Stability and stabilization of globular proteins in solution

Journal of Biotechnology 79 (2000) 193 – 203 www.elsevier.com/locate/jbiotec

Review article

Stability and stabilization of globular proteins in solution Rainer Jaenicke Institut fu¨r Biophysik und Physikalische Biochemie, Uni6ersita¨t Regensburg, Uni6ersitatsstrasse 31, D-93040 Regensburg, Germany Received 5 July 1999; received in revised form 22 October 1999; accepted 29 October 1999

Abstract Proteins are multifunctional: their amino acid sequences simultaneously determine folding, function and turnover. Correspondingly, evolution selected for compromises between rigidity (stability) and flexibility (folding/function/ degradation), to the result that generally the free energy of stabilization of globular proteins in solution is the equivalent to only a few weak intermolecular interactions. Additional increments may come from extrinsic factors such as ligands or specific compatible solutes. Apart from the enthalpic effects, entropy may play a role by reducing the flexibility (cystine bridges, increased proline content), or by water release from residues buried upon folding and association. Additional quaternary interactions and closer packing are typical characteristics of proteins from thermophiles. In halophiles, protein stability and function are maintained by increased ion binding and glutamic acid content, both allowing the protein inventory to compete for water at high salt. Acidophiles and alkalophiles show neutral intracellular pH; proteins facing the outside extremes of pH possess anomalously high contents in ionizable amino acids. Global comparisons of the amino acid compositions and sequences of proteins from mesophiles and extremophiles did not result in general rules of protein stabilization, even after including complete genome sequences into the search. Obviously, proteins are individuals that optimize internal packing and external solvent interactions by very different mechanisms, each protein in its own way. Strategies deduced from specific ultrastable proteins allow stabilizing point mutations to be predicted. © 2000 Elsevier Science B.V. All rights reserved. Keywords: Corresponding states; Extremophiles; Hyperthermophiles; Protein folding; Stabilization; Thermostability

1. Ecological background Life on earth exhibits an enormous adaptive capacity: Except for centers of volcanic action, the surface of the earth, from the abyssal region of the deep sea to heights beyond the Himalaya, E-mail address: [email protected] (R. Jaenicke)

represents ‘biosphere’. In quantitative terms, the currently known limits of the biologically relevant physical variables are − 40°CB T B +115°C, PB 120 MPa,  1B pHB 11 and water activities\ 0.6, corresponding to salinities up to 6 M (Jaenicke, 1991a). During evolution, organisms achieved viability under the given extreme conditions either by ‘escaping’ or ‘compensating’ the stress, or by enhancing the stability of their cellu-

0168-1656/00/$ - see front matter © 2000 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 8 - 1 6 5 6 ( 0 0 ) 0 0 2 3 6 - 4

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

194

Table 1 Degradative reactions important to irreversible protein denaturationa Reaction

Amino acids involved and characteristic reactions

Deamidation (formation of Asn, Gln (10% reactivity) especially in Asn-Gly isoaspartate) and Asn-Ser sequences Racemization Isomerization Glycation

Oxidation

Proteolysis Enzymatic Non-enzymatic Photodegradation

a

Comments

Independent of pH, product: iso-Asp as substrate of methyl transferase, leading to repair or clearance

Asp Pro (cis-trans isomerization)

Catalyzed by peptidyl prolyl-cis-trans isomerases (PPIs), (cf. Jaenicke, 1987, 1996a) Lys and other amino groups react with reducing Protein cross-linking by Maillard reaction, sugars involved in in vivo degradation, (cf. Barrett, 1985; Jaenicke, 1998) Cys X sulfenic “ cysteic A either direct Oxid. Thiolate mechanism catalyzed by Cu(II) or or SH/SS exchange (mixed disulphides) Fe(II), enzymatically by protein disulphide isomerase (PDI, DsbA, DsbB) Met X sulfoxide “sulfone Significant both in vivo and in vitro, in the presence of oxygen containing radicals Polypeptides “ amino acids General, protease-specific caused by hydrolysis Caused by contaminating proteases or autolysis of peptide bond At C-terminal side of Asp at C-terminal Asn, Over a wide pH range between Asp and Pro Trp“kynurenin, “ N-formyl kynurenin Caused by non-ionizing or ionizing radiation, Tyr“ DOPA, dityrosin depending on the local micro-environment of cysteine “2 Cys the amino acids

For details and further references, see Volkin et al. (1995).

lar inventory. To survive extremes of temperature and pressure, there is no alternative besides mutative adaptation. The result of this adaptation is that the respective organisms require the extremes for their whole life cycle. Depending on the variables, organisms that require high and low temperature, high hydrostatic pressure and high salinity are called thermo- or psychrophiles, barophiles and halophiles. Given the long-term ‘experiment’ of evolution, the various types of extremophiles may teach us how to modify proteins to preserve their functional state under biotechnologically relevant extreme conditions. Focusing on proteins, the upper temperature and pressure-limits of growth correspond to the borderline of irreversible denaturation and chemical degradation. The adaptive strategies allowing metabolic and reproductive activity in spite of this challenge are still unresolved. Denaturation/renaturation studies on ultrastable model systems such as proteins from hyperthermophilic microorganisms may serve as a key to a deeper understanding

of the stability problem and its application in industrial practice (Jaenicke et al., 1996; Jaenicke and Bo¨hm, 1998, 2000).

