|Home | About | Journals | Submit | Contact Us | Français|
Single-particle neutron spectroscopy has contributed important experimental data on molecular dynamics in biological systems. The technique provides information on atomic and molecular motions in macromolecules on the picosecond to the nanosecond time scale, which are essential to biological function. Here, we report on recent neutron measurements performed directly in living cells by using isotope labelling to explore the dynamics of specific cellular components. The paper proposes an integrated view of results on atomic-scale cell water dynamics, internal and global macromolecular motions and solvent isotope effect on macromolecular dynamics. The work established the specific usefulness of the neutron scattering technique to get insight into biologically relevant dynamical features, in particular through comparative measurements. The method developed can now be applied to look for dynamical signatures related to cell characteristics in many different cell types and organelles.
The cell is the fundamental unit of life. It is a highly evolved system with a level of complexity that drives many open questions. Obtaining in vivo information at the molecular scale is challenging. One of the essential properties of the cell is its dynamic nature. At every length scale of matter, the cell is in constant motion, and the different length and time scales are closely connected. For example, cell motility and cell cycle are possible through molecular interactions in the dynamic cytoskeleton network. Actin polymerization is used to produce protrusive and contractile arrays that cooperate to drive cell motility (Small & Resch 2005). In interphase cells, actin arrays are also involved in pre-mitotic nuclear movement (Kennard & Cleary 1997) and nuclear movement following division (Kost & Chua 2002). More generally, biological macromolecules are animated by a large variety of motions on a wide range of time scales, going from the femtosecond to the second, and arising on various length scales. Proteins experience different conformations separated from each other by small structural changes on fast time scales. Atomic and molecular motions on the picosecond to the nanosecond (ps–ns) time scale lubricate larger conformational changes on slower, millisecond, time scales, which are necessary for important biological processes, including enzyme catalysis and intermolecular recognition (Jimenez et al. 2004; Tousignant & Pelletier 2004). Depending on the conformation, proteins will be functional or not.
Water makes up approximately 70 per cent of the cell mass. Water surrounds biological macromolecules, allows them to be stable and active, to interact with each other and to move inside cell cytoplasm and in membranes through lipid headgroup- and protein-hydration interactions. It also constitutes an essential dynamic component of the cell interior, and controversies about its dynamical state have animated the biological community for decades. The perception obtained from indirect experimental observations, which has influenced current thought in significant parts of the biology community, is that the cell interior is a gel or colloidal-like structure in which confined water is interfacial with properties that are significantly different from those of pure water. A 2006 book edited by Pollack et al. states in the preface ‘ … practically all cell water is interfacial’ (Pollack et al. 2006). The claim, however, remains controversial as reviewed by Ball (2008). The apparent diffusion coefficient of water in biological tissues and in the brain has been measured by nuclear magnetic resonance (NMR). The measurements revealed that cell water diffusion, on a micrometric scale, is reduced by a factor of between 2 and 10 compared with pure water. The reduction could be explained partly by tortuosity effects, macromolecular crowding and confinement effects (reviewed by Le Bihan 2007). In contrast, very little is known about cell water dynamics on submicrometric and subnanosecond scales. To this day, only a few techniques can be applied, mainly NMR spectroscopy and neutron scattering. These methods constitute the key experimental techniques that are capable of providing a detailed view of cell water dynamics.
Experimental molecular dynamics studies in biological systems are attracting growing interest. Single-particle neutron spectroscopy has revealed itself as an extremely rich technique for dynamics measurements on living cells. The method provides a unique combination of atomic length-scale resolution and subnanosecond time-scale resolution. Thermal neutrons provide wavelengths of a few angströms that correspond to interatomic distances, and energies of a few millielectron volts that correspond to thermal fluctuations in biological samples. In these conditions and because of the absence of charge of the neutron, the neutron–matter interaction is non-destructive. Single-particle neutron spectroscopy provides a direct measurement of atomic and molecular motions on the ps–ns time scale.
The scattering signal is dominated by the scattering of hydrogen nuclei, which have an incoherent cross section one order of magnitude larger than the cross section of any other nucleus or isotope, including deuterium. The particularity is very useful to probe the dynamics of biological macromolecules as well as the motions of the surrounding water. The dynamics of a specific part of a complex system can also be enhanced by taking advantage of selective isotopic hydrogen–deuterium (H–D) labelling. Isotope labelling provides a key advantage of the neutron scattering method. A deuteration facility is provided at the Institut Laue-Langevin (ILL) to the users to help them with the preparation of specifically labelled samples, from entire cells to purified components such as proteins or membranes (see http://www.ill.eu/sites/deuteration/index.htm for further information).
