NMR Studies of the Order and Dynamics of Dipalmitoylphosphatidylchoiine Bilayers as a Function of Pressure

We have used 2H NMR methods to examine the order and dynamics of dipalmitoylphosphatidylcholine (DPPC) in multilamellar and small unilamellar vesicles in water as a function of pressure. Multipulse 2H NMR techniques were used with selectively deuterated DPPC on both chains at positions C-2, C-9, or C-13, to obtain lineshapes, spin-lattice relaxation times (T1), and spin-spin relaxation times (T2) at 50 °C from 1 bar to 5.2 kbar pressure. This pressure range allowed us to explore the phase behavior of DPPC from the liquid crystalline (LC) phase through various gel phases (Gl, Gll, Glll, GX), including the interdigited Gi phase. Pressure has an ordering effect: on all chain segments in all the phases. In the LC phase, the order parameter (SCD) decreases from C-2 > C-9 > C-13, while in the gel phases SCD decreases from C-9 > C-13 > C-2, indicating that in the gel phases the middle segments of the chains are more restricted in their motions than the ends. In the LC phase, T1 and T2 values for all segments decrease with pressure and have an order from C-13 > C-9 > C-2. These results suggest that similar conformational motions and molecular rotational motions occur in the LC state in all segments, but have increased amplitudes and frequencies toward the methyl ends. At the phase transitions, discontinuities and abrupt reversal of the slopes for the T1 or T2 dependences on pressure indicate major changes in motional modes and rates for DPPC molecules in the different structures. In the second part of this study, we have measured the lateral diffusion of DPPC in sonicated vesicles in D2O as a function of pressure. The spin-lattice relaxation rate in the rotating frame T−11p was plotted as a function of the square root of the spin-locking field angular frequency (ω1)1/2, and the lateral diffusion coefficient (D) was calculated from the slope. Pressure effects are observed on lateral diffusion in the LC phase (D = 5.4 − 2 × 10−9 cm2 seconds, from 1 to 300 bar) but are negligible in the GI phase (D ≈ 1.0 × 10−9 cm2 seconds, from 400 to 800 bar).

Sequence, Salt, Charge, and the Stability of DNA at High Pressure

The effect of pressure on the helix-coil transition temperature (Tm) is reported for the double-stranded polymers poly(dA)poly(dT), poly[d(A-T)], poly[d(l-C], and poly[d(G-C] and triple-stranded poly(dA)2poly(dT). The Tm increases as a function of pressure, implying a positive volume change for the transition and leading to the conclusion that the molar volume of the coil form is larger than the molar volume of the helix. From the change in Tm as a function of pressure, molar volume changes of the transition (ΔVt) are calculated using the Clapeyron equation and calorimetrically determined enthalpies. For the doublestranded polymers, ΔVt, increases in the order poly[d(l-C] < polyt[d(A-T)] < poly(dA)-poly(dT) < polylcl(G-C)]. The value of ΔVt, for the triple-stranded to single-stranded transition of poly(dA) 2poly(dT) is larger than that of poly[d(G-C)I. The magnitude of ΔVt increases with salt concentration in all cases studied; however, the change of ΔVt with salt concentration depends on the sequence of the DNA and the number of strands involved in the transition. In the model proposed to explain the results, the overall molar volume change of the transition is a function of a negative volume change arising from changes in the electrostatic interactions of the DNA strands, and a positive volume change due to unstacking the bases. The model predicted the direction of the change in the ΔVt for several experiments. The magnitude of AVJ increases with counter ion radius, thus for polyld(A-T)], ΔVt, increases in the series Na+ , K+, Cs+, The ΔVt also increases if the charge on the phosphodiester groups is removed. The kinetics of the formation of double-stranded (dA)19(dT)19 in 50 mM NaCI are slowed approximately 14-fold at 200 MPa relative to atmospheric pressure. The implied volume of activation of +37 ml mol−l in the direction of this change is also in agreement with the proposed model. The stability of double- and triple-stranded DNA helices in water around neutral pH depends on the base composition and sequence, as well as on the ionic strength of the solution. Each of these dependencies also defines how DNA interacts with water.