2. Temperature limit of protein stability Proteins, independent of their mesophilic or extremophilic origin, consist exclusively of the 20 canonical natural amino acids. In the multicomponent system of the cytosol, they may undergo a wide variety of covalent modifications, most of which are favored at elevated temperature or extremes of pH (Table 1). At temperatures beyond 100°C, hydrothermal degradation outruns biosynthesis (Jaenicke, 1991a, 1998). Thus, hyperthermophiles must compensate for the decomposition of amino acids either by using compatible protectants or by enhanced synthesis and repair. Little is known about the chemistry involved, and even less is known about protection and repair (Ichikawa and Clarke, 1998). Applying tempera-

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

195

Fig. 1. Hierarchy of protein structure determining both sequential folding and incremental stabilization of polypeptide chains. (A) Roman numbers mark the levels of Linderstro¨m-Lang’s nomenclature ignoring supersecondary structure and subdomains. (B) Arabic numbers along the vertical arrow refer to the pathway of folding and association which is parallelled by increasing stability: (1) helix formation and pairing of b-strands; (2) formation of supersecondary structure; (3) hydrophobic collapse to molten globule state of subdomains and domains; (4) formation of tertiary structure by docking of subdomains and domains; (5) association of subunits. (C) Two-step model for protein folding. The hypothetical structures of the polypeptide chain in the unfolded (U), molten globule (MG) and native state (N) are shown along with the free energy landscape (Peng et al., 1995; Jaenicke, 1999).

tures \ 100°C, the thermal stabilities of the common amino acids are (Val, Leu) \Ile \ Tyr \ Lys \His\ Met\ Thr \Ser \Trp \ (Asp, Glu, Arg, Cys). In many cases, the half-lives of the degradation reactions are significantly shorter than the generation time of hyperthermophilic microorganisms (Bernhardt et al., 1984); obviously, the life time of biomolecules needs to be compatible with their resynthesis. The temperature at which ATP hydrolysis becomes the limiting factor for viability lies between 110 and 140°C (Leibrock et al., 1995). This temperature limit coincides with the temperature range at which the hydrophobic hydration of proteins vanishes (Jaenicke, 1991a). Apparently, both the integrity of the natural amino acids and the formation of the hydrophobic core upon protein folding are essential for viability. Extrinsic factors and compatible solutes may enhance the stability, this way shifting limits of growth (Carpenter et al., 1993).

3. Intermolecular interactions and protein stability Proteins exhibit marginal stabilities equivalent to only a small number of weak intermolecular interactions (Dill, 1990; Jaenicke, 1991a). Average

values for the Gibbs free energy of stabilization (DG°stab) of medium size globular proteins are on the order of 50 kJ mol − 1 (Pfeil, 1998). In this respect, proteins from extremophiles do not differ strongly from their mesophilic counterparts. Their adaptation, either intrinsic or through interaction with extrinsic factors, is accompanied by increases in DG°stab between 10 and 100 kJ mol − 1, again the equivalent of only a few non-covalent interactions (Jaenicke and Bo¨hm, 2000). With regard to the intrinsic effects, stabilization may involve all levels of the hierarchy of protein structure, local packing of the polypeptide chain, secondary and supersecondary structural elements, domains and subunits (Fig. 1) (Jaenicke, 1987; Ragone and Colonna, 1995; Ragone, 1997; Facchiano et al., 1998; Jaenicke, 1998, 1999). The driving forces that are responsible for protein folding reflect the given hierarchy; they engage (i) in the nearest neighbor short-range interactions (throughchain as well as through-space) that lead to optimum packing and minimum cavity volume; and (ii) in the entropy effects associated with water release from hydrophobic surfaces and charge clusters which tend to form the core of the protein and peripheral ion pairs (Jaenicke, 1991b, 1996a; Bieri and Kiefhaber, 1999; Jaenicke and

196

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

Bo¨hm, 2000). Evidently, both the enthalpic and entropic contributions to the free energy of stabilization are affected by the biologically relevant extremes of physical conditions. A quantitative analysis of the effects of temperature, pressure, charge and water activity on the various types of intermolecular interactions is presently not possible. Even at a fixed set of the four variables, no unequivocal balance sheet of the contributions of the various weak interactions to DG°stab can be given, simply because both DG° and DDG° are marginal differences of the large energies piled up by the vast number of attractive and repulsive forces between all atoms in a given protein molecule1. Significant contributions come from electrostatic forces between polar and ionized groups (hydrogen bonds, dipole interactions, ion pairs) and hydrophobic effects involving nonpolar residues. The physical nature of the latter is now considered as being entropic and enthalpic due to significant contributions from van der Waals in-

Fig. 2. Temperature dependence of the free energy of stabilization of proteins (DH and DG in kJ mol − 1, DS in J mol − 1 K − 1). (A) DGstab versus T profile of sperm whale myoglobin, calculated from the contributions of the specific enthalpy and entropy of denaturation (Privalov and Gill, 1989). (B) Hypothetical DGstab versus T profiles for (a) mesophilic and (b–d) thermophilic proteins; Tm and T%m are the corresponding melting temperatures, and Topt and T%opt the optimal growth temperatures of the mesophilic and thermophilic organisms, respectively. (C) DGstab versus T profiles of selected mesophilic and hyperthermophilic proteins, illustrating the shifts observed for small (:70 residues) structurally related proteins: b-barrel DNA-binding protein Sso7d (–), all-b tyrosine kinases BtkSH3 (---) and Tec-SH3 ( –· –·), a-spectrin ( — ), CspA from B. subtilis ( ) and Csp from T. maritima (). The arrows at :30 and 80°C refer to the physiological optimum temperatures of the organisms, respectively (Knapp et al., 1998; D. Wassenberg and R. Jaenicke, unpublished results). 1