Single-particle neutron spectroscopy relies mainly on two types of techniques: elastic (EINS) and quasi-elastic (QENS) incoherent neutron scattering. In both experiments, the interference observed is that between an individual nucleus and itself at different positions as it moves in time. The information obtained is that of individual dynamics averaged over all nuclei in the sample, each contributing according to its scattering power (H nuclei will dominate). For this reason and to avoid confusion with incoherence in X-ray or electron diffraction, in the case of neutrons, we shall refer to the process as single-particle spectroscopy.
EINS is measured as a function of the scattering vector, Q. EINS experiments provide information on atomic mean square displacements (MSDs), u2, in a sample as a function of temperature, T (Zaccai 2000). The time scale examined depends on the energy resolution of the spectrometer, while the scattering vector, Q, range, in which the scattered intensity is measured, defines the length scale. A structural resilience, expressed as a mean effective force constant (in newtons per metre), k′, can be calculated from the inverse of the slope of u2 versus T (Zaccai 2000; Bicout & Zaccai 2001).
QENS is measured as a function of both energy transfer and scattering vector. QENS provides information on diffusive motions in biological systems (Bée 1988). On the energy scale, the QENS appears as a broad peak centred on zero energy transfer. In the case of simple exponential diffusion processes, the QENS spectra can be fitted mathematically by Lorentzian functions. The dependence of the Lorentzian half-width at half-maximum (HWHM) on Q contains information on the diffusion coefficient and the characteristic time of the motion.
With approximately 70 per cent of the total cell mass, water constitutes the malleable matrix in which macromolecules move and interact. Understanding the dynamic state of water in the cell interior is a major scientific challenge. Recently, water diffusive motions have been explored inside several types of cells, using neutron scattering and isotope labelling (Tehei et al. 2007; Jasnin et al. 2008a; Stadler et al. 2008). The study on E. coli shed light on the translational and rotational diffusive motions of water in the bacterial cytoplasm (Jasnin et al. 2008a). The experiments were performed on an atomic length scale, and on the picosecond to a few tens of picoseconds time scale. The translational diffusion parameters were extracted from the HWHM of the translational Lorentzian, ΓT, as a function of the scattering vector squared, Q2, at two temperatures (figure 1a). ΓT was best fitted using a jump-diffusion model, which describes diffusion between sites for the water protons with a mean residence time at each site (Bée 1988). Translational diffusion coefficients were found to be very close to those of pure water at corresponding temperatures, with residence times about twice longer (Teixeira et al. 1985; Bellissent-Funel et al. 1995). The higher residence times may reflect the longer times spent by the protons in the first hydration shell. The rotational parameters were extracted from the HWHM of the broad Lorentzian, ΓR + ΓT, as a function of Q2, at two temperatures (figure 1b). The rotational correlation times, τcor,R=1/ΓR, were close to the values extracted for the buffer under the same conditions and of the same order as the values measured for pure water (Sposito 1981; Teixeira et al. 1985). We concluded, therefore, that proton exchange between water layers takes place with the diffusion rates of pure-like water. The results established firmly that E. coli water diffusion was neither confined nor significantly slowed down compared with pure water. Cell water diffusion beyond the first hydration shell is similar to that of pure water at physiological temperature. The same conclusion has been reached using NMR spectroscopy by the group of Halle (Persson & Halle 2008). They have examined the rotational spin relaxation rate of E. coli water over a wide time range from the millisecond to the picosecond time scale. They have found that approximately 85 per cent of E. coli water presented rotational relaxation times similar to those of pure water. The studies provided new pieces of data that enlightened cell water dynamics at the molecular scale.
Translational and rotational diffusive motions of E. coli water. (a) HWHM of the translational Lorentzian, ΓT, as a function of the scattering vector squared, Q2, at 281 (squares) and 301 K (triangles). ΓT was best fitted (solid lines) ...
Similar studies were conducted on Haloarcula marismortui, an extreme halophilic organism from the Dead Sea, by using neutron scattering (Tehei et al. 2007) and NMR (Persson & Halle 2008), respectively. The studies revealed interesting differences between the translational and rotational motions of H. marismortui water. The neutron work showed that a significant part of the cell water presented a translational diffusion slowed down by two orders of magnitude compared with pure water (Tehei et al. 2007). The NMR work established that the rotational spin relaxation rate of H. marismortui water was very close to that found in pure water (Persson & Halle 2008). Both results are not in contradiction and could be partly explained by the high proportion of carboxyl groups along the halophilic structures and the high amount of intracellular K+ bound within them, which participate in very strong protein–salt ion–water interactions along halophilic protein surfaces (Madern et al. 2000). The difference observed between the two types of diffusive motions enhanced the importance of combining different experimental methods to draw a complete view of dynamics in complex systems such as entire cells.