Osmotic and Hydrostatic Pressure as Tools to Study Molecular Recognition

The question of molecular recognition is a central paradigm of molecular biology, playing central roles in most, if not all, cellular processes. Failed recognition events have been implicated in numerous disease states, ranging from flawed control of gene regulation and cellular proliferation to defects in specific metabolic activities. Historically, questions of molecular recognition have been approached through organic synthesis and through actual structural studies of biomolecular complexes. Fundamental insight into the mechanisms of molecular recognition can be realized through the use of broad interdisciplinary tools and techniques. In particular, the use of recombinant DNA technology in concert with hydrostatic and osmotic pressure methodologies have proven to be ideal for understanding the fundamental mechanisms of recognition. In our presentation, we will focus on recent results from our laboratory that examine three major classes of recognition events in biological systems: 1. Protein-protein recognition: here we seek to define the role of specific surface interactions; electrostatic, hydrogen bonding, and hydrophobic free energies provided through surface complimentarity, which define the specificity and affinity in the formation of complexes between the metalloproteins involved in electron transfer events in cytochrome P-450-dependent oxygenase catalysis and in the assembly of tetrameric hemoglobin. 2. Protein—small molecule recognition: here we seek to ascertain how the same fundamental forces of electrostatics, hydrogen bonding, and the hand-glove fit of a substrate into the active site of an enzyme can give rise to the observed high degree of control of regio- and stereo-specificity in catalysis and in the interfadal interactions of proteins at electrode interfaces. 3. Protein nucleic acid recognition: here again the same fundamental forces control recognition processes, but in this case we will focus on our exciting, recent discovery of a role for solvent water in mediating recognition between protein and nucleic acid components. Representative systems in the binding/ catalytic class of restriction endonucleases and recombinases will be discussed. In all cases, the use of pressure as a variable has provided unique understanding for the molecular details of these processes. Pressure, both hydrostatic and osmotic, has proven to be an enabling experimental technique in understanding the mechanistic origins of molecular recognition events.

Correlation Field Splitting of Chain Vibrations: Structure and Dynamics in Lipid Bilayers and Biomembranes

Pressure-tuning vibrational spectroscopy was first introduced to the study of structural and dynamic properties in biological systems from our laboratory about one decade ago. One of our efforts has been the search for spectral features and their pressure dependencies related to the structural and dynamic properties in biological systems. Pressure-induced correlation field splitting of the vibrational modes of methylene chains is one of the parameters that has been applied to monitor various structural and dynamic properties of a wide range of aqueous lipid bilayers and biomembranes in our laboratory. Correlation field splitting of the vibrational modes of the methylene chains in lipid bilayers is the result of vibrational coupling interactions among the ordered methylene chains with different site symmetry in the two-dimensional matrix. However, the basic theory and the characteristics of these interchain interactions in lipid bilayers still needed to be established. It was unknown whether the interchain interactions that result in the correlation field splitting take place within each lipid molecule or between neighboring molecules in the lamellar bilayers. The relative contributions of 'intramolecular and intermolecular interchain interactions to the correlation field splitting, and the effects of the long-range interchain interactions and interdigitation on the correlation field splitting, were also unknown. These problems have been resolved recently and are addressed in this chapter. Our laboratory has pioneered the study of structural and dynamic properties of biological systems by means of pressure-tuning vibrational spectroscopy (Wong et al., 1982). It is now well recognized that this spectroscopic technique is one of the most powerful physical methods for the study of biological and biomedical phenomena at the molecular level with enhanced resolution (Wong, 1984, 1987a, 1987b, 1987c, 1993). The biological systems we have studied by this method include not only various aqueous biomolecular assemblies but also whole cells and intact biological tissues (Rigas et al, 1990; Wong, 1984, 1987b, 1987c, 1993; Wong et al., 1991a, 1991b, 1993). We have found that the pressure-induced changes in many spectral features and parameters in both FTIR and Raman spectra of biological systems result from modifications in structure and dynamics at the molecular level.

Transient Enzyme Kinetics at High Pressure

In a detailed study of an enzyme reaction pathway, a measured composite rate constant, for example, kcat, can be interpreted in ways that lead to ambiguous conclusions. Two conditions must be met to solve this problem: (1) an elementary rate constant must be measured, and (2) a maximum number of physical-chemical parameters must be used to perturb the system under study. To gain access to elementary rate constants, cryobaroenzymology and/or transient methods, such as stopped-flow and flow-quench kinetics, can be used. Both perturbation and kinetics measurements performed under either high pressure or low temperatures can then be used to probe the thermodynamics of the interconversion of two successive intermediates to obtain parameters such as ΔG‡, ΔS‡, ΔH‡, and ΔV‡ The interdependence of the two major variables, namely temperature and pressure, is presented in this article, in which the role of organic cosolvents is considered as a third variable. During catalytic reactions, enzymes undergo a number of conformational changes related to their dynamic structural flexibility. This appears as a succession of different steps. A complete study of such processes, which generally are very rapid, consists of the exploration of the properties of these steps, including thermodynamic features obtained by the action of temperature and pressure. As long ago as 1950, Laidler (1950) formulated the first theoretical basis for explaining the responses of enzymes to high hydrostatic pressures. Chemists used this parameter extensively, and in the early stages of high-pressure kinetics they attempted to analyze the observed results on the basis of collision theory (Asano, 1991) or transition-state theory (Evans & Polanyi, 1935). These theories are still used to describe pressure effects on enzyme reactions. It is postulated that between two successive intermediates there is a labile transition state which governs the energetics of the reaction (Glastone et al., 1941). But we must remember that this theory was first applied only to simple homogeneous reactions in gases. For solutions, the treatment can require the introduction of other parameters such as the viscosity.