For a relatively small protein such as myoglobin the number of atoms is of the order of 5000; DG°stab amounts to 55 kJ/mol (Privalov, 1979).

teractions (Kauzmann, 1959; Baldwin, 1986; Privalov and Gill, 1989; Dill et al. 1990; Dill, 1990; Makhatadse and Privalov, 1996; Schellman, 1997). How the given potentials are accumulated to finally yield the experimental Gibbs free energy of stabilization (measured by calorimetry or spectroscopy) is still not fully understood, and estimates of the effects of temperature, pressure, charge and water activity on DG°stab remain an insoluble problem (Jaenicke, 1991b; Lifson et al., 1994; Murphy, 1995; Pace et al., 1996; Lifson, 1997). In the case of temperature, measurements of intermolecular forces have shown that the interaction energy per polar group can exceed the thermal energy (Leikin and Parsegian, 1994). Hydrogen bonds are favored at low temperature and become weaker as the temperature is increased. Due to the compensatory effects in the total energy balance, quantitative predictions with respect to the significance of any specific type of interaction cannot be made with confidence. Extremophilic proteins do not exhibit properties qualitatively different from non-extremophilic ones. Adaptive alterations tend to conserve the normal mesophilic characteristics. Considering the temperature dependence of the enthalpy and entropy of stabilization of globular proteins in solution, the Gibbs–Helmholtz equation predicts the DG° versus T profile to represent a skewed parabola, suggesting equilibrium denaturation transitions both at high and low temperatures (Fig. 2A) (Becktel and Schellman, 1987; Franks, 1995). Evidently, enhanced thermal stability (reflected by a shift of the denaturation temperature) may be accomplished by flattening the DG°stab versus T profile or by shifting it either towards higher temperature or higher free energy (Fig. 2B). In all three cases, Tm would be increased to the higher value T%m. So far experimental data corroborate the first two alternatives (Fig. 2C). Due to the parabolic shape of the stability profiles and the observation that proteins in their functional state ‘at work’ exhibit only marginal stabilities, the temperatures of maximum free energy are necessarily below the temperatures of optimal growth (Topt and T%opt). Comparing homologous proteins from mesophiles and extremophiles under their respective optimum

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

Fig. 3. Equilibrium transitions of two-domain and oligomeric proteins. (A) Urea-dependent unfolding/folding of bovine gBcrystallin at pH 2. N, U, I: native, fully and partially unfolded protein; u222 nm (D), s20, w ( ). Data for the N-terminal fragment parallel the I “U transition (Jaenicke, 1999). (B) GdmCl-dependent deactivation of tetrameric (M4) pig muscle LDH ( ) and the ‘proteolytic dimer’ (M2) lacking the N-terminal decapeptide () (Opitz et al., 1987). (C) GdmCl-induced deactivation of internal (), core-glycosylated ( ) and external () yeast invertase (with 0, 34, 65% carbohydrate) (Kern et al., 1992a,b).

conditions, in many cases a high degree of similarity in their physical and biochemical properties has been found (see below). The physiology of ‘corresponding states’ has been summarized in a most lucid way by G.N. Somero (1983, 1995).

4. Structural hierarchy and stabilization of proteins Several experimental approaches have been used to assign specific structural alterations to changes in stability: Selection of temperature-sensitive mutants; systematic variations of amino acid residues in the core or in the periphery of model proteins; crosslinking or joining of polypeptide chains; fragmentation of domain proteins or modifications of connecting peptides between domains; alterations of subunit interactions by mutagenesis or solvent perturbation (Matthews, 1996). As a result, protein stability has been found to be accomplished either by the covalent polypeptide backbone with or without disulfide-bridges, or by the above mentioned cumulative contributions of non-covalent interactions involved in the hierarchy of protein structure. In considering the increments of local structural elements, limited proteolysis may be used in order to determine the minimum length of

