Introduction
Understanding cellular adaptation to extreme environments remains a challenge for biology (Berezovsky and Nussinov, Reference Berezovsky and Nussinov2026). Salt has been used as a preservative for millennia, yet it is well known that certain moulds and microbes thrive in high salt concentrations and could lead to spoilage (Prasad and Seenayya, Reference Prasad and Seenayya2000). These salt-adapted organisms were not studied effectively until the 1970s, when an interest in life under extreme conditions was propelled by the isolation of microbes from salt lakes, acidic volcanic springs at high temperatures, and deep ocean hydrothermal vents at high pressures and temperatures. In parallel, the period saw the development of structural biology, a multidisciplinary field connecting physicists, chemists, and molecular biologists. State-of-the-art instrumentation was developed for electron microscopy, NMR, and X-ray and neutron diffraction. Methods, propelled by the phenomenal increase in computing power, were emerging for the collection and analysis of large amounts of data. Advances in technology were decisive. In the early days of X-ray macromolecular crystallography, Bragg spots were detected on film (Perutz et al., Reference Perutz1960); the development of a position-sensitive electronic detector for X-ray cameras by André Gabriel, based on the particle detectors of nuclear physics, revolutionised the field (Gabriel and Dupont, Reference Gabriel and Dupont1972). High-resolution position-sensitive detectors for electrons by Wasi Faruqi were similarly fundamental in the development of cryo-electron microscopy for the high-resolution structure determination of biomolecules (Faruqi and Henderson, Reference Faruqi and Henderson2007). Synchrotron radiation was applied to macromolecular crystallography; high-flux neutron sources were opening up to scientists for investigations of biological structures and dynamics. In a personal scientific memoir, Klug (Reference Klug2010) described the combination of X-ray diffraction and crystallographic (Fourier) electron microscopy to solve the structures of Tobacco Mosaic Virus and chromatin to a resolution of a few Ångstrøm. The approach was developed by Henderson and Unwin (Reference Henderson and Unwin1975) to solve the first structure of a transmembrane protein. This was followed by striking developments in electron cryo-microscopy (Subramaniam and Henderson, Reference Subramaniam and Henderson2000; McMullan et al., Reference Mcmullan2023; Patwardhan et al., Reference Patwardhan2025), which led to a steady rise in the number of structures solved, overtaking X-ray crystallography in 2025. Small-angle X-ray scattering (SAXS) and small-angle neutron scattering (SANS) deserve a special mention because they permitted structural studies in solution (Svergun et al., Reference Svergun2013). This was especially relevant for the study of biomolecules under extreme solvent conditions. Novel approaches in neutron spectroscopy provided data on macromolecular dynamics in solution (Grimaldo et al., Reference Grimaldo2019). Driven by discoveries in the molecular biology of extremophiles, biophysical methods continue to provide valuable data not only on adaptation to extreme conditions but also and especially on a variety of fundamental biomolecular systems relevant for human health, such as membrane receptors, for example.
In this review, I have put an emphasis on a selection of results and on the wider significance of biophysical studies on extremophiles, rather than on the detailed description of the structural biology methods, which continue to be in rapid expansion (Grimes et al., Reference Grimes2018; Metwalli et al., Reference Metwalli2020; Afonine et al., Reference Afonine2025; Da Silva et al., Reference Da Silva2026).
Biological background
Understanding how living organisms thrive in hypersaline lakes, or at 80 °C in volcanic hot springs, or even at 115 °C and under high hydrostatic pressure (HHP) near hydrothermal vents, 3,000 m below sea level, poses a major challenge. Already at the macromolecular level, polypeptide chains fold into active protein structures in restricted ranges of aqueous solvent salt composition (Von Hippel and Schleich, Reference von Hippel, Schleich, Timasheff and Fasman1969), temperature (Privalov and Khechinachvili, Reference Privalov and Khechinashvili1974) and pressure (Makhatadze et al., Reference Makhatadze2025). In order to address biomolecular adaptation beyond the mesophile range, the discovery of extremophiles led to original developments in biophysical methods for the characterisation of structures and interactions under extreme solvent concentration, temperature, and pressure. While these have mostly been studied with approaches focusing on specific proteins, biophysical approaches have also confirmed the link between environmental and protein adaptation at the more complex and diverse scale of the proteome. For example, Carré et al. (Reference Carré2021) developed nano-differential-scanning fluorimetry (nanoDSF) for the study of proteome stability and solubility in cell lysates of the halophilic archaeon Haloarcula marismortui proteome.
In the 1930s, Benjamin Wilkanski collected samples of seawater at various depths in the Dead Sea, a few kilometres from the mouth of the Jordan River. Following previous accounts of unicellular life with remarkably high salt tolerance, he planned to test the samples for microbial growth. His results, published in a report to Nature (Wilkanski, Reference Wilkanski1936), distinguished organisms in the cultures that were later shown to belong to algae of genus Dunaliella, and to halobacteria. Wilkanski changed his name to El-Azari Volcani and continued to publish on the microbiology of the Dead Sea (reviewed by Oren, Reference Oren1993). Mevarech et al. (Reference Mevarech1977) isolated a malate dehydrogenase from H. marismortui (Hm MalDH), which became the first halophilic protein to be studied extensively by biophysical methods (Zaccai, Reference Zaccai2013). Hm MalDH was found to display a surface rich in acidic amino acid residues and to bind significant amounts of salt and water. Recently, Ortega-Quintanilla and Millet (Reference Ortega-Quintanilla and Millet2025) discussed how halophilic organisms modify their proteome to withstand salt concentrations that inactivate non-adapted mesophilic proteins, and described the mechanistic link by which the halophilic amino acid composition optimises function in hypersaline environments, balancing the trade-off between stability, solubility, and catalytic function.
The spectacular colours of salt ponds in San Francisco Bay are also due to algae of genus Dunaliella and microorganisms of genus Halobacterium. Studies of bacteriorhodopsin, a light-activated proton pump in the purple membrane of Halobacterium halobium (later renamed Halobacterium salinarum) transformed membrane protein biophysics with about 4,500 publications in PubMed since its first description in 1973 by Dieter Oesterhelt and Walther Stoeckenius (Reference Oesterhelt and Stoeckenius1971, Reference Oesterhelt and Stoeckenius1973). When H. halobium was exposed to osmotic shock through step-wise lowering of the salt concentration, the cell wall broke up first, followed by the cell membrane; in distilled water, the sedimented residue contained intact purple sheets (Stoeckenius and Rowen, Reference Stoeckenius and Rowen1967). The colour was found to be due to a retinal-binding protein (the first discovered that was not associated with vision), named bacteriorhodopsin (BR) (Oesterhelt and Stoeckenius, Reference Oesterhelt and Stoeckenius1971); Oesterhelt and Stoeckenius (Reference Oesterhelt and Stoeckenius1973) went on to characterise its function as a light-activated proton pump. The exceptional organisation of BR on a two-dimensional crystalline lattice in PM led to the determination of the first structure of a membrane protein (Henderson and Unwin, Reference Henderson and Unwin1975). BR was also amongst the first membrane proteins to be sequenced (by pre-genomic methods) in a head-to-head race between the Massachusetts Institute of Technology (Khorana et al., Reference Khorana1979) and the Russian Academy of Sciences (Ovchinnikov et al., Reference Ovchinnikov1979).
Microbes of the genera Thermoplasma and Sulfolobus, which thrive in hot, low-pH environments such as volcanic springs in Yellowstone, Iceland, or Naples Solfatara, were isolated in the early 1970s by Langworthy et al. (Reference Langworthy1972). These authors found that Thermoplasma acidophilum lipids were different from other bacterial lipids; they had almost no fatty acid ester groups but apparently contained long chains of alkyl groups, ether-linked to glycerol: an observation that would later contribute to the classification of the organism amongst the archaebacteria (Woese et al., Reference Woese1978). Brock et al. (Reference Brock1972) characterised Sulfolobus by electron microscopy and found an unusual cell wall structure devoid of peptidoglycan. The biochemical and morphological characterisations established that Sulpholobus had no close relationship with previously described bacteria. Deep-sea hydrothermal vents were discovered in the late 1970s at depths 2,000–3,000 m (reviewed by Colín-García et al., Reference Colín-García2016). The most extreme hyperthermophiles live on the walls of such hydrothermal vents, at temperatures ≥90 °C. The current record growth temperature of 122 °C is held by Methanopyrus kandleri. The history of hyperthermophiles and their relevance for the origin of life are discussed by Stetter (Reference Stetter2013).
The proposal of a three-domain (Archaea, Bacteria, Eukarya, respectively) phylogenetic tree was based on molecular structure and sequence rather than cellular phenotype similarities (Woese et al., Reference Woese1990). The extreme halophilic and extreme thermophilic archaebacteria were classified within the archaeal domain based on the composition of their lipids (Langworthy et al., Reference Langworthy1972) and 16S RNA sequences (Magrum et al., Reference Magrum1978). Norse gods invaded the archaeal domain when high-throughput sequencing techniques revealed new lineages assembled from sediments in the vicinity of Loki’s Castle hydrothermal field on the edge of the North Mid-Atlantic Ridge, North-East of Iceland (Spang et al., Reference Spang2015). The analysis of these archaea revealed the presence of a relatively high number of genes with homologues exclusively in the Eukarya. Several new groups were subsequently described, forming a clade: Lokiarchaeota, Heimdallarchaeota, Thorarchaeota and Odinarchaeota, collectively named Asgard archaea, in reference to the Norse gods’ pantheon (Zaremba-Niedzwiedzka et al., Reference Zaremba-Niedzwiedzka2017). The discovery of Asgard archaea has greatly reshaped our understanding of the origin of eukaryotes. Asgards are currently considered the closest prokaryotic relatives of eukaryotes and represent the archaeal host lineage that participated in the endosymbiotic event leading to the first eukaryotic cell (López-García and Moreira, Reference López-García and Moreira2023) (Figure 1). In a recent review, Schleper et al. (Reference Schleper and Rodrigues-Oliveira2026) examine the consequences of the Asgard discovery in the broader context of archaeal molecular and cellular biology.
Phylogenetic tree of life. LUCA is the hypothesised Last Universal Common Ancestor. (From Wikimedia. Author: Crion. Used under Commons Licence.)
Figure 1. Long description
At the bottom center is the root labeled L U C A. From this point, three main branches radiate outward into colored sectors representing the domains of life.
1. Bacteria (Left): This large sector is divided into three sub-groups. The bottom light blue section includes C P R bacteria and Cyanobacteria. The middle purple section contains Firmicutes, Chloroflexi, and Actinobacteria. The top pink section, labeled Gram negative, includes Aquificae, Thermotogae, Fibrobacteres, and the F C B group (Chlorobi, Bacteroidetes). It also contains the P V C group (Planctomycetes, Verrucomicrobia, Chlamydiae), Spirochaetes, and Proteobacteria.
2. Archaea (Top Center): A light blue sector containing Nanoarchaeota (D P A N N), Euryarchaeota, Thaumarchaeota, Crenarchaeota, and Lokiarchaeota (A S G A R D).
3. Eukaryota (Right): This sector is divided into three main groups. The top purple section contains Excavata. The middle green section, labeled Diaphoretickes, includes Archaeplastida (Glaucophyta, Chloroplastida, Rhodophyceae), Hacrobia, and the S A R group (Stramenopiles, Alveolata, Rhizaria). The bottom purple section, labeled Amorphea, includes Amoebozoa and Opisthokonta (Fungi and Animalia).