QENS has been used to study internal molecular motions in hydrated protein powders (Doster et al. 1989; Andreani et al. 1995; Zanotti et al. 1997; Fitter 1999; Pérez et al. 1999; Dellerue et al. 2001; Paciaroni et al. 2003; Roh et al. 2006) and in the integral membrane protein bacteriorhodopsin in purple membrane stacks (Fitter et al. 1996a,b, 1997, 1998, 1999). Recently, we presented a dynamical mapping of macromolecular dynamics in vivo, on the full ps–ns time domain accessible through neutron scattering (Jasnin et al. 2008b). The study was performed on E. coli cells close to physiological conditions of temperature. The analysis allowed the separation of global and internal dynamical processes, by using a combination of different neutron spectrometers (figure 2). The apparent global macromolecular diffusion in the E. coli cytoplasm was extracted from the HWHM of the associated Lorentzian, Γsd, at two temperatures (figure 2a). We found a translational diffusion coefficient, D = (1.06 ± 0.11)×10−7 cm2 s−1 at 303 K, which was consistent with the value of 1.2 × 10−7 cm2 s−1 found for haemoglobin in red blood cells at 310 K (Doster & Longeville 2007). Physiological internal motions were found to be a combination of localized reorientational and diffusive motions, occurring from a few picoseconds to approximately 100 ps. The motions observed in a few picoseconds corresponded to restricted jump-diffusion motions, as revealed by the profile of the HWHM of the associated Lorentzian, Γf, as a function of Q2 (figure 2b). They can arise, for example, from the contribution of small groups such as methyl groups, as well as from hydrogen bond breaking and formation in macromolecular structures. Localized reorientational motions emerged on longer, a few tens of picoseconds, time scales (see figure 2c,d and its caption for the details of the analysis). They arise from all protons participating in the rotational motions of small groups (such as protons in methyl groups), in stochastic reorientations of large molecular subunits (such as polypeptide side chains, fatty acid chains or other molecular subunits) as well as in librations of buried groups, relative displacements of globular domains, sugar conformational changes or RNA global bending (McCammon & Harvey 1987).
Global and internal macromolecular motions in E. coli. The data were collected using three neutron spectrometers at the ILL and ISIS neutron sources. (a) Global motions. HWHM of the Lorentzian associated with the apparent self-diffusion of macromolecules, ...
The comparison with previous in vitro work revealed differences between physiological internal motions and the motions extracted in hydrated powders and in solution (Zanotti et al. 1997; Fitter et al. 1998, 1999; Pérez et al. 1999; Appavou et al. 2006; Tehei et al. 2006). On the picosecond time domain, both internal molecular flexibility and diffusion rates are increased in the cell interior compared with the parameters extracted from fully hydrated powders (Zanotti et al. 1997; Fitter et al. 1998, 1999; Pérez et al. 1999). The result showed that the large amount of cell water contributes to picosecond internal macromolecular motions. It suggested that the cell water hydrogen-bond network influences the hydrogen-bond dynamics in macromolecular structures and contributes to physiological structural flexibility on the picosecond time scale. In contrast, picosecond internal flexibility measured in E. coli was found reduced when compared with that measured in solution (Pérez et al. 1999; Appavou et al. 2006; Tehei et al. 2006). The result indicated that weak forces owing to the vicinity of macromolecules may attenuate the lubricating effect of cell water. The exploration of in vivo dynamics appears all the more crucial because the work revealed that standard sample preparations, such as powders and solutions, do not represent accurately the physiological environment, whose complexity participates in the functional dynamics necessary for biological activity.
Solvent interactions are essential for macromolecular stability through hydration, hydrogen bonds, van der Waals interactions, ion binding and the hydrophobic effect. Macromolecular folding is the result of a delicate balance between hydration and intramolecular interactions. Heavy water (D2O) is often used in biophysical studies to replace natural abundance water (H2O) to prevent signal contamination by water H nuclei or to study exchange or other isotope effects. The differences in properties between H2O and D2O molecules, however, are expected to influence macromolecular stabilization and dynamics. The few studies carried out to date in solution have reported that dynamics–stability relationships depend on the protein type. Halophilic proteins were shown to have higher effective resilience in D2O (Tehei et al. 2001), which can be explained by the particularly strong interactions with hydrated salt ions. In contrast, in the case of bovine serum albumin (BSA), heavy water was found to reduce structural resilience (Tehei et al. 2001), which suggested a bigger role of entropy in D2O. Tehei and co-workers have also shown that stability was increased for the two types of proteins in D2O, revealing that higher stability was not necessarily associated with higher resilience.