197

a polypeptide chain that is required to still form an intrinsically stable native-like structure. In this context, NMR analysis has shown that oligopeptides down to six residues may form stable (nonrandom) conformations, supporting the idea that local structures may serve as ‘seeds’ in the folding process (Wright et al., 1988; Jaenicke, 1991b; Chakrabartty and Baldwin, 1995). However, it is important to note that these short fragments do not necessarily adopt the same conformation in unrelated protein structures; e.g. reverse turn motifs observed in small peptides seem to be absent in the known three-dimensional structures of proteins containing these sequences (Creighton, 1988). With regard to the stability of protein fragments, it has been known for a long time that proteins are cooperative structures showing mutual stabilization of structural elements. To find out at which fragment size native-like structure is no longer formed, thermolysin was used as a model. The N-terminal portion of the enzyme is found to stabilize the all-helical C-terminal domain. This may be shortened to the 62-residues three-helix bundle without losing much of its native (secondary) structure. In shortening the polypeptide chain, the free energy of stabilization drops steadily to half its value, while the temperature limit of denaturation is only shifted from 87 to 64°C. At a size of 20 residues, no residual structure is left; instead, ‘wrong interactions’ cause aggregation (Vita et al., 1989). As illustrated in Fig. 1, the incremental stabilization observed for fragments or subdomains holds also at the higher levels of the hierarchy of protein structure. Striking examples illustrating the mutual stabilization of domains are the modular eye-lens crystallins and their homologs (Jaenicke, 1999). In the case of the monomeric gB-crystallin, the two-domain protein shows bimodal denaturation transitions with significant mutual stabilization of the domains (Fig. 3A). The same holds for interactions between subunits in oligomeric or multimeric proteins such as lactate dehydrogenase (LDH) (Jaenicke, 1999) or tobacco mosaic virus protein (TMVP) (Lauffer, 1966, 1967). Taking tetrameric LDH as an example, Fig. 3B illustrates the difference in stability between the ‘proteolytic dimer’ and the native

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

198

dimer of dimers. The monomer is structured only as a folding intermediate on the assembly pathway. The same holds for the separate domains which exhibit an even shorter life-time so that the nicked monomer is only obtained in joint reconstitution experiments, i.e. through ‘mutual chaperoning’ (Opitz et al., 1987). There are examples where stabilization due to subunit association goes to the extreme: phage, viruses, chromatin, ferritin, bacterial surface layers, etc. In all these cases, the monomers have average size, stability and flexibility. This way, the complex structures exhibit their characteristic low turnover, while the capsomers or subunits may still perform their essential morphopoietic or catalytic functions, without being inhibited with respect to translocation, targeting, processing, etc. In addition to the sequence-encoded increments discussed so far, extrinsic factors such as ions, cofactors, metabolites, compatible solutes and covalent conjugates may contribute to protein stability (Hensel and Ko¨nig, 1988; Jaenicke, 1998). As an example, Fig. 3C illustrates the effect of glycosylation on the thermal stability of invertase from yeast. It is significant because it shows that glycosylation alters the yield rather than the mechanism of protein folding and association; in this function, it can take over the role of molecular chaperones (Kern et al., 1992a,b, 1993).

5. Forces and mechanisms Stability refers to the maintenance of the spatially defined functional state under extreme conTable 2 Relative amino thermophilesa

acid

Charged residues (DEKRH) Polar/uncharged residues (GSTNQYC) Hydrophobic residues (LMIVWPAF)

compositions

of

mesophiles

Mesophiles

Thermophiles

24.11%

29.84%

31.15

26.79

44.74

43.36

and

a One-letter abbreviations of amino-acid residues in brackets. For details, see Deckert et al. (1996).

ditions. High-resolution 3D structures gained in the crystalline state and in solution may be determined to a resolution better than 1 A, . However, even at this level of precision, there is no way to calculate the free energy of stabilization from the coordinates, nor can dynamic data be established that would allow the unique structure-function relationship of proteins to be elucidated in an unambiguous way. The reason is that in general the polypeptide chain fluctuates between preferred conformations with amplitudes and angles up to 50 A, and 20°, respectively (Huber, 1988). In the crystal and in concentrated solution, molecular properties may be perturbed, because lattice forces or mere intermolecular interactions may enforce local positions of the polypeptide chain which are not representative of the functionally relevant solution structure. The central issue in the adaptation of biomolecules to extreme conditions is the conservation of their functional state, which means a well-balanced compromise between stability and flexibility (Tsou, 1988, 1998; Jaenicke, 1991a). As has been mentioned, basic mechanisms of molecular adaptation are changes in packing density, charge distribution, hydrophobic surface area, and in the ratio of polar/non-polar or acidic/basic residues. Available complete genome sequences give an idea how different temperature optima of mesophiles and thermophiles are reflected at the level of the amino acid compositions of their protein repertoire (Table 2). Some correlations seem to hold; for example, compared to mesophiles, genomes of thermophiles encode higher levels of charged amino acids, primarily at the expense of uncharged polar residues. Available sequence-based attempts to define ‘traffic rules’ of protein stabilization have so far been statistically insignificant (Bo¨hm and Jaenicke, 1994). This is not surprising, because the comparable order of magnitude of DG°stab and DDG°stab does not suggest dramatic changes in the intermolecular interactions, even in the most extreme cases. This prediction is stressed by the fact that, in terms of the stability increments per residue, the thermal energy (kT) exceeds the intrinsic stabilization. This indicates that the overall stability of a polypeptide chain must involve cooperativity,