Water, water, everywhere
Liquid water is vital for the biochemistry of life (Ball, Reference Ball2010). Its existence and dynamical properties constitute a necessary condition for organisms to thrive in any environment. The thermodynamics of the effect of the aqueous solvent environment on macromolecule stability is determined by the Gibbs free energy change,
$ \Delta \mathrm{G} $
,
where T is temperature,
$ \Delta \mathrm{H} $
is enthalpy change, and
$ \Delta \mathrm{S} $
is entropy change. A reaction will proceed in the direction of a decrease in free energy (negative
$ \Delta \mathrm{G} $
). Qualitatively, the enthalpy change will be negative if bonds are formed during the reaction; an increase in accessible configurations (disorder) will lead to an increase in entropy. In the case of protein unfolding in the absence of solvent, for example, bonds will be broken (positive enthalpy change) and chain disorder will increase (positive entropy change). The folded-to-unfolded transition will take place at the temperature where
$ \Delta \mathrm{H}=T\Delta \mathrm{S} $
. In the presence of solvent (the biological situation), the thermodynamics applies to the total, protein plus solvent, system. Dynamic hydrogen-bond networks in liquid water provide a remarkable entropy sink, which favours folding of polypeptide chains and formation of membranes. As a solvent, water also constitutes the support to specify the acidity or basicity (pH) of a solution, as well as to dissociate salt molecules into ions.
Hydration water
Hydration water plays a significant role around lipid headgroups and on the surface of biological membranes, especially with respect to membrane protein dynamics (Zaccai and Gilmore, Reference Zaccai and Gilmore1979; Zaccai, Reference Zaccai1987). The hydration dependence of BR dynamics has been studied intensively and is discussed in Section ‘Molecular dynamics of BR’.
Small-angle X-ray or neutron scattering (SAXS or SANS) have the advantage of being able to measure macromolecular structures in solution. Contributions to the scattering of different parts of a particle depend on contrast, the difference in electron density (for SAXS) or neutron coherent scattering length density (for SANS) between the part and solvent. The usual application of contrast variation is in SANS by using H2O/D2O exchange; the neutron scattering density of H2O is −0.561 × 1010 cm−2, that of D2O is 5.755 × 1010 cm−2, that of protein is between 2 and 3 × 1010 cm−2, making it practically invisible in ~40% D2O, its contrast match point. The contrast match of nucleic acids is ~65% D2O, that of lipids ~12% D2O (Svergun et al., Reference Svergun2013). By combining SAXS and SANS on various soluble proteins, Svergun et al. (Reference Svergun1998) established the existence of a first hydration shell with an average density ~10% larger than that of bulk solvent. Comparisons with other studies suggested that this may be a general property of aqueous interfaces. The molecular dynamics of the hydration shell and bulk water were reviewed by Bellissent-Funel (Reference Bellissent-Funel2001). Wang (Reference Wang2026) reviewed the behaviour of water confined in low-dimensional materials. Such water displays structural and dynamical properties that diverge fundamentally from bulk liquid water. The hydrogen bond network, in particular, is reshaped inducing molecular ordering, which alters dielectric, vibrational, and transport properties. The review by Wang (Reference Wang2026) presents a comparative synthesis of confined water, from single-molecule encapsulation in molecular cages and single-file flow in carbon nanotubes to layered phases trapped between atomically flat van der Waals crystals. By unifying results across structural, spectroscopic, and transport studies, it highlights its implications for nano-fluidics, energy storage, and biological hydration.
Does the crowded cytoplasm modify the properties of water beyond the immediate hydration shells of macromolecules? It is an important question because most biochemical studies were performed in fractionated systems in dilute solution, an environment different from the crowded cellular interior. Also, dynamics studies mentioned in the previous paragraph were performed in vitro, in hydrated powders or solution samples. In order to understand biophysical adaptation to extreme environments, it was of significant interest to measure protein dynamics in vivo (review by Zaccai, Reference Zaccai2020).
Water in cells
The cytoplasm of all cells contains at least 70% water, which constitutes the intracellular matrix in which biological molecules interact. Understanding its dynamic state is a major scientific challenge. Jasnin et al. (Reference Jasnin2008) used neutron spectroscopy to measure water dynamics, in vivo, in the cytoplasm of Escherichia coli. In such experiments, the main signal comes from the neutron incoherent scattering cross-section of H (80.3 barns); D, with a much lower neutron incoherent cross-section (2 barns), is practically invisible. In order to focus on water motions, the samples used by Jasnin et al. (Reference Jasnin2008) consisted of centrifuged pellets of cells from deuterated cultures, suspended in H2O. Experimental timescales covered motions on the picosecond time scale (corresponding to bulk water motions) to the nanosecond time scale corresponding to interfacial (hydration shell) water. Escherichia coli cells contain approximately 2.6 g of water per gram of macromolecule, and the first hydration layer accounts for about 0.5 g of water per gram of macromolecule. On average, cell water would correspond to about five layers of water molecules around macromolecular surfaces. The neutron data were fitted by a jump diffusion model, which yielded a residence time on a site and translational diffusion coefficient between sites. The residence time was found to be higher by a factor of 2 than that of pure water. This average might reflect the longer time spent by the water molecules in the hydration shells. Once released, however, exchange with external layers would take place with a diffusion rate similar to that of pure water. In contrast to a widespread opinion held at the time that water is ‘tamed’ by macromolecular confinement, the neutron measurements established that water diffusion beyond hydration layers within E. coli cells is similar to that of pure water at physiological temperature. Jasnin et al. (Reference Jasnin2008) emphasised the fundamental biological relevance of their results. The remarkable dynamics of bulk water and its properties as an entropy sink are essential for various aspects of cell viability, such as macromolecular folding, stabilisation and activity, transport, membrane formation and protein insertion into membranes. Whenever hydration shells have been detected it has always been after the macromolecules were fully synthesised, correctly folded, stabilised, and the bulk water subsequently dried or frozen out. The data show that the water between macromolecules in vivo in E. coli has properties that are essentially the same as those of bulk water and can therefore participate as such in the interactions mentioned previously.
In a neutron spectroscopy study on planarians and other living animal cells, Mamontov (Reference Mamontov2018) found that microscopic diffusivity within the cells was well-defined and could be separated into a fast component corresponding to the sum of hydration, cytoplasmic and extra-cellular water and a slower component corresponding to cell constituents.
Water diffusion was also measured in cells of an extreme halophile. Studies on the distribution of cell electrolytes during growth of a halobacterium culture had suggested extensive potassium ion-water binding within the cytoplasm (Ginzburg et al., Reference Ginzburg1970). Tehei et al. (Reference Tehei2007) characterised intracellular water dynamics in H. marismortui, by neutron spectrometry. Similar to the E. coli experiments of Jasnin et al. (Reference Jasnin2008), samples consisted of centrifuged pellets of halobacterial cells from fully deuterated cultures, suspended in H2O. Water diffusion at 285 K was measured on two neutron spectrometers, on the 10 ps and 1 ns time scales, respectively. From the 10 ps data, a translational diffusion constant of 1.3 × 105 cm2 s−1 was determined at 285 K. This value is close to that found for the bulk water fraction in E. coli, as well as that of the water in the 3.5 M NaCl solution bathing the cells (Frank Gabel, private communication). A surprisingly large fraction of slowly diffusing water was reported from the 10 ns data. At 285 K the water H nuclei of this component displayed a residence time of 411 ps (compared with a few picoseconds in bulk or hydration water). At 300 K, the residence time dropped to 243 ps and was associated with a translational diffusion of 9.3 × 10−8 cm2 s−1, or 250 times lower than that of bulk water. It was hypothesised that the electrolyte experiments of Ginzburg and Ginzburg (Reference Ginzburg, Ginzburg, Eisenberg, Katchalsi-Katzir and Manson1975) as well as the slow mobility of a large part of H. marismortui cell water indicated the existence of specific hydrated potassium ion binding structures, perhaps similar to those in the potassium channel (MacKinnon, Reference Mackinnon2003). Persson and Halle (Reference Persson and Halle2008) applied water 2H (D) and 17O spin relaxation NMR and molecular dynamics simulations to study water diffusion in living cells of E. coli and H. marismortui. Their data were similar for the bacterium and halophile, and they concluded that there was no ‘slow’ water in H. marismortui. NMR measured rotational diffusion of the water molecule whereas neutron spectroscopy measured linear diffusion and residence times (Tehei et al., Reference Tehei2007; Jasnin et al., Reference Jasnin2010). Persson and Halle (Reference Persson and Halle2008) argued that rotational and linear diffusion would be coupled; ‘slow’ water would also display slower rotational diffusion. As shown by the recent discovery of a plastic ice phase (Rescigno et al., Reference Rescigno2025) (see Section ‘Superheated water, supercritical water and high-pressure ice phases’), however, it is possible to maintain rotational dynamics even when water molecules are fixed in a crystalline ice structure (Figure 2). This suggests that the NMR data are not in contradiction with the neutron spectrometry discovery of ‘slow’ water in the extreme halophile.
Molecular dynamics snapshot showing the rotational disorder around oxygen atoms of water molecules in the ice lattice of the plastic ice phase. Adapted from Rescigno et al. (Reference Rescigno2025).
Figure 2. Long description
A three-dimensional molecular model displays a dense field of water molecules. Each molecule consists of a central red sphere representing an oxygen atom bonded to two smaller white cylinders representing hydrogen atoms. While the oxygen atoms are positioned in a roughly ordered hexagonal lattice, the hydrogen atoms are oriented in various directions, illustrating rotational disorder. In the bottom-left corner, a small three-axis coordinate system serves as a spatial anchor with a green arrow pointing up, a red arrow pointing right, and a blue circle indicating the z-axis. Some molecules show faint, semi-transparent ghosting effects behind them, suggesting movement or multiple superimposed rotational states within the simulation snapshot.
Significant studies of water diffusion were also performed on human tissue. Modern medical diagnostic imaging techniques rely on water diffusion to investigate the architecture of brain tissue Le Bihan, Reference Le Bihan2007; Gkintoni et al., Reference Gkintoni2026). Intrinsic tissue heterogeneity causes systematic deviations from bulk water diffusion behaviour. A number of theoretical and empirical approaches have been proposed to explain the non-Gaussian profile of the process. The aim of the work by Natali et al. (Reference Natali2019a) was to shed light on the physics of water diffusion in brain tissue from the micron scale (from NMR) to the atomic scale (from neutron spectroscopy). The coexistence of free-like and confined water populations in brain tissue extracted from a bovine right hemisphere was revealed at both scales. The results are relevant for improving the modelling of the physics driving intra- and extracellular water diffusion in brain tissue. In another study, Natali et al. (Reference Natali2019b) combined diffusion NMR and neutron spectroscopy to compare water dynamics between the two hemispheres of bovine brain, again on the micron scale of diffusion within the tissue to the atomic scale sensitive to local dynamics within and around cells. They interpreted the significant differences in water dynamics in terms of hemisphere-dependent cellular composition as well as specificity of neurological function such as preferential networking.
By using neutron spectroscopy, Marques et al. (Reference Marques2020) probed dynamical activity in healthy and various cancerous human cells. In addition to higher plasticity in the cancer cells, they found the dynamics of hydration water molecules to remain practically unaffected, while the rotational motions of cytoplasmic water undergo significant changes upon the normal-to-cancer transition.
