Ca2+/mol of CaM), but that the final compact structure of CaM with the enzyme requires saturating Ca2+. In addition, MLCK enzyme does bind to 2Ca2+ x B12QCaM, although this complex is more extended than the complex with native CaM. Our results support the hypothesis that CaM regulation of MLCK involves an initial binding step at less than saturating Ca2+ concentrations and a subsequent activation step at higher Ca2+ concentrations.     

Interactions in nonnative and truncated forms of staphylococcal nuclease as indicated by mutational free energy changes

Several mixed disulfide variants of staphylococcal nuclease have been produced by disulfide bond formation between nuclease V23C and methane, ethane, 1-propane, 1-n-butane, and 1-n-pentane thiols. Although CD spectroscopy shows that the native state is largely unperturbed, the stability toward urea-induced unfolding is highly dependent on the nature of the group at this position, with the methyl disulfide protein being the most stable. The variant produced by modification with iodoacetic acid, however, gives a CD spectrum indicative of an unfolded polypeptide. Thiol-disulfide exchange equilibrium constants between nuclease V23C and 2-hydroxyethyl disulfide have been measured as a function of urea concentration. Because thiol-disulfide exchange and unfolding are thermodynamically linked, the effects of a mutation (disulfide exchange) can be partitioned between various conformational states. In the case of unmodified V23C and the 2-hydroxyethyl protein mixed disulfide, significant effects in the nonnative states of nuclease are observed. Truncated forms of staphylococcal nuclease are thought to be partially folded and may be good models for early folding intermediates. We have characterized a truncated form of nuclease comprised of residues 1-135 with a V23C mutation after chemical modification of the cysteine residue. High-resolution size-exclusion chromatography indicates that modification brings about significant changes in the Stokes radius of the protein, and CD spectroscopy indicates considerable differences in the amount of secondary structure present. Measurement of the disulfide exchange equilibrium constant between this truncated protein and 2-hydroxyethyl disulfide indicate significant interactions between position 23 and the rest of the protein when the urea concentration is lower than 1.5 M.(ABSTRACT TRUNCATED AT 250 WORDS)     

Folding intermediates of wild-type and mutants of barnase. I. Use of phi-value analysis and m-values to probe the cooperative nature of the folding pre-equilibrium

It is difficult to determine whether transient folding intermediates have a cooperative (or first-order) folding transition without measuring their rates of formation directly. An intermediate I could be formed by a second-order transition from a denatured state D that is progressively changed into I as conditions are changed. We have not been able to monitor the rate of formation of the folding intermediate of barnase directly, but have analysed its reactivity and the equilibrium constant for its formation over a combination of wide ranges of temperature, concentration of denaturant and structural variation. Phase diagrams have been constructed for wild-type and 16 mutant proteins to map out the nature of the energy landscape of the denatured state. The free energy of unfolding of I, delta GD-I, changes with [urea] according to a highly cooperative transition. Further, mD-I (= delta delta GD-I/delta [urea]) for wild-type and several mutants is relatively insensitive to temperature, as would be expected for an intermediate that is formed cooperatively, rather than one that melts out according to a second-order transition. The phi-values for the formation of I change abruptly through the folding transitions rather than have the smooth changes expected for a second-order transition. There is a subset of mutants for which both mD-I and phi-value analysis indicate that a second intermediate becomes populated close to the melting temperatures of the native proteins. The folding intermediate of barnase is, thus, a relatively discrete and compact entity which is formed cooperatively.     

Titration properties and thermodynamics of the transition state for folding: comparison of two-state and multi-state folding pathways

CI2 folds and unfolds as a single cooperative unit by simple two-state kinetics, which enables the properties of the transition state to be measured from both the forward and backward rate constants. We have examined how the free energy of the transition state for the folding of chymotrypsin inhibitor 2 (CI2) changes with pH and temperature. In addition to the standard thermodynamic quantities, we have measured the overall acid-titration properties of the transition state and its heat capacity relative to both the denatured and native states. We were able to determine the latter by a method analogous to a well-established procedure for measuring the change in heat capacity for equilibrium unfolding: the enthalpy of activation of unfolding at different values of acid pH were plotted against the average temperature of each determination. Our results show that the transition state of CI2 has lost most of the electrostatic and van der Waals' interactions that are found in the native state, but it remains compact and this prevents water molecules from entering some parts of the hydrophobic core. The properties of the transition state of CI2 are then compared with the major folding transition state of the larger protein barnase, which folds by a multi-state mechanism, with the accumulation of a partly structured intermediate (Dphys or I). CI2 folds from a largely unstructured denatured state under physiological conditions via a transition state which is compact but relatively uniformly unstructured, with tertiary and secondary structure being formed in parallel. We term this an expanded pathway. Conversely, barnase folds from a largely structured denatured state in which elements of structure are well formed through a transition state that has islands of folded elements of structure. We term this a compact pathway. These two pathways may correspond to the two extreme ends of a continuous spectrum of protein folding mechanisms. Although the properties of the two transition states are very different, the activation barrier for folding (Dphys-->++) is very similar for both proteins.     