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

because the addition of stability increments per amino acid residue in the process of structure formation would not allow to overcome the thermal energy (Jaenicke and Lilie, 2000). If the molecular mass of a protein is too small to provide the necessary size of a cooperative unit, the structure may be stabilized by covalent crosslinking or complex formation with ligands or other components. Both improve the stability by decreasing the entropy of unfolding or dissociation. The requirement for flexibility is fulfilled by the subtle compensation of attractive and repulsive interactions. As extremophilic proteins are usually structurally homologous to their mesophilic counterparts, they exhibit similar properties. The essential adaptive alteration tends to shift the normal characteristics to the respective extreme, in the sense that under physiological conditions the molecular properties are similar. In the terminology of the ecologist this means that adaptation to extremes of physical conditions tends to maintain ‘corresponding states’ regarding overall topology, flexibility and solvation (see above). For the close relationship with respect to the spatial structure, numerous examples have been reported in recent years (cf. Adams and Kelly, 2000). In the present context, two examples with widely differing molecular weights may be sufficient to prove the case: (i) The 7 kDa single-domain cold-shock proteins from B. subtilis, B. caldolyticus and Thermotoga maritima exhibit DG°stab-values differing by factors of 2 and 2.5, while their all-b topology is the same, with r.m.s. deviations of their polypeptide chains below 1 A, (Perl et al., 1998; Jaenicke and Bo¨hm, 1998, 2000); (ii) Lactate dehydrogenases from pig, B. stearothermophilus and Thermotoga with Tm-values of 45, 60 and 91°C, and sequence identities of 40.5 and 39.5% show r.m.s.d. B2.2 A, (Auerbach et al., 1998). Regarding 3D structure and flexibility, glyceraldehyde phosphate dehydrogenase (GAPDH) from yeast and Thermotoga as well as isopropyl malate dehydrogenase (IPMDH) from E. coli and Thermus thermophilus have been studied in detail. Comparing Stokes radii, hydrogen – deuterium exchange and chaotropic activation, it is evident that the dynamic properties of the mesophilic enzymes at

199

:25°C come close to those of the (hyper-)thermophilic ones at : 70°C (Korndo¨rfer et al., 1995; Jaenicke, 1996b; Wallon et al., 1997; Jaenicke and Bo¨hm, 1998; Za´vodszky et al., 1998). Adaptation to specific solvation properties have been investigated in connection with halophiles whose proteins require high salt concentrations for activity and denature at low salt. This anomalous feature is correlated with an excess of acidic over basic residues and an increase in the number of intramolecular salt bridges compared to non-halophilic homologs (Eisenberg, 1995). The hypothetical tendency to compete with the highly concentrated salt solution for water of hydration by exposing acidic amino acid side chains has been confirmed by the significant increase in hydrogen bonds in the case of ferredoxin from H. marismortui (Frolow et al., 1996). Except for the need to maintain normal hydration, halophilic proteins do not seem to exhibit specific structural properties. The phenomenological description of extremophilic adaptation in terms of ‘corresponding states’ gives no insight regarding the molecular mechanism of the increase in rigidity of proteins. This information has been provided by B.W. Matthews’ detailed analysis of more than 1000 mutants of bacteriophage T4 lysozyme (Matthews, 1995, 1996; Jaenicke, 1999). Available high-resolution data on (hyper-)thermophilic proteins did not unravel new mechanisms. Instead, they confirmed how subtle the different attractive and repulsive interactions are balanced in native globular proteins, and how much even marginal local strain in the three-dimensional structure may affect protein stability. The following conclusions in terms of a ‘repertoire of adaptive mechanisms’ may be drawn: (i) due to the large number of hydrogen bonds in proteins, the marginal difference in H-bond strength between water–water and water–protein will be magnified to an energy change that may exceed DG°stab. Available DDG°-values prove : 5 additional hydrogen bonds to be sufficient for thermophilic stabilization; (ii) a-helices, extended b-structures and helix-dipole interactions contribute significantly to protein stability; (iii) since only :70% of the theoretically available hydrophobic contri-

200

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

butions are realized as a consequence of the balance of favorable contributions to DG°stab on protein folding, there is ample space for additional optimization in terms of improved core packing and decrease in hydrophobic surface; (iv) an increase in the state of association may contribute to (thermal) stability; (v) the same holds for shortened loops and decreased cavity volumes; (vi) to a first approximation, increments of stabilization are additive (Bryan, 1995; Murphy, 1995; Pace et al., 1996; Colacino and Crichton, 1997; Jaenicke, 1999). For protein engineering, the given results provide guidelines, however, unequivocal predictions or general strategies for protein stabilization are still not available (Bo¨hm and Jaenicke, 1994).

6. Protein folding and protein stability The driving forces responsible for protein folding and protein stabilization are the same, because along the pathway of folding and association the polypeptide chain gains increasing stability (Jaenicke, 1999). With the small increments of the weak local interactions in mind, the question arises how self-organization can proceed at elevated temperature or pressure. Compared to temperature, biologically relevant changes in pressure have no significant effects because the reaction volumes and activation volumes involved in most biochemical reactions are comparatively small (Groß and Jaenicke, 1994). In order to simulate the effect of high temperature on folding, the in vitro denaturation/renaturation of hyperthermophilic GAPDH from Thermotoga maritima has been studied at 0–100°C. Refolding over a wide temperature range was found to yield the native state, even beyond the physiological temperature, thus indicating that the anomalous thermal stability refers not only to the native state, but also to intermediates on the pathway of folding and association. At 0°C, the enzyme was found to be trapped as a tetrameric intermediate with molten globule-like properties; upon shifting the temperature beyond :10°C, the native state is reached in a rapid reaction (Rehaber and Jaenicke, 1992; Jaenicke, 1996b). Regarding the folding kinetics,