Superheated water, supercritical water and high-pressure ice phases
Water boils at 100 °C under 1 bar (100 kPa) pressure. Increasing pressure results in superheated water with a higher boiling point. At 10 bar (1 MPa), for example, superheated water remains liquid to 175 °C. The ambient water temperature at great depths on the sea floor is ~2 °C. However, water emerging from hydrothermal vents can reach temperatures >400 °C (Colín-García et al., Reference Colín-García2016). At depths ~3,000 m, (hydrostatic pressure ~ 300 bar or 30 MPa), the temperature and pressure values are close to the critical point in the phase diagram of water. Water above the critical point will be in a supercritical fluid state, with physical properties between those of a gas and those of a liquid.
Jupiter’s icy moons, Ganymede and Callisto, and Saturn’s Titan may contain a water fraction up to half their total mass in oceans hidden below an outer layer of ice (Soderlund et al., Reference Soderlund2024). The pressure at the base of their hydrosphere would range from 1 to 2 GPa for the large icy moons, to several tens of GPa for Earth-sized exoplanets. On the phase diagram of water, ICEVII appears to be the relevant phase for the high pressures and temperatures of planetary interiors. Neutron spectroscopy measurements, conducted at temperatures between 450 and 600 K and pressures up to 6 GPa revealed that the crystalline ICEVII phase is plastic, with water molecules exhibiting picosecond rotational dynamics (Rescigno et al., Reference Rescigno2025): a structure which combines the solid properties of ice with rotational dynamics similar to that of liquid water (Figure 2).
Compatible solutes
Biological membranes support considerable electrochemical gradients between the interior and exterior of cells, but are fragile with respect to osmotic stress. Organisms isolated from salt marshes or marine environments compensate for osmolarity fluctuations in the environment by producing low molecular weight osmoprotectants called compatible solutes because they do not interfere with cell biochemistry and metabolism (Empadinhas et al., Reference Empadinhas and da Costa2008). They belong to different chemical families, such as polyols, sugars like trehalose, amino acids, betaines, and ectoines, which maintain osmotic equilibrium in many halotolerant and dry-tolerant bacteria, algae and fungi. Their accumulation in plants increases survival rates during stress due to drought. In tardigrades, brine shrimp, and nematodes, osmoprotectants can allow cells to survive when completely dehydrated. Special water interactions have been assumed for compatible solutes in order to explain both their protective effects on biomolecules and the fact that they do not interfere with cell biochemistry.
Trehalose is a disaccharide present in a wide variety of organisms, including bacteria, yeast, fungi, insects, invertebrates, and lower and higher plants. It has been shown that trehalose can protect proteins and cellular membranes from inactivation or denaturation caused by a variety of stress conditions, including desiccation, dehydration, heat, cold, and oxidation. Neutron spectroscopy experiments on myoglobin dynamics suggested a mechanism for the protective effect of trehalose (Cordone et al., Reference Cordone1999). The hydrated protein displayed harmonic dynamics, typical of the solid state, at temperatures up to about 200 K where a so-called dynamical transition occurs with mean-square displacements rising more rapidly with temperature, typical of a fluid state. It was hypothesised that the large fluctuations were required for activity. At high temperatures they lead to denaturation. The addition of trehalose had a striking effect: the dynamical transition was suppressed, and the solid-like harmonic regime was maintained to high temperatures, locking in the protein dynamics and protecting it from denaturation.
Ectoine is a compatible solute in the cytosol of halotolerant eubacteria of the genus Halomonas, isolated from saline soils, salt lakes or the deep sea. Zhang et al. (Reference Zhang2022) studied the osmoadaptation strategies of Halomonas alkaliphila from Xiaochaidan Salt Lake by using transcriptome and ectoine analysis. Divergent mechanisms were involved in osmoadaptation at different salinities of the microbe. Ectoine synthesis and accumulation were significantly increased under salt stress. Increased transcriptional expression of ectoine synthesis genes may play a key role in high salinity-induced osmoadaptation. At extremely high salinity (18% NaCl), 5-hydroxyectoine and ectoine worked together to maintain cell survival. Another halomonas species, Halomonas titanicae, is destroying the wreck of the Titanic by turning it into rust (Sánchez-Porro et al., Reference Sánchez-Porro2010). Zaccai et al. (Reference Zaccai2016) examined the biophysical effects of ectoine on the hydration shells of maltose-binding protein (MBP) and of the purple membrane (PM) from H. salinarum. SANS and membrane diffraction data provided model-independent quantitative structural evidence for the current hypothesis that ectoine molecules are preferentially excluded from the PM and MBP hydration shells, while neutron spectroscopy established that membrane molecular dynamics were preserved by the presence of ectoine outside the membrane hydration layer. Neutron atomic-level liquids diffraction of molar ectoine solutions revealed modifications in H-bond geometry, which led to a significant weakening of water–water interactions in the presence of the solute. An interpretation is suggested that the enthalpy-driven H-bond structuring effect on water in direct contact with ectoine, shown previously by infrared spectroscopy, and the entropy contributed by the weaker inter-water H-bonding, shown by neutron diffraction, together, thermodynamically favour the high solubility of the ectoine molecule and its preferential exclusion from macromolecular surfaces (Zaccai et al., Reference Zaccai2016).
The dihydrate form of ectoine is not stable under ambient conditions; it loses the bound water molecules to transform into a highly hygroscopic anhydrate form. In order to further examine the behaviour, Hützler et al. (Reference Hützler2020) solved the crystal structures of the dihydrate and anhydrate forms of ectoine by X-ray and neutron diffraction, and quantum mechanical calculations. In the dihydrate form, the carboxylate group in the molecule adopts a less stable conformation compared to that in the anhydrate. The results from crystallography together with the previous H-bonding data from liquid diffraction allowed the authors to propose a pathway for water uptake and loss in the solid state of ectoine.
If salt is a preservative, how do halophiles survive?
Since ancient times, salt has been valued as an essential nutrient that enhances flavour in small quantities and inhibits spoilage in large quantities. It was used for mummification by the Egyptians, and generally to preserve meat, fish and vegetables up to the widespread use of refrigeration. The Madaba map, a 6th-century floor mosaic from the Church of Madaba in Jordan, shows a boat harvesting salt on the Dead Sea at the point of junction with the River Jordan. The pile of salt on the boat is flecked with pink, indicating its contamination by halophilic life forms (Figure 3). At the end of the Middle Ages, salt wars raged in Europe, such as the one between Ferrara and Venice; in the 19th century, the San Elizario Salt War was a range war in the United States. In the 21st century, salt caravans still ply between Timbuktu and salt mines in Mali.
A small part of the Madaba Mosaic, a 6th-century floor map from the church of Madaba in Jordan. The picture shows a boat on the Dead Sea at the point of junction with the River Jordan. The lumps on the boat represent salt obtained by evaporation from Dead Sea water. (Michael Avi-Yonah, Israel Exploration Society, Jerusalem 1954).
Figure 3. Long description
The mosaic uses earth-toned tesserae in shades of tan, brown, and cream. At the North, the River Jordan is depicted as a narrow horizontal stream that widens as it flows West and South into the Dead Sea. The water is represented by undulating horizontal lines of dark and light tiles. Along the Northern bank, there are stylized palm trees and Greek inscriptions. In the center-right, a wooden boat with a curved prow and two oars is positioned on the water. Inside the boat, two large, rounded mounds of white and light-tan tiles represent salt. A human figure is partially visible behind the salt mounds, with one arm extended toward the stern. In the Southeast corner, four lines of modern white sans-serif text are overlaid on the mosaic: Multiple strain microbial biofilms, Agonists of a peptide hormone, Role of intestinal microbes in obesity, and Listeria immune system evasion.
When dissolved in water, salt molecules dissociate into ions, which play vital roles in physiology. Sodium is the main ion found in extracellular fluid, universally, while potassium is the main ion inside cells. Regulated ion transport between the two sides of the cell membrane through protein ion pumps and channels leads to osmo-electric potential gradients that contribute to bioenergetics and are critical for nerve and muscle function, as well as to control body hydration and blood pH. Ionic concentrations are kept under tight control by various mechanisms of significance in disease (Zhou and Trebak, Reference Zhou and Trebak2024). Ionic disruptions lead to medical emergencies through dehydration, overhydration, or cardiac and neurological complications. Because of their medical relevance and important industrial applications, such as in battery design, the study of ionic solutions is an important field of physical chemistry. In mesophiles, high salt concentrations strongly inhibit biochemical interactions such as protein folding and stabilisation or protein-nucleic acid contacts. In contrast, the proteins of extreme halophiles are strongly dependent on high salt concentrations for their stability. In a recent review, Hu and Vila Verde (Reference Hu and Vila Verde2026) critically examine experimental and computational studies investigating the relationship between the halophilic proteome and thermodynamic stability of halophilic proteins as a function of salt concentration.
Halophilic malate dehydrogenase (Hm MalDH)
Hm MalDH in solution
Malate dehydrogenase from H. marismortui (Hm MalDH) was purified and first characterised by Mevarech et al. (Reference Mevarech1977) and Mevarech and Neumann (Reference Mevarech and Neumann1977) by applying an adapted procedure because the protein denatured below 2 M NaCl. Cendrin et al. (Reference Cendrin1993) cloned, sequenced and expressed, in E. coli, the gene coding for Hm MalDH. Large amounts of a soluble but inactive form of the protein were produced. Activation of the enzyme was achieved by increasing the salt concentration to 3 M NaCl. The recombinant protein was purified and shown to be indistinguishable from the native enzyme isolated from H. marismortui. The work represents the first example of a successful expression of a halobacterial gene coding for a soluble protein in E. coli and its recovery as a functional enzyme. Bonnete et al. (Reference Bonnete1993) combined analytical ultracentrifugation (AUC), SAXS and SANS contrast variation to propose a structural model for Hm MalDH in solution, in which protein-water-salt binding appears as an important stabilisation feature. An analysis of the combined data provided quantitative thermodynamic and structural parameters. In multimolar NaCl or KCl, Hm MalDH forms a tetramer that binds ~0.1 g water and ~0.4 g salt per g protein. Bovine serum albumin (BSA) was studied in parallel by the same methods, as a model for non-halophilic proteins. It was found to bind comparable amounts of water but an order of magnitude less salt. In contrast to an earlier assumption that the protein in a halophilic environment would possess a hydration shell to exclude salt, it essentially recruits salt ions, apparently required for its stabilisation, and it dissociates and unfolds when NaCl or KCl concentration falls below ~2 M. Hm MalDH is active, soluble and mildly unstable in an unusually wide range of salt conditions and temperatures, making it a particularly interesting model for the study of solvent effects on protein stability. The Hofmeister series is a classification of salt ions in order of their ability to precipitate proteins. Bonnete et al. (Reference Bonnete1994) studied its loss-of-activity kinetics due to tetramer dissociation and concomitant unfolding in various salts of the Hofmeister series, in H2O and D2O. In all cases, stability increased when D2O replaced H2O and with increasing salt concentration. Each salt condition, however, was associated with a particular thermodynamic regime depending on the position of the ions on the Hofmeister series and specific binding to the protein. Bonnete et al. (Reference Bonnete1994) concluded that stabilisation mechanisms in Hm MalDH adapted to solvent conditions. Ebel et al. (Reference Ebel2002) and Ebel et al. (Reference Ebel and Zaccai2004), by exploring the effects of crowding in extremophiles, established a linkage between solvation, on the one hand, and weak protein–protein interactions, stability, and dynamics, on the other.