Oligomers of the cytoplasmic fragment from the Escherichia coli aspartate receptor dissociate through an unfolded transition state

The kinetic and equilibrium properties of a clustering process were studied as a function of temperature for two point mutants of a 31 kDa fragment derived from the cytoplasmic region of the Escherichia coli aspartate receptor (C-fragment), which were shown previously to have a greater tendency to form clusters relative to the wild-type C-fragment [Long, D. G., & Weis, R. M. (1992) Biochemistry 31, 9904-9911]. The clustering equilibria were different for the two C-fragments. Monomers of a serine-461 to leucine (S461L) mutant C-fragment were in equilibrium with dimers, while monomers of a S325L C-fragment were in equilibrium with trimers. The positive values for delta H degree, delta S degree, and delta Cp degree of dissociation estimated from a van't Hoff analysis, and the differences in the CD spectra of isolated monomers and oligomers, demonstrated that the monomers were less well-folded than the clustered forms. The oligomer dissociation rate exhibited a marked temperature dependence over the range from 4 to 30 degrees C and was remarkably slow at low temperatures; e.g. t1/2 of dimer dissociation for the S461L C-fragment was 85 h at 4 degrees C. The values for delta H degree +2, delta S degree +2, and delta Cp degree +2 derived from the temperature dependence of the dissociation rate were comparable to the corresponding parameters determined in a DSC study of C-fragment denaturation [Wu, J., Long, D. G., & Weis, R. M. (1995) Biochemistry 34, 3056-3065], which indicated that the transition state resembled thermally denatured C-fragment. Octyl glucoside accelerated the dissociation rate by 3-5-fold presumably by lowering the barrier to dissociation. This acceleration and the positive value of delta Cp degree +2 were interpreted as evidence for an increase in solvent accessible hydrophobic groups in the transition state. The molecular basis for the slow rate of dissociation is proposed to result from the conversion of intermolecular coiled coils in the oligomers to an intramolecular coiled coil in the monomer.     

Non-native local interactions in protein folding and stability: introducing a helical tendency in the all beta-sheet alpha-spectrin SH3 domain

The relative importance of secondary structure interactions versus tertiary interactions for stabilising and guiding the folding process is a matter for discussion. Phenomenological models of protein folding assign an important role to local contacts in protein folding and stability. On the other hand, simplistic lattice simulations find that secondary structure is mainly the product of protein compaction and that optimisation of folding speed seems to require small contributions of local contacts to the stability of the folded state. To examine the extent to which secondary structure propensities influence protein folding and stability, we have designed mutations that introduce a strong non-native helical propensity in the first 19 residues of the alpha-spectrin SH3 domain. The mutant proteins have the same three-dimensional structure as the wild-type, but they are less stable and have less co-operative folding transitions. There seems to be a relationship between the non-native helical propensity and the compaction of the denatured state. This suggests that in the denatured ensemble under native conditions there is a significant proportion of compact structures with non-native secondary structures. Our results demonstrate that non-local interactions can overcome strong non-native secondary structure propensities and, more important, that optimisation of folding speed and co-operativity requires the latter to be relatively small.     

Structural characterisation and comparison of the native and A-states of equine lysozyme

Native state 1H NMR resonance assignments for 125 of the 129 residues of equine lysozyme have enabled measurement of the hydrogen exchange kinetics for over 60 backbone amide and three tryptophan indole hydrogen atoms in the native state. Native holo equine lysozyme hydrogen exchange protection factors are as large as 10(6), the most protected residues being located in elements of secondary structure. High exchange protection in the domain interface correlates with the binding of Ca2+ in this region. Equine lysozyme differs from most non-Ca2+ binding lysozymes in forming a highly populated partially folded state at low pH. The protein in this A-state at pH 2.0 has been found to bind 1-anilino-naphthalene-8-sulphonate with the enhancement of fluorescent intensity and blue shift in the spectral maximum characteristic of molten globules. NMR spectra indicate that the A-state is globally much less ordered than native equine lysozyme but does not contain significant regions of random coil structure. The amides most protected against hydrogen exchange in the A-state (protection factors up to 10(2) at 5 degrees C) correspond to residues of three of the four alpha-helices of the native state; the side-chains of these residues form a hydrophobic cluster that includes five aromatic residues. Circular dichroism and tryptophan fluorescence indicate that these residues are substantially more constrained than similar residues in "classical" molten globules. Taken together, the data suggest a model for the A-state of equine lysozyme in which a more ordered core is surrounded by a less ordered but still compact polypeptide chain.     

Thermally denatured ribonuclease A retains secondary structure as shown by FTIR

Fourier transform-infrared (FTIR) spectroscopy has been used to test for the presence of nonrandom structure in thermally denatured ribonuclease A (RNase A) at pH* 2.0 (uncorrected pH measured in D2O). The amide I spectral region of the native and thermally denatured protein was compared. A substantial decrease in the amount of beta-sheet and alpha-helix and a corresponding increase in the amount of turn and unordered structure was observed on thermal denaturation. The results indicate that thermally denatured RNase A contains significant amounts of secondary structure (11% helix and 17% beta-sheet), consistent with previous results reported for circular dichroism, and with a relatively compact structure, as revealed by dynamic light scattering. These results are in contrast to those of amide protection experiments reported recently [Robertson, A.D., & Baldwin, R.L. (1991) Biochemistry 30, 9907-9914] which indicated no stable hydrogen-bonded structure under these experimental conditions. Possible explanations for this apparent discrepancy are given.