available data allow the conclusion that increasing intrinsic stability is reflected by a decrease in the rate of unfolding. In this context, mutant studies have shown that enhanced stability may be determined kinetically rather than thermodynamically (Pappenberger et al., 1997). Comparing the unfolding and folding kinetics of the cold-shock proteins from B. subtilis, B. caldolyticus and Thermotoga maritima (see above), the rates of unfolding were shown to be inversely proportional to the thermal stability. On the other hand, despite numerous sequence variations among the three proteins, folding occurred extremely fast (halftime : 1 ms) with closely similar kinetics (Perl et al., 1998). In the case of large multidomain or oligomeric proteins, kinetic partitioning, i.e. aggregation as a side reaction, competes with in vitro folding (Jaenicke, 1996a). As a consequence, due to (partial) irreversibility, thermodynamic stability data cannot be determined. In vivo, heat-shock proteins take care of misfolding and aggregation. They have also been discovered in (hyper-)thermophiles (Adams and Kelly, 2000). Their extremely high expression levels clearly suggest important cellular functions, however, their detailed mechanism is still unresolved (Trent et al., 1997; Jaenicke and Bo¨hm, 1998; Minuth et al., 1998, 1999).

7. Conclusions Proteins, due to the delicate balance of stabilizing and destabilizing interactions, are only marginally stable. Contributions to the net free energy of stabilization range from local interactions at the level of elements of secondary structure and subdomains to interactions between domains and subunits. Enhanced intrinsic stability in thermophiles is the cumulative effect of minute improvements of local interactions: higher packing efficiency (mainly through van der Waals interactions), networks of ion pairs and/or hydrogen bonds (including a-helix stabilization), and reduction of conformational strain (loop stabilization). Thus, thermostability corresponds to increased rigidity at low temperature which is shifted to

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

normal flexibility at the physiological temperature level. Evidently, evolutionary adaptation tends to maintain corresponding states with respect to conformational flexibility, that way optimizing biological function under specific conditions. The picture that has been emerging from intensive in vitro studies shows a mosaic of a wide variety of combinations of adaptive strategies which has been tested and supported by the prediction of mutagenesis effects; however, general rules of stabilization cannot be given (yet).

Acknowledgements This review is dedicated to Professor Hans Neurath on the occasion of his 90th birthday. I should like to thank the Deutsche Forschungsgemeinschaft and the Fonds der Chemischen Industrie for continuous support of research summarized in this review. Fruitful discussions with Drs G. Auerbach, G. Bo¨hm, R. Huber, M. Kretschmar, T. Oshima, C.N. Pace, V. Rehaber, F.X. Schmid, B. Schuler, H. Schurig, G.N. Somero, K.O. Stetter, M. Wenk and H. Zuber are gratefully acknowledged.

References Adams, M.W.W., Kelly, R.M., 2000. Hyperthermophilic enzymes. Meth. Enzymol. (in press). Auerbach, G., Ostendorp, R., Prade, L., et al., 1998. LDH from the hyperthermophilic bacterium T. maritima: The crystal structure at 2.1 A, resolution reveals strategies for intrinsic stabilization. Structure 6, 769–781. Baldwin, R.L., 1986. Temperature dependence of the hydrophobic interaction in protein folding. Proc. Natl. Acad. Sci. USA 83, 8069 – 8072. Barrett, G.C., 1985. Chemistry and Biochemistry of Amino Acids. Chapman and Hall, New York, pp. 354–375. Becktel, W.J., Schellman, J.A., 1987. Protein stability curves. Biopolymers 26, 1859–1877. Bernhardt, G., Lu¨demann, H.-D., Jaenicke, R., Ko¨nig, H., Stetter, K.O., 1984. Biomolecules are unstable under ‘black smoker’ conditions. Naturwissenschaften 71, 583–586. Bieri, O., Kiefhaber, T., 1999. Elementary steps in protein folding. Biol Chem 380, 923–929. Bo¨hm, G., Jaenicke, R., 1994. On the relevance of sequence statistics for the properties of extremophilic proteins. Int. J. Pept. Protein Res. 43, 97–106.