Hm MalDH crystal structures
SAXS, SANS and AUC are solution methods that provided important results on the thermodynamics and low-resolution structure of Hm MalDH and its interactions in different salt solvents. Macromolecular crystallography, though limited to solvent environments compatible with crystallisation, significantly provides atomic-resolution views of interactions when these are ordered in the crystal. For example, whereas solution methods accurately weigh the total amount of hydration water in Hm MalDH, the crystal structure will only show the smaller number of water molecules that are ordered in the crystal. Diffracting crystals of Hm MalDH were obtained by using MPD (2-methyl-2,4-pentanediol), and its structure was determined by X-ray crystallography (Dym et al., Reference Dym1995). Compared to that of non-halophilic malate dehydrogenases, the structure of Hm MalDH revealed a significant excess of acidic over basic residues on the enzyme surface (Figure 4) and the presence of more salt bridges. Crystallographic studies on selected mutants were subsequently performed in order to explore the role of protein-solvent interactions in the stability of the enzyme. Richard et al. (Reference Richard2000) highlighted novel protein-solvent features in the structure of the E267R stability mutant of apo (-NADH) Hm MalDH at 2.6 Å resolution compared to that of the apo-wild type at 2.9 Å resolution. The tetramer appears to be stabilised by ordered water molecule networks and inter-subunit complex salt bridges ‘locked’ in by bound solvent chloride and sodium ions. The E267R mutation points into a central ordered water cavity, disrupting protein-solvent interactions. Analysis of the crystal structures confirmed that halophilic adaptation exploited the high ionic concentration of the environment for stabilisation. Irimia et al. (Reference Irimia2003) solved the structure of the (R207S, R292S) mutant of Hm MalDH at 1.95 Å resolution to examine the role of salt bridges and solvent ions in quaternary structure stabilisation (Figure 5). The mutations disrupt two inter-dimeric salt bridge clusters, essential for wild-type tetramer stabilisation. Previous solution experiments on the double mutant in solution had shown the tetrameric structure in 4 M NaCl to dissociate into active dimers when the salt concentration dropped to 2 M NaCl. The crystal structure shows a tetramer that, in the absence of the salt bridge clusters, appears to be stabilised by a network of ordered water molecules and by chloride binding at the dimer–dimer interface. The double mutant and wild-type dimer folds are essentially identical. Chloride ions are also observed at the monomer–monomer interfaces of the mutant, contributing to the stability of each dimer against low salt dissociation. The results further supported the hypothesis that extensive binding of water and salt is an important feature of adaptation to a halophilic environment.
The surface of Hm MalDH showing the excess of acidic (negatively charged) residues in red over basic (positively charged) residues in blue.
Figure 4. Long description
A space-filling molecular model of the H m Mal D H protein tetramer. The structure is roughly rectangular with rounded edges, composed of four interconnected subunits. The surface is primarily light gray, representing neutral residues. Red clusters, representing negatively charged acidic residues, are densely and uniformly distributed across the entire surface. Blue clusters, representing positively charged basic residues, are much fewer in number and appear as isolated patches. The red patches significantly outnumber the blue patches across all four quadrants of the protein surface, illustrating a highly acidic exterior.
Salt bridge clusters and chloride ion binding at the monomer–monomer and dimer–dimer interfaces. (a) Chloride ions are shown as large yellow spheres; water molecules are represented by small red spheres. (b) Zoomed in on a salt bridge cluster in the dimer–dimer interface. Adapted from Irimia et al. (Reference Irimia2003).
Figure 5. Long description
Panel a shows a tetrameric protein structure composed of four subunits labeled A, B, C, and D. Subunits A (top-left) and C (bottom-right) are colored blue, while subunits B (top-right) and D (bottom-left) are colored pink. Large yellow spheres representing chloride ions are positioned at the interfaces between the subunits. Small red dots representing water molecules surround the complex. Panel b is a circular inset showing a zoomed-in view of the salt bridge cluster. It features ball-and-stick models of specific amino acid residues. At the top are Asn 1 8 6 A and Asp 2 0 9 A. To the right is Glu 3 0 1 D. At the bottom is Ser 2 9 2 D and Gly 2 6 6 D. A dashed line indicates a distance of 2.75 Angstroms between the nitrogen atom of Asn 1 8 6 A and the oxygen atom of Gly 2 6 6 D.
Malate dehydrogenase from a halophilic bacterium
In order to investigate if similar stabilisation features were found in non-archaeal halophilic malate dehydrogenases, Madern and Zaccai (Reference Madern and Zaccai2004) purified and characterised the enzyme from the bacterium, Salinibacter ruber (Sr MalDH). The phenotype of this bacterium is remarkably similar to that of the haloarchaea (Mongodin et al., Reference Mongodin2005). The protein was found to be a tetramer similar to Hm MaldH by sedimentation velocity measurements. The amino acid composition, however, does not have the strongly acidic character of Hm MaldH, and the enzyme is stable and active in the absence of salt. Mongodin et al. (Reference Mongodin2005) found, however, that genes related to the transport and metabolism of glycerol and glycine betaine, the most common compatible solutes in moderate halophiles and halotolerant species, were present in the genome of Sr MaldH. It is likely that the mesophile-like surface of Sr MalDH is due to the cytoplasmic osmolyte being a compatible solute, rather than highly concentrated KCl as in archaeal extreme halophiles.
The LDH-MalDH superfamily and ancestral MalDH
The work on halophilic adaptation of Hm MalDH inspired the exploration of relationships between malate and lactate dehydrogenases (Brochier-Armanet and Madern, Reference Brochier-Armanet and Madern2021). l-Malate (MalDH) and l-lactate (LDH) dehydrogenases belong to a large superfamily of NAD-dependent enzymes that has been characterised in organisms belonging to the three Life domains. LDHs are tetramers, whereas MalDHs can be either dimeric or tetrameric. In order to gain insight into molecular relationships between them, Madern et al. (Reference Madern2000) studied folding intermediates of a mutant of Hm MalDH, a protein with LDH-like structure and MalDH enzymatic activity. Hm MalDH is a tetramer, two dimers interacting mainly via complex salt bridge clusters. These salt bridges are disrupted in the mutant. Neutron scattering and AUC showed that while the enzyme remained a tetramer in 4 M NaCl, stable oligomeric intermediates could be trapped under specific conditions of pH, low NaCl, or temperature. The properties of the dimeric intermediate were compared to those of dimeric intermediates of LDH and dimeric MalDHs. A detailed analysis of the putative dimer-dimer contacts in these enzymes provided insight into why some can form tetramers and others cannot. Katava et al. (Reference Katava2020) combined experiments and molecular dynamics simulations to unveil the allosteric propensity of a thermophilic MDH. At its working temperature, MDH is not allosteric; it adopts, however, a compact structure and reorganises the active state towards functional conformations similar to homologous allosteric LDH. A single-point mutation confers on the MDH cooperative behaviour that mimics an allosteric LDH. The study revealed that not only does thermophilic MDH use temperature as an external parameter to regulate its functionality in a similar way that allosteric LDHs use substrate/cofactor binding but also that the MDH scaffold possesses intrinsic allosteric potential. Coquille et al. (Reference Coquille2025) focused on deciphering the evolutionary mechanisms by which allosteric properties were acquired in the LDH/MalDH superfamily. By introducing mutations associated to the emergence of allosteric LDHs into the non-allosteric MalDH from M. kandleri, the enzymatic profile was gradually shifted towards that of typical of allosteric LDHs (see also Pereira et al., Reference Pereira2026, in Section ‘Crystal structures reveal the diversity of adaptation mechanisms’). By adding a final single mutation to fine-tune the flexibility of the catalytic site, Coquille et al. (Reference Coquille2025) obtained an enzyme with both sigmoid (homotropic) and hyperbolic (heterotrophic) substrate activation profiles. The work highlighted the role of the underlying protein dynamics.
Blanquart et al. (Reference Blanquart2021) described the evolutionary history of MalDHs within the halobacteria. They resurrected nine ancestors along the inferred MalDH phylogeny, including the Last Common Ancestral Halobacterial MalDH (LCAHa) and compared their biochemical properties with those of five modern halobacterial MalDHs, and monitored their stability, oligomeric states, and enzymatic properties as a function of salt type and concentration. Diverse evolutionary processes, such as amino acid replacement, gene duplication, loss and replacement of the MalDH gene owing to horizontal transfer, resulted in significant differences in solubility, stability, and catalytic properties between these enzymes in the three Halobacteriales, Haloferacales, Natrialbales orders since the LCAHa common ancestor.
The search for traces of life: the protective effect of salt on biological macromolecules
Tehei et al. (Reference Tehei2002) found that trapping Hm MalDH in ‘dry’ salt crystals protects the enzyme against thermal denaturation. Similar protection was not observed for the homologous mesophilic enzyme. High salt concentrations also protected transfer RNA molecules against thermal degradation, allowing activity to be recovered. The authors discussed the observations in the context of cell-free biological macromolecules in the environment and the search for traces of life in planetary exploration.
Bacteriorhodopsin (BR) in the purple membrane (PM) of H. salinarum
The discovery of PM was a priceless gift to structural biophysicists and spectroscopists. The membrane is stable and easy to obtain in large quantities. The function of BR is associated with a millisecond cycle of colour changes, called the photocycle. The challenge was to relate the photocycle with structural changes associated with proton transfer against the membrane potential. The retinal is covalently attached to the protein via a protonated Schiff base. The chromophore isomerises upon absorption of a photon; the Schiff base then releases a proton to the extracellular medium and is subsequently re-protonated from the cytoplasm. The photocycle proceeds from the dark-adapted state, bR570, through intermediates, K590, L570, M550, N560, O640, where the numbers correspond to the light absorption maxima in nanometres (Figure 6).
On the left-hand side: the structure of BR in PM, showing the alpha helices as purple ribbons, the lipids in ball-and-stick representation and water molecules as blue balls. On the right-hand-side is shown the photocycle with the names, and absorption maxima (in nanometres) of the intermediate states. The proton is bound on the cytoplasmic side of the membrane (CP) and released on the extra-cellular side (EC). Adapted from Wood et al. (Reference Wood2007).
Figure 6. Long description
The left panel displays a three-dimensional molecular model of B R embedded in a purple membrane. Purple ribbon-like alpha helices form the protein core, surrounded by light green ball-and-stick lipid molecules and blue spherical water molecules. A vertical black arrow bisects the structure, pointing from the top label C P down to the bottom label E C, terminating at a proton symbol H super plus. This indicates proton transport from the cytoplasmic to the extracellular side.
The right panel, titled Photocycle, depicts a circular pathway of intermediate states. Starting at the top and moving clockwise:
- B R sub 570 absorbs light h nu to reach state J.
- State J transitions to K via a double-headed arrow.
- State K transitions to L via a double-headed arrow.
- State L leads to M sub 412, where a proton H super plus is released to the E C side.
- M sub 412 transitions to N.
- State N receives a proton H super plus from the C P side to transition to O.
- State O returns to the initial B R sub 570 state.
Structural studies on intact PM
The purple membrane provided a unique opportunity to study the structure, to a few Ångstrøm resolution, of a membrane protein in situ. The organisation of BR on a two-dimensional crystalline lattice led to the determination by electron diffraction of the first and, so far, only structure of a membrane protein, in its natural environment with lipids and water, revealing that it crossed the membrane in seven α-helices (Henderson and Unwin, Reference Henderson and Unwin1975).