High-Pressure NMR Studies of the Dissociation of Arc Represser and the Cold Denaturation of Ribonuclease A

We begin this article with a brief discussion of the specialized high-resolution NMR instrumentation developed for high-pressure studies of biochemical systems. We then present the potential for the unique information content of high-pressure NMR spectroscopy as illustrated by the results of two NMR studies performed recently in our laboratory. Different denatured states of Arc represser are characterized by one-dimensional (1D) and two-dimensional (2D) NMR. Increasing pressure promotes sequential changes in the structure of Arc represser: from the native dimer through a predissociated state to a denaturated molten globule monomer. A compact state (molten globule) of Arc represser is obtained in the dissociation of Arc represser by pressure, whereas high temperature and urea induce dissociation and unfolding to less structured conformations. The presence of NOEs (Nuclear Overhauser Enhancement) in the β-sheet region in the dissociated state suggests that the intersubunit β-sheet (residues 6–14) in the native dimer is replaced by an intramonomer β-sheet. Changes in 2D NMR spectra prior to dissociation indicate the existence of a predissociated state that may represent an intermediate stage in the folding and subunit association pathway of Arc represser. The cold denaturation study of ribonuclease A has shown that high pressure can be utilized not only to perturb the protein structure in a controlled way but also to lower the freezing point of aqueous protein solutions substantially. As a result, one can access subzero temperatures and carry out cold denaturation studies of proteins. The results of the NMR study of the reversible cold denaturation are compared with the heat and pressure denaturation of bovine pancreatic ribonuclease A. High-resolution NMR spectra of complex molecules in the liquid phase usually exhibit a great deal of structure and yield a wealth of information about the molecule. Therefore, it is not surprising that multinuclear high-resolution Fourier transform NMR spectroscopy at high pressure represents the most promising technique in studies of the pressure effects on biochemical systems (Jonas & Jonas, 1994). The high information content of the various advanced NMR techniques, including 2D NMR techniques such as NOESY, COSY, and ROESY, have yet to be fully exploited in high-pressure NMR experiments.

High-Pressure FTIR Studies of the Secondary Structure of Proteins

Infrared spectra of five globular proteins (bovine pancreas ribonuclease A, horse skeletal muscle myoglobin, bovine pancreas insulin, horse heart cytochrome c, egg white lysozyme) in 5% D2O solutions (pD 7.0) were measured as a function of pressure up to 1470 MPa at 30 °C. According to the second-derivative spectral changes in the observed amide I band of the proteins, which indicate that the α-helix and β-sheet substructures of the secondary structures break dramatically into the random coil conformation, ribonuclease A and myoglobin are denatured reversibly at 850 MPa and 350 MPa, respectively. Lysozyme denatures partially and reversibly at 670 MPa, as shown by decrease in the α-helix and β-turn substructures, but no change occurs in the random coil and β-sheet substructures. The secondary structure of cytochrome c is not disrupted at pressures up to 1470 MPa, and partial transformation of the α-helix of insulin to random coil starts at 960 MPa. Hydrogen-deuterium exchange of protons on the amide groups in the protein interior is increased by external pressure and is associated with the pressure-induced protein conformational changes. A number of studies on the effects of pressure on protein denaturation have been carried out using various high-pressure detection methods: ultraviolet absorbance spectroscopy (Brandts et al., 1970; Hawley, 1971), visible absorbance spectroscopy (Zipp & Kauzmann, 1973), fluorescence intensity spectroscopy (Li et al., 1976), polarization fluorescence spectroscopy (Chryssomallis et al., 1981), and enzyme activity assays (Taniguchi & Suzuki, 1983; Makimoto et al., 1989). These techniques have the great advantage of being applicable to pressure-induced reversible denaturation of proteins to identify the thermodynamic parameters, especially the volume change and compressibility of a protein in solution, because the experiments can be run under dilute conditions at a protein concentration of less than 0.05% w/v. Therefore, these data reflect the intramolecular phenomena of reversible pressure changes and provide the volume changes accompanying the denaturation of proteins, which are due to the difference in partial molal (specific) volume between the native and denatured proteins in solution.