201

Bryan, P.N., 1995. Site-directed mutagenesis to study protein folding and stability. In: Protein Stability and Folding. Meth. Mol. Biol. 40, 271 – 289. Carpenter, J.F., Clegg, J.S., Crowe, J.H., Somero, G.N., 1993. Compatible solutes and macromolecular stability. Cryobiology 30, 201 – 241. Chakrabartty, A., Baldwin, R.L., 1995. Stability of a-helices. Adv. Protein Chem. 46, 141 – 176. Colacino, F., Crichton, R.R., 1997. Enzyme thermostabilization: the state of the art. Biotechnol. Genet. Eng. Rev. 14, 211 – 277. Creighton, T.E., 1988. On the relevance of non-random polypeptide conformations for protein folding. Biophys. Chem. 31, 155 – 162. Deckert, G., Warren, P.V., 1996. et al., The complete genome sequence of the hyperthermophilic bacterium Aquifex aeolicus. Nature 392, 353 – 358. Dill, K.A., 1990. Dominant forces in protein folding. Biochemistry 29, 7133 – 7155. Dill, K.A., Privalov, P.L., Gill, S.J., Murphy, K.P., 1990. The meaning of hydrophobicity. Science 250, 297 – 298. Eisenberg, H., 1995. Life in unusual environments: progress in understanding the structure and function of enzymes from extreme halophilic bacteria. Arch. Biochem. Biophys. 318, 1 – 5. Facchiano, A.M., Colonna, G., Ragone, R., 1998. Helix stabilizing factors and stabilization of thermophilic proteins: an X-ray based study. Protein Eng. 11, 753 – 760. Franks, F., 1995. Protein destabilization at low temperatures. Adv. Protein Chem. 46, 105 – 139. Frolow, F., Harel, M., Sussman, J.L., Mevarech, M., Shoham, M., 1996. Insights into protein adaptation to a saturated salt environment from the crystal structure of a halophilic 2Fe-2S ferredoxin. Nat. Struct. Biol. 3, 452 – 458. Groß, M., Jaenicke, R., 1994. Proteins under pressure: The influence of high hydrostatic pressure on structure, function and assembly of proteins and protein complexes. Eur. J. Biochem. 221, 617 – 630. Hensel, R., Ko¨nig, H., 1988. Thermoadaptation of methanogenic bacteria by intracellular ion concentration. FEMS Microbiol. Lett. 49, 75 – 79. Huber, R., 1988. Flexibility and rigidity of proteins and protein-pigment complexes. Angew. Chem. Int. Ed. 27, 79 – 99. Ichikawa, J.K., Clarke, S., 1998. A highly active protein repair enzyme from the extreme thermophile T. maritima. Arch. Biophys. Biochem. 358, 222 – 231. Jaenicke, R., 1987. Folding and association of proteins. Progr. Biophys. Mol. Biol. 49, 117 – 237. Jaenicke, R., 1991a. Protein stability and molecular adaptation to extreme conditions. Eur. J. Biochem. 202, 715 – 728. Jaenicke, R., 1991b. Protein folding: local structures, domains, subunits and assemblies. Biochemistry 30, 3147 – 3161. Jaenicke, R., 1996a. Protein folding and association: significance of in vitro studies for self-organization and targeting in the cell. Curr. Top. Cell. Reg. 34, 209 – 314.

202

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203

Jaenicke, R., 1996b. GAPDH from T. maritima: strategies of protein stabilization. FEMS Microbiol. Rev. 18, 215–224. Jaenicke, R., 1998. What ultrastable proteins teach us about protein stabilization. Biochemistry (Moscow) 63, 312–321. Jaenicke, R., 1999. Folding and stability of domain proteins. Progr. Biophys. Mol. Biol. 71, 155–241. Jaenicke, R., Bo¨hm, G., 1998. Stabilization of proteins: what extremophiles teach us about protein stability. Curr. Opin. Struct. Biol. 8, 738 – 748. Jaenicke, R., Bo¨hm, G., 2000. Thermostability of proteins. Meth. Enzymol. (in press). Jaenicke, R., Lilie, H., 2000. Folding and association of oligomeric and multimeric proteins. Adv Protein Chem (in press). Jaenicke, R., Schurig, H., Beaucamp, N., Ostendorp, R., 1996. Structure and stability of ‘hyperstable proteins’: glycolytic enzymes from the hyperthermophilic bacterium T. maritima. Adv. Protein. Chem. 48, 181–269. Kauzmann, W., 1959. Some factors in the interpretation of protein denaturation. Adv. Protein. Chem. 14, 1–63. Kern, G., Schmidt, M., Buchner, J., Jaenicke, R., 1992a. Glycosylation inhibits the interaction of invertase with the chaperone GroEL. FEBS Lett. 305, 203–205. Kern, G., Schu¨lke, N., Schmid, F.X., Jaenicke, R., 1992b. Quaternary structure and stability of internal, external and core-glycosylated invertase from yeast. Protein Sci. 1, 120– 131. Kern, G., Kern, D., Jaenicke, R., Seckler, R., 1993. Kinetics of folding and association of differently glycosylated variants of invertase from C. cere6isiae. Protein Sci. 2, 1862– 1868. Knapp, S., Mattson, P.T., Christova, P., et al., 1998. Thermal unfolding of small proteins with SH3 domain folding. Proteins: Struct. Funct. Genet. 31, 309–319. Korndo¨rfer, I., Steipe, B., Huber, R., Tomschy, A., Jaenicke, R., 1995. The crystal structure of holo-GAPDH from the hyperthermophilic bacterium T. maritima. J. Mol. Biol. 246, 512 – 521. Lauffer, M.A., 1966. Polymerization-depolymerization of TMVP. VII: a model. Biochemistry 5, 2440–2446. Lauffer, M.A., 1967. Polymerization-depolymerization of TMVP: refinement of a model. Chimia 21, 460–462. Leibrock, E., Bayer, P., Lu¨demann, H.-D., 1995. Non-enzymatic hydrolysis of ATP at high temperatures and high pressure. Biophys. Chem. 54, 175–180. Leikin, S., Parsegian, V.A., 1994. Temperature-induced complementarity as a mechanism for biomolecular assembly. Proteins 19, 73 – 76. Lifson, S., 1997. Wanderings in the fields of science. Comprehensive Biochemistry. Semenza Ed. 40, 1–55. Lifson, S., Felder, C.E., Libmann, J., Shanzer, A., 1994. The role of energy calculations in the design, synthesis and study of biologically active Fe(III) carriers. In: Wipff, G. (Ed.), Computational Approaches to Supramolecular Chemistry. Kulver Acad. Publ, Dordrecht, pp. 349–376. Makhatadse, G.I., Privalov, P.L., 1996. Energetics of protein structure. Adv Protein Chem 47, 307–345.