Neutron diffraction with specific deuterium labelling permitted to locate the areas of hydration and water molecules in the membrane (Zaccai and Gilmore, Reference Zaccai and Gilmore1979; Zaccai, Reference Zaccai1987; Papadopoulos et al., Reference Papadopoulos1990), the retinal binding site in projection (Jubb et al., Reference Jubb1984), and specific amino acid residues in order to correlate primary and secondary structure of BR within the purple membrane (Engelman and Zaccai, Reference Engelman and Zaccai1980; Popot et al., Reference Popot1989) contributing to essential structural information on membrane protein folding. Conformational changes during the photocycle were determined by trapping intermediates and solving their atomic-resolution structures by electron crystallography (Subramaniam and Henderson, Reference Subramaniam and Henderson2000) and X-ray crystallography (Neutze et al., Reference Neutze2002). Popot and Engelman (Reference Popot and Engelman1990, Reference Popot and Engelman2000) reviewed their work on reconstituted BR from helical fragments in lipids or with amphipols in lipid-free environments in order to establish a model for helical membrane protein folding, stability, and evolution. They conceptualised helical membrane protein folding and oligomerisation as involving two energetically distinct stages – the formation of stable transmembrane helices and their subsequent side-to-side association. Helix interactions with the bilayer, with prosthetic groups, and with each other were examined in the context of available evidence to conclude that the two-stage model was a useful approach for discussions of stability, folding concepts, and as a basis for understanding membrane protein evolution.
BR crystallography
BR was amongst the first targets in the development of lipidic cubic phases to solubilise and crystallise membrane proteins (Landau and Rosenbusch, Reference Landau and Rosenbusch1996; Pebay-Peyroula et al., Reference Pebay-Peyroula1997; Heberle et al., Reference Heberle1998), and amphipols for membrane protein solubilisation, folding, and structural studies by cryo-EM, NMR, or X-ray diffraction (Zoonens and Popot, Reference Zoonens and Popot2014; Popot, Reference Popot2018). Neutze et al. (Reference Neutze2002) and Wickstrand et al. (Reference Wickstrand2015) critically reviewed the light-induced structural modifications found in X-ray crystallographic and electron microscopy studies of BR. Three-dimensional structures of intermediates in the photocycle of BR provided a detailed structural picture of the proton pump in action. The sequence of conformational changes following the photoisomerisation of the all-trans retinal chromophore (covalently bound to Lys216 via a protonated Schiff base) to a 13-cis configuration is described. Initial changes are located near the protein’s active site, where a key water molecule that serves as a keystone for the ground state of BR is disordered. It is important both for stabilising the structure of the extracellular half of the protein and for maintaining the high pKa of the Schiff base (the primary proton donor) and the low pKa of Asp85 (the primary proton acceptor). Coupled structural rearrangements propagate out from the active site towards the extracellular half of the protein, facilitating the mechanism of proton release to the extracellular medium. On the cytoplasmic half of the protein, structural rearrangements define a pathway for the re-protonation of the Schiff base by opening the proton transport channel and enabling a proton to be taken up from the cytoplasm.
Molecular dynamics of BR
Neutron spectroscopy experiments on intact PM initially concentrated on how the global molecular dynamics of BR depended on its water/lipid environment. Following Doster et al. (Reference Doster1989), who revealed a dynamical transition in myoglobin, a soluble protein in a hydrated powder sample, by neutron scattering, Ferrand et al. (Reference Ferrand1993) found a similar transition in BR, a membrane protein in its natural environment. In hydrated PM, a dynamical transition was observed from a harmonic to a softer, anharmonic fluctuation regime at about 200 K. In the dry membrane, however, the transition was abolished, and the fluctuations remained harmonic above 300 K, indicating the necessity of hydration for soft dynamics (Zaccai, Reference Zaccai2000). Zaccai (Reference Zaccai1987) had characterised the structure and hydration of PM under different conditions by neutron diffraction and H2O/D2O exchange. The unit cell dimension of the BR lattice was found to decrease by the same amount (2%) upon drying the membranes at room temperature or when they were cooled to liquid nitrogen temperatures. The reasons for the similar contraction are different, however. In the dry membranes at room temperature, the lipid headgroups are dehydrated, and the decrease in dimension is due to the lipids packing closer to each other. In the wet membrane at liquid nitrogen temperature, the water of hydration around lipid headgroups remains in place, and the decrease in dimension upon cooling is due to thermal contraction only. The data suggested that functional BR, in the wet state at room temperature, has a relatively soft environment that would allow large amplitude motions of the protein; in the inactive, dry membranes at room temperature, large amplitude protein motions would be inhibited by the close-packed environment, similarly to the contracted membrane at liquid nitrogen temperatures. Lehnert et al. (Reference Lehnert1998) further analysed PM molecular dynamics at various hydration levels and discovered two main motion populations that responded differently to water binding. Striking correlations appeared between the nanosecond to picosecond thermal motions and the slower kinetic constants (in the millisecond time range) of relaxations and conformational changes occurring during the photocycle.
The dynamics of selected labelled amino acids within BR was measured by Réat et al. (Reference Réat1998). Fully deuterated PM was prepared biosynthetically in H. salinarum with BR containing H-retinal, H-tryptophan, and H-methionine. The labelled amino acids mainly cluster around the retinal in the functional core of the protein. Above the dynamical transition, the labelled groups appeared to be significantly more resilient than the rest of the protein. The results quantified the dynamical heterogeneity of BR with a more rigid core region, which would control the stereo-specific selection of retinal conformations. This was later confirmed by the crystallographic work on intermediates (see Section ‘BR crystallography’).
Wood et al. (Reference Wood2007, Reference Wood2008a) studied the coupling of protein and hydration water dynamics in purple membrane and in maltose-binding protein, respectively, by a combination of neutron spectroscopy and molecular dynamics simulations. Wood et al. (Reference Wood2008b) examined the hydration dependence of active core fluctuations within bacteriorhodopsin by using H/D specific labelling. Two motion populations were characterised. The lower amplitude motions were found to be the same for both the labelled active core of BR and the global membrane. The larger amplitude dynamics of the labelled part, however, were found to be more resilient than the average membrane, which again suggested their functional importance in maintaining a less fluctuating environment around the retinal to stabilise its conformational changes. The response to hydration was characterised, showing that the labelled part is not shielded from hydration effects. The results suggest that the inhibition of high-amplitude motions may play a key role in the slowing down of the photocycle and the proton pumping activity of BR at low hydration. Wood et al. (Reference Wood2008c) combined neutron scattering and molecular dynamics simulations on three PM samples with specifically labelled leucine, isoleucine and tyrosine residues, respectively. Motions of the labelled residues on the pico- to nanosecond time scale were examined as a function of temperature from 20 to 300 K. The three labelled amino acids displayed different temperature dependence: isoleucine residues have larger displacements compared to the global PM above 120 K; leucine residues have displacements similar to that of PM in the entire temperature range studied; and tyrosine residues have displacements smaller than that of the average membrane in an intermediate temperature range. Experimental features were mostly well reproduced by molecular dynamics simulations performed at five temperatures, which allowed the dynamical characterisation of the amino acids under study as a function of local environment. The resulting dynamical map of bacteriorhodopsin revealed that movements of a specific residue are determined by both its environment and its residue type.
The hydration and dynamics of a triple mutant of BR (D96G/F171C/F219L) in PM were investigated by neutron scattering coupled with H2O/D2O exchange (Lehnert et al., Reference Lehnert2005). The mutant, which is active in proton transport, has an ‘open’ ground-state structure similar to that of the M intermediate in the photocycle (Figure 6) of the wild type. The experiments demonstrated an increased proton channel hydration in the mutant PM compared with wild-type PM, in both high (86%) and low (57%) relative humidity. The authors suggested that this is due to the smaller side chains of the mutant residues liberating space for more water molecules in the proton channel; the extra water would then be able to participate in the proton translocation network. The global dynamical behaviour of the mutant BR on the 100 ps time scale, as a function of relative humidity, was found to be identical to that of the wild type, showing that the ‘open’ BR structure and additional water molecules in the proton channel still maintain the resilient active core of the protein.
The molecular dynamics of BR is probably the best-studied for a membrane protein proton pump. Bondar and Smith (Reference Bondar and Smith2025) recently reviewed over four decades of research on the light-driven protein, which inspired the development of key experimental and computational methodologies, now widely used in membrane protein studies. Ion translocation across membranes against electrochemical gradients is vital for bioenergetics. In the case of BR, the proton pump harnesses an energetically favourable reaction, the absorption of a photon by the retinal, to pump protons against the membrane potential. The overall directionality of the pump had to be considered in order to decipher the reaction mechanism by using molecular dynamics computations, with general implications for active-state geometries in cellular ion pumps. Conflict between computational approaches and experimental results applied to the BR proton transfer steps in the early photocycle intermediates was resolved by detailed quantum mechanical/molecular mechanical computation. Key to understanding how the pump achieves directional transfer of protons, the proton transfer and reorganisation of the internal hydrogen-bond network needed to be considered within the context of the energy landscape of the complete reaction cycle.
BR-inspired biology and potential medical applications
The structure of BR in seven transmembrane helices turned out to be shared by the G protein-coupled receptors (GPCRS), the largest superfamily of cell surface membrane receptors, encoded by approximately 1,000 genes. GPCRs mediate vital biological functions of signal transduction triggered by various extracellular stimulants such as photons, ions, lipids, neurotransmitters, hormones, peptides, and odorants (Zhang et al., Reference Zhang2024).
BR was the first retinal-binding protein discovered that was not associated with vision in animals. Rhodopsins, originally named as the pigments of vision, now encompass a variety of photoactive retinal-binding proteins in animals and microbes. They include cation and anion passive or active transporters, sensor and signal-transducing enzymes, light-gated channels and light-driven pumps that are the main tools in optogenetics (Kandori, Reference Kandori2019). Optogenetics is a biological technique used to characterise and manipulate the activity of neurons or other cell types with light. This is achieved by expression of rhodopsin-gated ion channels, pumps or enzymes in the target cells (Chen et al., Reference Chen2026).The 2026 Nobel Prize in Physiology and Medicine was awarded to KARL DEISSEROTH (The Howard Hughes Medical Institute and Stanford University, USA), PETER HEGEMANN (Humboldt University of Berlin, Germany) and GEORG NAGEL (University of Würzburg, Germany) “for their discoveries concerning light-gated ion channels and optogenetics”.
Following intensive studies of BR and the superfamily of retinal-binding membrane proteins (review by Oesterhelt, Reference Oesterhelt1998), the Oesterhelt laboratory explored phototaxis (directed motion along a gradient towards more favourable conditions, in response to a light signal) in H. salinarum (Sundberg et al., Reference Sundberg1990), with the systems biology ultimate aim to describe all the steps between an outside stimulus, protein expression and the response network. In the extreme halophile, phototaxis is driven by a rotating flagellar motor. The clockwise or anticlockwise direction of the rotation is switched on depending on if the colour of the light is attractive or repellent. Streif et al. (Reference Streif2009) published a quantitative analysis of signal transduction in motile and phototactic cells by computerised light stimulation and model-based tracking. Schlesner et al. (Reference Schlesner2012) identified the genes and studied the protein interactions in the H. salinarum taxis signalling network. The interaction analysis revealed that the core signalling proteins are involved in different protein complexes with dynamic interchanges between them. Altegoer et al. (Reference Altegoer2022) published structural insights into the mechanism of archaeal rotational switching. Dieter Oesterhelt passed away in November 2022. He had participated in the first characterisation of PM and BR and continued his innovative contributions to membrane biology to the very end of his life.