Use of Partial Molar Volumes of Model Compounds in the Interpretation of High-Pressure Effects on Proteins

The transfer of liquid hydrocarbons into water is accompanied by a large decrease in volume at 25 °C and atmospheric pressure, with typical values for ΔV°tr of — 2.0 ml mol methylene−1. Considering the large amount of apolar surface that is exposed when a globular protein unfolds, the hydrocarbon transfer results imply that the change in volume accompanying the unfolding process (ΔV°obs) should be highly negative under these conditions. However, experimental data on the pressure denaturation of proteins typically yield relatively small values of ΔV°obs at atmospheric pressure and 25 °C. We analyze this apparent inconsistency in terms of a simple thermodynamic dissection of the partial molar volume. This approach allows the volume effects that result from solute-solvent interactions to be determined from experimental partial molar volumes. The use of absolute quantities (partial molar volumes) circumvents assumptions associated with the use of results from transfer experiments. An important finding is that hydration of apolar species is less dense than bulk water. This discovery leads to the conclusion that the contribution to ΔV°obs for protein unfolding from the hydration of apolar surfaces is highly positive, contrary to predictions based on transfer data. Further, hydration of polar surfaces makes a positive contribution to ΔV°obs. The large, positive term from the differential hydration of the folded and unfolded states is compensated by the difference in free volume of the protein in the two states. This finding provides a new framework for interpreting pressure effects on macromolecules. The full characterization of a macromolecular system requires knowledge of the effect of pressure on the system. The thermodynamic information obtained from using pressure as a perturbation is a volume change for the particular reaction being studied. The observed volume change, ΔV°obs, for protein unfolding may provide insight into the mechanisms that determine the three-dimensional structure of the folded state. Pressure denaturation experiments have been demonstrated for a number of proteins, including ribonuclease A (Gill & Glogovsky, 1965; Brandts et al., 1970), chymotrypsinogen (Hawley, 1971), metmyoglobin (Zipp & Kauzmann, 1973), and, more recently, lysozyme (Samarasinghe et al., 1992) and staphylococcal nuclease (Royer et al., 1993).

Resolution of the Ambiguity of van’t Hoff Plots by the Effect of Pressure on the Equilibrium

The change in the Gibbs free energy function, ΔG, of chemical reaction is determined by the difference between the heats respectively released to and absorbed from the environment, and separation of the enthalpy and entropy changes that these changes represent cannot be achieved without specific hypotheses as to their relations. The determination of the enthalpy of reaction by the plot of ΔG/T against 1/T (van’t Hoff plot) implicitly assumes that the enthalpy ΔH and entropy ΔS are temperature independent, and this assumption leads to very large errors when this is not the case and ΔH « TΔS. It is therefore inapplicable to the reactions of molecules, such as proteins, that have thermally activated local motions. The concepts offered previously by the author to relate the entropy and enthalpy changes in protein associations are reviewed briefly and applied to account for the temperature dependence of ΔH and ΔS. It is shown that two different values of the enthalpy computed in that manner correspond to each value of the apparent van’t Hoff enthalpy, but that the choice between the two is easily made by reference to the volume change on reaction. The enthalpies of association of subunit pairs of seven oligomers are all found to be positive and much more uniformly related to the size of the intersubunit surface than those previously assigned by use of the classical van’t Hoff plot.

The Effects of Increased Viscosity on the Function of Integral Membrane Proteins

For many years the idea that the activity of integral membrane proteins is regulated by the fluidity of the lipid matrix was popular and appeared to be quite rational. However, as information about the effect of viscosity on the function of different membrane proteins became available, the correlation between the two became increasingly unclear. The purpose of this article is to readdress this issue in light of our recent pressure and temperature studies. This chapter is divided into seven parts: (1) the effect of viscosity on enzyme activity; (2) the effect of viscosity on the local motions of proteins; (3) characterization of membrane viscosity; (4) demonstration of changes in protein-lipid contacts brought about by changes in viscosity; (5) an example of a protein in which the viscosity appears to stabilize a particular conformational state: (6) relations between membrane viscosity and protein function; and (7) conclusions. The effect of viscosity (η) on the rate (k) of a chemical reaction was first given by Kramers (1940): . . . k=A/ηe−Ea/RT (1) . . . In this expression, viscosity will affect the rate of a reaction by limiting the rate of diffusion of reactants. Viscosity will thus modify the frequency factor (A) and should not affect the activation energy. This expression has been applied to aqueous soluble enzymes (for example, Gavish, 1979; Gavish & Werber, 1979; Somogyi et al., 1984), and it appears that, in general, enzymes obey Kramers’s relation, although in some cases the exponent of η is less than one. Viscosity can affect enzymatic rates not only by limiting the diffusion of substrates but also by damping internal motions of the protein chains. It seems reasonable that a high enough viscosities, the protein would be damped sufficiently so that large activation energies will be required for the backbone motions that allow substrates and products to diffuse into and out of the active site. This viscosity-induced increase in activation energy was shown by studies of the reassociation of carbon monoxide and dioxygen to the heme site of myoglobin after flash photodissociation (Austin et al., 1975; Beece et al., 1980).