Matthews, B.W., 1995. Studies on protein stability with T4 lysozyme. Adv. Protein Chem. 46, 249 – 278. Matthews, B.W., 1996. Structural and genetic analysis of the folding and function of T4 lysozyme. FASEB J. 10, 35 – 41. Minuth, T., Frey, G., Lindner, P., Stetter, K.O., Jaenicke, R., 1998. Recombinant ultrastable chaperonin from the archaeon Pyrodictium occultum shows chaperone activity in vitro. Eur. J. Biochem. 258, 837 – 845. Minuth, T., Henn, M., Rutkat, K., et al., 1999. The recombinant thermosome from the hyperthermophilic archaeon Methanopyrus kandleri: in vitro analysis of its chaperone activity. Biol. Chem. 380, 55 – 62. Murphy, K.P., 1995. Non-covalent forces important to the conformational stability of protein structures. In: Shirley, B.A. (Ed.), Protein Stability and Folding Meth. Mol. Biol, vol. 40, pp. 1 – 34. Opitz, U., Rudolph, R., Jaenicke, R., Ericsson, L., Neurath, H., 1987. Proteolytic dimers of LDH: characterization, folding and reconstitution or the truncated and nicked polypeptide chain. Biochemistry 26, 1399 – 1406. Pace, C.N., Shirley, B.A., McNutt, M., Gajiwala, K., 1996. Forces contributing to the conformational stability of proteins. FASEB J. 10, 75 – 83. Pappenberger, G., Schurig, H., Jaenicke, R., 1997. Disruption of an ionic network leads to accelerated thermal denaturation of GAPDH of the hyper-thermophilic bacterium T. maritima. J. Mol. Biol. 274, 676 – 683. Peng, Z.-Y., Wu, L.C., Schulman, B.A., Kim, P.L., 1995. Does the molten globule have a native-like tertiary fold? Phil. Trans. R. Soc. Lond. B. Biol. Sci. 348, 43 – 47. Perl, D., Welker, C., Schindler, T., et al., 1998. Conservation of rapid two-state folding in mesophilic, thermophilic and hyperthermophilic cold-shock proteins. Nat. Struct. Biol. 5, 229 – 235. Pfeil, W., 1998. Protein Stability and Folding: A Collection of Thermodynamic Data. Springer, Berlin, Heidelberg, New York. Privalov, P.L., 1979. Stability of proteins. Adv. Protein Chem. 33, 167 – 241. Privalov, P.L., Gill, S.J., 1989. The hydrophobic effect: a reappraisal. Pure Appl. Chem. 61, 1097 – 1104. Ragone, R., 1997. Protein association and stability of thermophilic enzymes. Protein Pept. Lett. 4, 277 – 279. Ragone, R., Colonna, G., 1995. Do globular proteins require some structural peculiarity to best function at high temperatures? J. Am. Chem. Soc. 117, 16 – 20. Rehaber, V., Jaenicke, R., 1992. Stability and reconstitution of GAPDH from the eubacterium T. maritima. J. Biol. Chem. 267, 10999 – 11006. Schellman, J.A., 1997. Temperature, stability and the hydrophobic interaction. Biophys. J. 73, 2960 – 2964. Somero, G.N., 1983. Environmental adaptation of protein: strategies for the conservation of critical functional and structural traits. Comp. Biochem. Physiol. A 76, 621 – 633. Somero, G.N., 1995. Proteins and temperature. Annu. Rev. Physiol. 57, 43 – 68.

R. Jaenicke / Journal of Biotechnology 79 (2000) 193–203 Trent, J.D., Kagawa, H.K., Yaoi, T., Olle, E., Zaluzec, N.J., 1997. Chaperonin filaments: The archaeal cytoskeleton. Proc. Natl. Acad. Sci. USA 94, 5383–5388. Tsou, C.-L., 1988. Folding of the nascent peptide chain into a biologically active protein. Biochemistry 27, 1809–1812. Tsou, C.-L., 1998. The role of active site flexibility in enzyme catalysis. Biochemistry (Moscow) 63, 253–258. Vita, C., Fontana, A., Jaenicke, R., 1989. Folding of thermolysin fragments: hydrodynamic properties of isolated domains and subdomains. Eur. J. Biochem. 183, 513–518. Volkin, D.B., Mach, H., Middaugh, C.R., 1995. Degradative covalent reactions important to protein stability. In: Protein Stability and Folding. Meth. Mol. Biol. 40, 35–63.

.

203

Wallon, G., Kryger, G., Lovett, S.T., Oshima, T., Ringe, D., Petsko, G.A., 1997. Crystal structures of E. coli and S. typhimurium IPMDH and comparison with their thermophilic counterpart from T. thermophilus. J. Mol. Biol. 266, 1016 – 1031. Wright, P.E., Dyson, H.-J., Lerner, R.A., 1988. Conformation of peptide fragments of proteins in aqueous solution: Implications for initiation of protein folding. Biochemistry 27, 7167 – 7175. Za´vodszky, P., Kardos, J., Svingor, A., Petsko, G.A., 1998. Adjustment of conformational flexibility is a key event in the thermal adaptation of proteins. Proc. Natl. Acad. Sci. USA 95, 7406 – 7411.