Low pH insertion peptides (pHLIP) form a family that harnesses the high acidity at the surface of cells within the tumour microenvironment (TME) as a biomarker target (Reshetnyak et al., Reference Reshetnyak2024). They can be employed to deliver imaging or therapeutic cargo to the TME or into cancer cells. PhLIPs, not being directed to specific receptors, simultaneously target stromal, cancer and some immune cells, with a significant synergistic effect. They can also be placed on the surface of nanoparticles for targeted intracellular payloads. The first member of the pHLIP family was discovered during fundamental studies of membrane protein folding on BR. Seven polypeptides corresponding to the BR alpha helices were synthesised, and each was studied in phospholipid vesicles in order to test if it would be individually stable in a membrane. A peptide containing just the third helix of BR spontaneously distributed, in a pH-dependent way, between three conformations: a water-soluble conformation, a peripherally bound non-helical conformation, and an alpha-helical one. The latter inserted rapidly and reversibly across the membrane, with a pK of 6.0 (reviewed by Reshetnyak et al. (Reference Reshetnyak2024)).
Extremes of pressure and temperature
Bacteria and archaea inhabiting the deep-sea vent ecosystem experience HHP as well as extreme temperatures, pH, and salinity (Jebbar et al., Reference Jebbar2015). In contrast to mesophilic proteins, in which folding thermodynamics and stability have been extensively studied by calorimetry (Privalov and Khechinachvili, Reference Privalov and Khechinashvili1974), few hyperthermostable proteins have been thoroughly characterised in terms of their in vitro folding processes and their thermodynamic stability profiles. Luke et al. (Reference Luke2007) discussed examples and concluded that despite the small set of studied systems, it had become clear that superslow protein unfolding was a dominant strategy to allow these proteins to function at extreme temperatures. Ramakrishna Reddy et al. (Reference Ramakrishna Reddy2026) systematically investigated the molecular basis of thermal adaptation by constructing independent datasets of four mesophilic–thermophilic protein pairs, defined by sequence identity and optimal growth temperature. The hierarchical rule-based method identified stabilising residues as the primary distinguishing factor, followed by electrostatic energy, volume, and localised electrical effects. HHP is known to impact the structure of several cellular components and functions, such as membrane fluidity, protein activity and structure. Thermodynamically, the impact of pressure resembles a lowering of temperature, since it reinforces the structure of proteins. Makhatadze et al. (Reference Makhatadze2025) characterised the thermodynamic effects of pressure and temperature on alpha helices. HHP favours the lower volume state. Pressure effects on protein stability depend on the volume changes between the native and denatured states. The volume of the denatured state can depend on pH and ionic composition of the solvent, temperature, as well as on pressure. The authors used circular dichroism spectroscopy on a series of model peptides to focus on the pressure and temperature dependence of the helix-coil transition. They found that low temperatures and low pressure stabilise the helical structure, suggesting that the volume change of the helix-coil transition is positive. Above >40 °C, however, volume changes become negative, and pressure destabilises the helical structure. In psychrophilic bacterial strains adapted to the bottom of the ocean far from hydrothermal vents, the stabilising effect of pressure at low temperature correlates with the increase in the fraction of high-helix-forming amino acid residues at the expense of low-helix-forming residues (Makhatadze et al., Reference Makhatadze2025).
The impact of HHP as an alternative to thermal treatment on proteins is a major area of study in food science (review by (Rufino Vieira et al., Reference Rufino Vieira2024).
Crystal structures reveal the diversity of adaptation mechanisms
From the mid-1990s to 2025, a few hundred structures with the keyword hyperthermophile were deposited in the Protein Data Bank. The structural basis of protein thermostability has always been of scientific and biotechnological interest. One approach has been to compare differences in the X-ray crystal structures of orthologous proteins from hyperthermophilic and mesophilic organisms in order to examine stabilising interactions. Hennig et al. (Reference Hennig1997) solved the structure of dimeric phosphoribosyl anthranilate isomerase (PRAI) from the hyperthermophile Thermotoga maritima (tPRAI) to 2.0 Å resolution. The comparison with the known 2.0 Å structure of PRAI from E. coli showed an increased number of salt bridges and significant differences at the level of the fold. In tPRAI, symmetry-related loops protrude reciprocally into cavities of the other subunit to provide for multiple hydrophobic interactions stabilising the dimer.
Coquelle et al. (Reference Coquelle2007) made a comparative biochemical and structural analysis of LDHs from Champsocephalus gunnari (an Antarctic fish), Deinococcus radiodurans (a mesophilic bacterium) and Thermus thermophilus (a thermophilic bacterium). The respective temperature adaptation from hot to cold environments resulted from a decrease in activation enthalpy and an increase in activation entropy. In the crystal structures of the LDHs, pairwise comparisons between thermophilic versus mesophilic and mesophilic versus psychrophilic (cool-loving) revealed that temperature adaptation is due to a few amino acid substitutions localised in critical regions. These substitutions play a role in conformational stability, in local flexibility, or in both. Going from hot- to cool-adapted enzymes, the substitutions decreased the number of ion pairs, reduced the size of ionic networks, and created unfavourable interactions involving charged residues, inducing strong local disorder. The analysis shed light on how evolutionary processes shift the subtle balance between overall stability and structural flexibility.
The crystal structure of the malate dehydrogenase-NAD complex from the hyperthermophilic archaeon Archeoglobus fulgidus (Af MalDH) was solved at 2.9 Å resolution by Irimia et al. (Reference Irimia2004). Unexpectedly, the protein was shown to be dimeric. All other known members of the MalDH family are homo-tetramers like Hm MalDH. The association of the dimers in Af MalDH to form a tetramer is prevented by several deletions at the level of two essential loops within [LDH-like] MalDH enzymes. In order to gain insight into high-temperature adaptation, the structure was compared to Hm MalDH as well as to several thermostable and hyperthermostable homologues. Af MalDH displays a reduction of the solvent-exposed surface, an optimised compact hydrophobic core, a high number of hydrogen bonds, and includes a large number of ion pairs at the protein surface. These features occur concomitantly with a reduced number of residues in the protein subunit due to several deletions in loop regions. The loops are further stiffened by ion pair links with secondary structure elements.
The nitrogenase reductase NifH catalyses ATP-dependent electron delivery to the Mo-nitrogenase, a reaction central to biological N2 fixation. NifHs have been studied extensively in bacteria but structural information about their archaeal counterparts was limited. Maslac et al. (Reference Maslac2024) characterised NifHs from three Methanococcales, providing X-ray crystal structures from meso-, thermo-, and hyperthermophilic methanogens: Methanococcus maripaludis (37 °C), Methanothermococcus thermolithotrophicus (65 °C), and Methanocaldococcus infernus (85 °C). The native structures from M. thermolithotrophicus exhibit high flexibility. In contrast, the complexes of NifH with MgADP obtained from the three methanogens are superposable, more rigid, and present remarkable structural conservation with their homologues. In comparison to the NifH from the phylogenetically distant Methanosarcina acetivorans, these reductases do not cross-react significantly with Mo-nitrogenase from A. vinelandii. However, they associate with bacterial nitrogenase when a ligand is added to mimic a transient reactive state. Detailed surface analyses suggested that subtle substitutions would affect optimal binding during the catalytic cycle between the NifH from Methanococcales and the bacterial nitrogenase, implying differences in the N2 -machinery from these ancient archaea.
Grundman et al. (Reference Grundman2025) studied the structure and stability under extreme conditions of a truncated haemoglobin (TrHb) from the psychro-piezophile Shewanella bentica by electronic absorption spectroscopy, X-ray crystallography, NMR spectroscopy and SAXS. The protein self-associates weakly in solution; it contains an extensive network of hydrophobic tunnels connecting the active site to the surface; amino acid replacements at the dimeric interface confirmed this region to be the site of intermolecular interactions. HHP dissociated the quaternary structure while the protomers resisted unfolding and haem loss. Preservation of structural integrity under HHP was also observed in non-piezophilic TrHbs, which suggests that this ancient property is derived from functional requirements. The study broadened the perception of the TrHb lineage and the resistance of globins to extreme environmental conditions.
Understanding the evolution of proteins to adapt their stability and function to changing temperatures remains a central question in molecular biology (Akanuma, Reference Akanuma2026). Pereira et al. (Reference Pereira2026) investigated molecular mechanisms of thermal adaptation in MalDHs by solving the structures of a reconstituted ancestral MalDH from Methanococcales archaea and present-day MalDHs from M. infernus (M. inf) and Methanocaldococcus jannaschii (M. jan). The comparison highlighted the role of salt bridges in thermal adaptation. Furthermore, M. inf and M. jan, MalDHs show structural features that resemble the extended or compact states typical of allosteric lactate dehydrogenases. To test whether or not there is a link between thermal adaptation and the emergence of allosteric regulation, Pereira et al. characterised two M. jan MalDH mesophilic-like mutants, structurally and by molecular dynamics simulations. The results indicated an equilibrium between non-competent and competent catalytic site configurations, depending on temperature. At low temperature the wt M. jan MalDH selects non-competent conformers, whereas high temperature favours active conformers. In contrast, the M. jan MalDH mutants explore competent conformers for catalysis at a lower temperature, a phenomenon that fits well with their biochemical behaviour. The results suggest that thermal adaptation and the evolution of allostery are strongly linked via modulation of the protein conformational landscape.
Proteases and proteasomes
Proteases and proteasomes are key players of regulated protein degradation in all domains of Life. Research on oligomeric protease complexes derived from hyperthermophilic archaea has significantly advanced our understanding of how specific self-association of proteins is one of the ways to optimise and regulate biological activity.
The TET peptidases form unique tetrahedral structures of large self-compartmentalised dodecameric complexes. These metallopeptidases, members of the M42 family, are widely distributed in prokaryotes. TETs may participate in protein homeostasis and amino acid recycling by processing peptides downstream of the proteasome and other related proteolytic complexes. Three different versions of TET complexes, with different substrate specificities, were found to coexist in the cytosol of the hyperthermophilic archaeon Pyrococcus horikoshii (PhTET1, PhTET2, PhTET3). Appolaire et al. (Reference Appolaire2014) applied an elegant approach based on SANS with deuterium labelling and contrast variation to elucidate the quaternary structure organisation and assembly pathways of the 468 kDa hetero-oligomeric, symmetric PhTET2-PhTET3 aminopeptidase complexes. The results revealed that the topology of the PhTET2 and PhTET3 dimeric building blocks within the complexes is not casual but rather suggests that the quaternary arrangement optimises catalytic efficiency towards peptide substrates. Basbous et al. (Reference Basbous2018) identified a recombinant TET complex in P. horikoshii (PhTET4), which self-assembled into a tetrahedral edifice similar to other TET complexes. It is a strict hyperthermophilic, glycyl aminopeptidase that, unlike other peptidases from the same family, is activated only by nickel ions. The hyperthermophilic autotrophic archaeon Methanocaldococcus jannaschii possesses a single gene encoding an M42 peptidase. MjTET is the first of the family to exhibit broad amino acid specificity, including activity on aromatic residues. Atalah et al. (Reference Atalah2026) compared the catalytic constants of MjTET and those of its close analogues in P. horikoshii. The specialised P. horikoshii TETs displayed higher catalytic efficiencies than the generalist MjTET. The structure of MjTET was resolved to 3 Å by cryo-EM and compared with available models of the four P. horikoshii TETs to reveal features underlying substrate specificity. The structural analysis, combined with mutagenesis studies, revealed a previously uncharacterised loop in the catalytic domain that contributes to substrate discrimination. Chagny et al. (Reference Chagny2025) established sequence criteria for high-throughput identification of TET peptidases and revealed their wide taxonomic distribution in archaea. In addition to the five TETs that had been identified in Thermococcales, they functionally characterised six additional TETs, including four from Asgardarchaeota. The results fell into two main functional classes: generalist TETs with broad-spectrum activities and specialised TETs with selective activities. Combining functional and phylogenetic analyses, a classification of TET peptidases into 11 groups was proposed. Notably, one generalist group was identified as ancestral in archaea from which specialised TETs arose following gene duplication or horizontal transfer.