We explored the solvent isotope effect on average macromolecular dynamics in the E. coli interior (Jasnin et al. 2008c). Measurements were performed on living E. coli containing H2O and D2O, respectively, close to physiological conditions of temperature. Both flexibility, expressed as u2 values, and structural resilience, expressed as k′ values, were found to be smaller in D2O than in H2O (figure 3). A difference was expected because the driving forces behind macromolecular stabilization and dynamics are different in H2O and D2O. In D2O, the hydrophobic effect is known to be stronger than in H2O: it favours the burial of non-polar surface groups as well as their van der Waals packing in the macromolecule core. This may lead to the observed smaller u2 values. The smaller resilience value suggested a larger entropy content in the D2O case owing to increased sampling of macromolecular conformational substates. In contrast, in H2O, stronger resilience associated with larger fluctuation amplitudes suggested a bigger role of hydration bond interaction in dynamics. The increased flexibility measured in the H2O case could, however, arise partly from the contribution of exchangeable hydrogens exposed to the solvent, as well as from the differences in viscosity between H2O and D2O solvent. The study confirmed the significant role played by the solvent in macromolecular dynamics and provided further evidence of the importance to keep it in mind when replacing H2O by D2O for experimental studies.
Solvent isotope effect on macromolecular dynamics in E. coli. MSD, u2, as a function of temperature, T, from E. coli containing H2O (triangles) and E. coli containing D2O (squares). Macromolecular flexibility, expressed as ...
The work on E. coli water dynamics revealed that the water between the macromolecules in the intracellular environment has essentially bulk-like properties on the atomic length scale (Jasnin et al. 2008a). It was concluded that liquid water flows freely inside cells, within the crowded macromolecular environment. It is very likely that the remarkable properties of bulk water are essential for cellular viability, in particular for functional macromolecular folding, stabilization and activity, transport, membrane formation and protein insertion into membranes. The results demonstrated that intracellular water can participate as such in all the interactions described above. Furthermore, as revealed by the comparison with powder samples, cell water was shown to contribute to internal macromolecular motions on the picosecond time scale (Jasnin et al. 2008b). This could be in accordance with the cell water hydrogen-bond network influencing hydrogen-bond dynamics in macromolecular structures. The comparison with dilute protein solutions suggested that weak forces owing to macromolecular crowding may attenuate the lubricating effect of cell water. The results from the solvent isotope effect on macromolecular dynamics in E. coli support the presence of water with bulk-like properties in the cell (Jasnin et al. 2008c). Replacing H2O by D2O has similar effects on cell macromolecules, as has been observed in the dilute solution of BSA (Tehei et al. 2001). The lower apparent resilience indicates that the increased stability in D2O is entropy driven, and results from the increased sampling of conformational substates.
It must be pointed out, however, that the results on protein–water interactions and dynamics strongly depend on the type of organism and its environment. This has been shown by the previous work on halophilic proteins (Tehei et al. 2001) and the halophilic organism H. marismortui (Tehei et al. 2007). In these systems, protein and water dynamics are strongly affected by the molecular interactions with the surrounding salt ions. The highest stability of halophilic proteins in D2O is enthalpy driven by ion hydration and results in a higher apparent resilience (Tehei et al. 2001). A slow water component has been discovered in H. marismortui, which is not present in E. coli, probably resulting from similar protein–salt ion interactions (Tehei et al. 2007).
We concluded that intracellular molecular dynamics is the result of a delicate balance between the structural and dynamical properties of bulk water and macromolecular hydration and interactions.
The studies on extremophilic bacteria and E. coli revealed that it was possible to get insight into molecular dynamical features in complex systems such as entire cells. Escherichia coli has the advantage of being easy to prepare, can be cultivated in fully or partially deuterated medium and can be obtained in the large quantities required for the neutron scattering measurements. Escherichia coli was therefore an ideal system for establishing the feasibility of in vivo neutron studies. Nevertheless, the work on bacteria constitutes the first step in the in vivo neutron studies. It is now essential to go further by exploring different cells and organelles, including eukaryotes and cellular tissue. The measurements will permit exploring protein dynamics in the presence of cytoskeleton or nucleus or as a function of the cellular state in different organelles. The feasibility of such measurements relies on the appropriate choice of the cell sample. The cells should be easy to prepare in bulk quantities, considering the few hundreds of milligrams of quantities required for the neutron measurements.
The author wishes to thank Dr Giuseppe Zaccai for critical reading of the manuscript and helpful comments.
One contribution of 13 to a Theme Supplement ‘Biological physics at large facilities’.