AAA+ ATPases are ubiquitous hexameric unfoldases that, in complex with proteases, form the proteasome protein degradation machinery in archaea and eukarya. Ibrahim et al. (Reference Ibrahim2017) combined SANS, temperature activation and fluorescence spectroscopy to monitor unfolding of a green fluorescent protein substrate by the archaeal AAA+ PAN unfoldase, on the sub-minute time scale. The results support a model in which the large enzymes unfold the substrates in a reversible power stroke mechanism involving several of their subunits. Mahieu et al. (Reference Mahieu2020) applied time-resolved SANS of a deuterium-labelled substrate and an unlabelled archaeal PAN-20S system to obtain direct structural information on substrate states during ATP-driven unfolding and subsequent proteolysis in solution. Krüger et al. (Reference Krüger2023a, Reference Krüger2023b) used NMR to monitor GFP processing, in real time, by the archaeal PAN unfoldase and the PAN–20S degradation system. PAN-dependent unfolding of GFP does not involve the release of partially folded GFP molecules resulting from futile unfolding attempts. In contrast, once stably engaged with PAN, GFP molecules are efficiently transferred to the proteolytic chamber of the 20S subunit. This is essential to guarantee that unfolded but not proteolysed proteins are not released into the cell, where they would form toxic aggregates. The results of Krüger et al. are in good agreement with previous real-time SANS data (Ibrahim et al., Reference Ibrahim2017) with the further advantage of allowing the investigation of substrates and products at amino-acid resolution.
Molecular dynamics adaptation to extreme environments
Neutron spectroscopy studies of extremophilic protein molecular dynamics have shed light on the diversity of responses to environmental conditions. The mean square fluctuation and resilience (effective force constant) calculated from neutron scattering data are, in fact, thermodynamic parameters informing the atomic statistical mechanics of the sample. The link was made by Gabel et al. (Reference Gabel2009), who combined neutron scattering and differential scanning calorimetry data on the denaturation of native human butyrylcholinesterase and its ‘aged’ soman conjugate. These authors found a direct correlation between the temperature dependence of the mean square displacements and that of the calorimetric specific heat. The transition entropy and enthalpy extracted from a two-state model fit of the neutron data were, within experimental accuracy, compatible with the values determined by differential scanning calorimetry.
Hm MalDH dynamics adapts to solvent environment
The well-characterised extreme halophilic enzyme Hm MalDH provided the opportunity to evaluate how its thermal molecular dynamics depended on salt and water isotope solvent composition (Tehei et al., Reference Tehei2001). By using neutron spectroscopy to measure dynamics and circular dichroism (CD) to measure loss of secondary structure, the authors examined the dynamics and denaturation of Hm MalDH in molar NaCl or KCl in H2O or D2O. BSA was also measured in molar NaCl as a mesophilic control. In the case of Hm MalDH, dynamic resilience (from neutron scattering) increased progressively with increasing stability (from CD), from 2 M NaCl in H2O, via 2 M KCl in D2O to 2 M NaCl in D2O, reflecting the strength of ionic interactions and the stronger hydrophobic interactions in D2O. Interestingly, the opposite was observed for BSA; its resilience is higher in H2O where it is less stable than in D2O. The ‘softer’ dynamics for BSA in D2O indicated that higher thermostability is associated with higher entropic fluctuations, due to the stronger hydrophobic effect in D2O. In Hm MalDH, higher stability is associated with increased resilience indicating the dominance of enthalpic terms arising from bonded interactions like salt bridges and the hydrated ion binding previously shown in the halophilic protein.
Adaptation to temperature and pressure in bacteria
Tehei et al. (Reference Tehei2004) applied neutron spectroscopy as a function of temperature to compare mean square fluctuation (MSF) dynamics in live bacterial cells of the psychrophile, Aquaspirillum arcticum; the mesophiles, E. coli and Proteus mirabilis; the thermophile, T. thermophilus; and the hyperthermophile, Aquifex pyrofilus (Figure 7). Note the deviation from the line of the MSF in the psychrophile corresponding to denaturation in the proteome above 20 °C. The resilience (effective force constant <k′>), calculated from the slope of the MSF with temperature, increased from 0.2 N/m for the psychrophiles via 0.4 N/m for the mesophiles to 0.6 N/m for the hyperthermophile. The increased resilience at high physiological temperature showed the dominance of enthalpic rather than entropic terms in the stabilisation free energy, while functional flexibility is maintained by temperature (Figure 8a). Thermal adaptation was evident from the observation that the extrapolated root mean square atomic fluctuation was found to be similar (~1 Å on the 0.1 ns time-scale) for each organism at its physiological temperature – 4 °C, 37 °C, 75 °C, 85 °C, respectively (Figure 8b). For the psychrophiles, which thrive at low physiological temperatures, evolution would have selected low resilience, sufficient for stabilisation, in order to allow sufficient flexibility for function.
Mean square fluctuations (MSF) plotted against absolute temperature for A. arcticum (A), P. mirabilis (B), T. thermophilus (C), and A. pyrofilus (D). Effective mean force constants, describing mean macromolecular resilience, were calculated from the slopes of the straight-line fits in the temperature region where the bacterial proteins are stable. From Tehei et al. (Reference Tehei2004).
Figure 7. Long description
A four-panel multi-plot showing data for A. arcticum (a), P. mirabilis (b), T. thermophilus (c), and A. pyrofilus (d).
Each panel shares the same axes:
- The vertical Y-axis represents M S F denoted as <u super 2> in Angstroms squared, ranging from 1.0 to 2.5.
- The primary horizontal X-axis at the bottom represents absolute temperature T in Kelvin (K), ranging from 270 to 310.
- A secondary horizontal X-axis at the top represents temperature T in degrees Celsius, ranging from 0 to 40.
Panel (a) A. arcticum: Data points show a linear increase from approximately 1.2 to 1.6 Angstroms squared between 275 and 305 Kelvin. A single outlier point sits higher at 2.0. The calculated force constant is <k prime> = (0,21 plus or minus 0,03) N/m.
Panel (b) P. mirabilis: Data points show a steady linear increase from 1.3 to 1.6 Angstroms squared. The calculated force constant is <k prime> = (0,39 plus or minus 0,01) N/m.
Panel (c) T. thermophilus: Data points show a very shallow linear increase, remaining relatively flat near 1.4 Angstroms squared. The calculated force constant is <k prime> = (0,67 plus or minus 0,11) N/m.
Panel (d) A. pyrofilus: Data points show a shallow linear increase from 1.2 to 1.4 Angstroms squared. The calculated force constant is <k prime> = (0,60 plus or minus 0,01) N/m.
(a) The mean effective force constant is lowest for the psychrophile bacteria (softest) and highest (stiffest) for the thermophile. (b) Similar rms fluctuation amplitudes extrapolated at respective physiological temperatures. From Tehei et al. (Reference Tehei2004).
Figure 8. Long description
Panel a is a bar graph showing the mean effective force constant k prime in Newtons per meter on the y-axis, ranging from 0.0 to 0.8. The x-axis lists three categories. First, Psychrophile has a single white bar at approximately 0.18 with a blue error bar. Second, Mesophile has two bars, a white bar at 0.42 and a green hatched bar at 0.39. Third, Thermo forward slash Hyperthermophile has a white bar at 0.68 with a large red error bar and a red hatched bar at 0.60. Below each category is a small square inset showing a microscopic cellular cross-section in red, green, and yellow tones respectively.
Panel b is a bar graph showing r m s Fluctuation in Angstroms on the y-axis, ranging from 0.0 to 1.6. The x-axis categories match panel a. Psychrophile shows a single white bar at 1.1. Mesophile shows a white bar and a green hatched bar, both approximately 1.25. Thermo forward slash Hyperthermophile shows a white bar and a red hatched bar, both approximately 1.28. Below these are three identical blue-toned microscopic cellular cross-section insets.
Caviglia et al. (Reference Caviglia2025) found that in E. coli (EC), a mesophile, protein diffusion in the cytoplasm undergoes a pronounced slowdown at temperatures near cellular death, coinciding with the early stages of unfolding. These authors compared the phenomenon by CD and neutron spectroscopy in EC and cells from the psychrophile Psychrobacter arcticus (PA), and the hyperthermophile Aquifex aeolicus (AA), respectively. A dynamic arrest closely mirroring the behaviour of EC, occurring near its corresponding cell death temperature of ~370 K, was observed in AA. In contrast, the behaviour of PA was markedly different; protein diffusion abruptly declined well above ~295 K, the lethal threshold of the psychrophile. In all three organisms, the arrest of protein diffusive dynamics is driven by the unfolding of a small fraction of the proteome. In PA, the arrest appeared to be decoupled from cell death, which likely resulted from the loss of biological activity in a few key enzymes already within the low-temperature range.
In the cold deep-sea or at hydrothermal vents, pressure and temperature thermodynamic parameters are often coupled. Koduru et al. (Reference Koduru2026) addressed this question in an investigation of the functional and folding properties of an exonuclease, Cnase, from the Gram-positive piezophile, Carnobacterium sp. AT7, isolated at 2,500 m depth and ∼2 °C, in the Aleutian Trench in the North Pacific. They found that Cnase is a bona fide exonuclease, despite its high negative charge, and that it largely conserves the structure and fold of its well-studied mesophilic homolog, staphylococcal nuclease, Snase, in spite of significant differences in their sequences.
Adaptation to temperature in an archaeal hyperthermophile malate dehydrogenase and mesophile rabbit muscle malate dehydrogenase
Tehei et al. (Reference Tehei2005) and Tehei and Zaccai (Reference Tehei and Zaccai2005) extended the study of dynamics to hyperthermophilic malate dehydrogenase from Methanococcus jannaschii and a mesophilic homologue, lactate dehydrogenase from Oryctolagus cuniculus (rabbit) muscle. Again, root mean square fluctuations were found to be similar for both enzymes at their respective optimal activity temperature. Resilience values are higher by an order of magnitude for the high hyperthermophile (1.5 N/m for M. jannaschii malate dehydrogenase, compared to 0.15 N/m for O. cunniculus lactate dehydrogenase. Similar to the bacterial case (Tehei et al., Reference Tehei2004), thermoadaptation appears to have been achieved by evolution through selection of appropriate structural resilience in order to preserve specific protein structure while allowing the conformational flexibility required for activity.
Adaptation to heat and salt stress in H. salinarum
Marty et al. (Reference Marty2013) investigated in vivo molecular dynamics in H. salinarum cells under thermal stress by neutron spectroscopy combined with microbiological characterisation. Modifications were detected with respect to unstressed cells, reflecting a softening of protein structures consistent with partial denaturation. The results validated the neutron scattering method for the study of molecular dynamics in the proteome of living cells. In a further in vivo study, Vauclare et al. (Reference Vauclare2015) addressed molecular dynamics adaptation to salt stress in H. salinarum cells. The extreme halophile accumulates molar concentrations of KCl in the cytosol to counterbalance the external osmolarity due to NaCl. As a consequence, cytosolic proteins are permanently exposed to high salt conditions. In non-adapted systems, such conditions might well promote protein aggregation and/or denaturation. In vitro studies, however, like the ones on Hm MalDH (see Section ‘Halophilic malate dehydrogenase (Hm MalDH)’) have shown that proteins from extreme halophiles are themselves obligate halophiles. In their natural setting, halophilic organisms often face significant variations in environmental salt concentration. The results from neutron scattering coupled with microbiological characterisation show that deleterious effects already occur in the H. salinarum proteome molecular dynamics when the external salt concentration is still relatively high, suggesting the onset of survival mechanisms quite early on when the environmental salt concentration decreases.
Molecular dynamics comparison of five haloarchaeal species
Carré et al. (Reference Carré2025) pursued the in vivo biophysical description of the proteome in five haloarchaeal species with respect to protein halophilic character (deduced from mass spectrometry and sequence traits), intracellular salinity (from inductively coupled plasma atomic emission spectroscopy analysis), protein stability (measured by nano-differential-scanning-fluorimetry) and global resilience and flexibility (assessed from neutron spectroscopy). The five strains with diverse NaCl and Mg2+ growth requirements were ranked on a proteome halophilicity scale as follows: H. sodomense > H. salinarum > H. marismortui ≅ H. volcanii > H. mediterranei. High intracellular accumulation of Mg2+ and low intracellular accumulation of K+ were correlated with higher stability and resilience. Sequence traits associated with mean proteome halophilicity, such as decreased hydrophobicity and increased acidity, weighted by the relative abundance of each protein, were also correlated with stability and resilience. Haloferax mediterranei, however, was found to be an exception as its proteome showed the highest in cellulo molecular stability and resilience associated with the fewest sequence traits related to halophilicity, highlighting the significance of the intracellular salt environment in determining proteome biophysical properties.
Adaptation to high hydrostatic pressure
Martinez et al. (Reference Martinez2016) studied adaptive strategies to HHP via molecular dynamics in the piezophilic archaeon, Thermococcus barophilus, and the closely related pressure-sensitive Thermococcus kodakarensis, by neutron scattering, at 0.1 and 40 MPa (400 bar). T. kodakarensis exhibited higher intrinsic stability, with reduced hydration water dynamics and reduced fast protein and lipid dynamics under pressure. In contrast, the proteome of T. barophilus responds differently to pressure. While the diffusion coefficient of hydration water was reduced, as in T. kodakarensis, the fast protein and lipid dynamics were slightly enhanced with increasing pressure, indicating a surprising decoupling between hydration water and proteome molecular flexibility under pressure. Martinez et al. (Reference Martinez2016) proposed that the high flexibility of the T. barophilus proteome associated with reduced hydration water dynamics under HHP may be one of the aspects of molecular adaptation to HHP.
Peters et al. (Reference Peters2026) reviewed their recent work on the molecular bases for adaptation to temperature and HHP in archaeal membranes, and on proteome and cell-wide adaptation strategies seen in adapted versus non-adapted species.
Conclusion and future directions
The field of biophysics in extremophile research is vast. In this review, I have limited myself to the topics I know best, particularly those related to macromolecular adaptation to extreme salinity, temperature, or pressure. Adaptation to other severe stresses, such as radiation resistance in the bacterium Dienococcus radiodurance, for example, is just as important to understand. Gaifas et al. (Reference Gaifas2025) used a combination of advanced fluorescence microscopy and cryogenic electron tomography to study the composition and structural organisation of the various layers composing D. radiodurans’ cell envelope and uncovered the mechanisms that drive its unique mode of division. Their findings enhanced the comprehension of a fundamental process required for bacterial proliferation and raised questions about the evolutionary advantages of an unusual division mode and its role in the organism’s remarkable resilience.
Extremophile research continues to be driven by biological exploration and discoveries, on the one hand, and progress in methods and instrumental developments, on the other.
With respect to structural biology, I quote, in full, the abstract of Grimes et al. (Reference Grimes2018):
Macromolecular crystallography (MX) has been a motor for biology for over half a century and this continues apace. A series of revolutions, including the production of recombinant proteins and cryo-crystallography, have meant that MX has repeatedly reinvented itself to dramatically increase its reach. Over the last 30 years synchrotron radiation has nucleated a succession of advances, ranging from detectors to optics and automation. These advances, in turn, open up opportunities. For instance, a further order of magnitude could perhaps be gained in signal to noise for general synchrotron experiments. In addition, X-ray free-electron lasers offer to capture fragments of reciprocal space without radiation damage, and open up the subpicosecond regime of protein dynamics and activity. But electrons have recently stolen the limelight: so is X-ray crystallography in rude health, or will imaging methods, especially single-particle electron microscopy, render it obsolete for the most interesting biology, whilst electron diffraction enables structure determination from even the smallest crystals? We will lay out some information to help you decide.
With respect to SAXS and SANS, instrumentation has been built for time-resolved studies (see, for example, Mahieu et al., Reference Mahieu2020), and simultaneous X-ray and neutron scattering measurements (Metwalli et al., Reference Metwalli2020). Contrast variation methods are also being developed for SAXS studies of macromolecular assemblies of variable internal electron density, such as protein-RNA/DNA complexes, or detergent/lipid-solubilised membrane proteins (Da Silva et al., Reference Da Silva2026). The solvent electron density is adjusted by the addition (up to molar concentrations) of small, electron-rich molecules such as sugar, glycerol, salt, or heavy-atom compounds like lanthanides. An important issue is whether or not the conformational state of the biomolecule under study is maintained after long incubation times in the presence of the contrast agents. Da Silva et al. (Reference Da Silva2026) studied a bacterial endo-β-1,4-mannanase and a malate dehydrogenase in solution with high concentrations of sucrose and with a lanthanide compound. They found that the global conformation and the oligomeric state of the enzymes were preserved over the entire concentration range of the contrast agents, while their enzymatic activities remained unaltered even after incubation times up to 24 h.
At the same time, biologists continue to discover new avenues of research in the study of extremophiles. Planetary exploration stimulated considerable interest in the extremophile scientific community (Jebbar et al., Reference Jebbar2020). The icy moons of the outer Solar System harbour potentially habitable environments in vast oceans below a layer of ice (Soderlund et al., Reference Soderlund2024). These are characterised by extremes in temperature, pressure, pH, and other physico-chemical conditions. From an icy moon perspective, Jebbar et al. (Reference Jebbar2020) reviewed the study of microbial diversity and biosignatures in terrestrial extreme environments. Closer exploration of the moons is on the way. The Jupiter Icy Moons Explorer (JUICE) of ESA (the European Space Agency) was launched in 2023; it will arrive at Jupiter and its moons in 2031 and will investigate the interior of the three icy Galilean moons, Ganymede, Callisto, and Europa (Van Hoolst et al., Reference van Hoolst2024).
While sharing the interest in the search for extra-terrestrial life (Maurel and Zaccai, Reference Maurel and Zaccai2001), I note that, after 50 years of research since the discovery of extremophiles in the 1970s, when I started my career, it is an understatement to write that there is still much to learn about life on our own planet.
The deep sea is the largest carbon sink on Earth. It hosts abundant biodiversity that buttresses the carbon cycle. The subsurface accounts for about 90% of the biomass of archaea and bacteria. Hilmi et al. (Reference Hilmi2023) discuss the unique biophysical, socioeconomic and governance characteristics of the deep sea in an essay published in Frontiers in Climate. They critically assess the opportunities to mitigate climate change through carbon fluxes and storage. This includes carbon transferred from the atmosphere to deep water by organic and inorganic carbon pumps, carbon sequestered in the skeletons and bodies of deep-sea organisms, or carbon buried within sediments or captured in carbonate rock.
The dark habitats underground, known as the deep biosphere, are the most poorly understood on Earth, in particular with respect to their influence on global biogeochemical processes (Edwards et al., Reference Edwards2012). New technologies have emerged, permitting researchers to do active, manipulable experimentation in situ within the subsurface, and new realms are being investigated from deep in South African mines to sediments in the middle of oceanic gyres (rotating currents) (Edwards et al., Reference Edwards2012).
Thermal adaptation of Methanococcales proteins has already been discussed in the main body of this review. Because of the crucial biogeochemical role of methanogens there is now growing interest in the structural biology of their metabolic processes. Methanogens are anaerobic methane-producing archaea that derive energy from methanogenesis, a biological process responsible for most global methane emissions. Methane is a potent greenhouse gas and a high-energy fuel, and methanogens could help solve the dual challenge posed by climate change and energy shortage (Lyu et al., Reference Lyu2022). Anaerobic methanotrophic archaea (ANME) are crucial to carbon cycling on Earth. They oxidise methane in anoxic niches by transferring electrons to nitrate, metal oxides, or sulphate-reducing bacteria. Methane capture, the first reaction in Anaerobic Oxidation of Methane (AOM), has been proposed to be carried out by the Methyl-Coenzyme M Reductase (MCR). No ANMEs have been isolated so far, hampering the biochemical investigation of anaerobic methane oxidation. Circumventing the isolation barrier, Müller et al. (Reference Müller2025) purified MCRs directly from microbial enrichments to compare them with methanogenic and methanotrophic homologues. The native structures obtained at atomic resolution revealed the detailed architecture of the methane-capturing enzyme. Despite their physiological differences, these ANMEs have highly conserved MCR structures, similar to homologues from methanogenic Methanosarcinales. Labelling with gaseous krypton did not reveal any internal channels that would facilitate alkane diffusion to the active site, as observed in the ethane-specialised enzyme. Based on their data, Müller et al. (Reference Müller2025) concluded that methanotrophic MCRs should follow the same radical reaction mechanism proposed for the methane-generating homologues. The findings highlight the importance of studying natural systems.
In conclusion, it is appropriate to quote the abstract of a Consensus Statement published in Nature Reviews Microbiology (Cavicchioli et al., Reference Cavicchioli2019):
In the Anthropocene, in which we now live, climate change is impacting most life on Earth. Microorganisms support the existence of all higher trophic life forms. To understand how humans and other life forms on Earth (including those we are yet to discover) can withstand anthropogenic climate change, it is vital to incorporate knowledge of the microbial ‘unseen majority’. We must learn not just how microorganisms affect climate change (including production and consumption of greenhouse gases) but also how they will be affected by climate change and other human activities. This Consensus Statement documents the central role and global importance of microorganisms in climate change biology. It also puts humanity on notice that the impact of climate change will depend heavily on responses of microorganisms, which are essential for achieving an environmentally sustainable future.
Acknowledgements
I am happy to express my gratitude to Richard Henderson, Serge Perez, Jean-Luc Popot, Nathan Zaccai, and my colleagues, Bruno Franzetti and Dominique Madern, for their critical reading of the manuscript, as well as to Frank Gabel for his unwavering support. I am especially grateful to the QRB editor and referees, whose comments and requests for revision have substantially enriched the manuscript.
Financial support
I acknowledge financial support from the Extremophilic and Large Molecular Assemblies (ELMA) group at the Institut de Biologie Structurale (CEA, CNRS, UGA), Grenoble, France, and from the Biology Deuteration Chemistry and Soft Matter (BDCS) group at the Institut Laue Langevin, Grenoble, France.
Competing interests
The author declares no competing interests.
Open access
