Environmental adaptation of proteins: Strategies for the conservation of critical functional and structural traits

Camp. Biochem. Physiol. Vol. %A. No. 3, pp. Q-633.

1983

0300.9629:83 93.00 + 0.00

,

Printed in Great Britain

1983 Pergamon Press Ltd

ENVIRONMENTAL ADAPTATION OF PROTEINS: STRATEGIES FOR THE CONSERVATION OF CRITICAL FUNCTIONAL AND STRUCTURAL TRAITS GEORGE Marine

Biology

Research

N.

SOMERO

Division A-002, Scripps Institution of Oceanography, San Diego, La Jolla, CA 92093. USA (Received

I8 FrbruarJ

University

of California,

1983)

Abstract-l. Comparative studies of lactate dehydrogenases (LDHs) and skeletal muscle actins from vertebrates adapted to widely different temperatures and hydrostatic pressures reveal major conservative trends in protein evolution and adaptation. For enzymes, ligand binding, as estimated by apparent Michaelis constant (K,,,) values, is strongly conserved at physiological temperatures. pressures, intracellular pH values and osmotic compositions of different organisms. The catalytic rate constants (k,,, values) of enzyme homologues are highest for enzymes of low-body-temperature organisms. a trend that can be interpreted in terms of temperature compensation of metabolism. For skeletal muscle actins. the enthalpy and entropy changes accompanying the assembly of filamentous (F) actin from globular (G) actin are highest in high-body-temperature species and especially low in polar and deep-sea fishes. The thermal stability of G-actin is positively correlated with adaptation temperature. except in the case of actins of deep-sea fishes, which are also highly heat stable. Hydrophobic interactions between actin subunits may be of reduced importance in low-body-temperature animals and, especially. in deep-sea fishes. The differences in enthalpy and entropy changes during the G-to-F transformation favor a close conservation of the equilibrium constant for actin assembly under physiological conditions of temperature and pressure for different species. These adaptive patterns in enzymes and actin are likely to reflect changes in protein primary structure. 2. The appropriate values for protein traits such as ligand binding abilities and catalytic rates are also shown to be established by the composition of the low molecular weight constituents of the cytosol. For example, the use of a combination of urea and methylamine solutes for osmoregulation by marine elasmobranchs is shown to be a mechanism which permits the conservation of key protein traits at high osmolarities. The methylamine solutes such as trimethylamine-N-oxide have effects on proteins opposite to those of urea, and at the approximately 2:l concentration ratio of urea to methylamines, these counteracting effects are virtually complete. 3. Regulation of hydrogen ion activity (pH) also is shown to play a major role in the conservation of critical protein traits. The importance of temperature-dependent pH in ectotherms is discussed in terms of stabilizing binding abilities and maintaining correct regulatory and structural sensitivities of proteins, The buffering capacity of tissues reflects the potential of the tissue for generating acidic end-products during anaerobic metabolism. Skeletal muscle, especially white locomotory muscle of fishes. is highly buffered relative to red locomotory muscle and heart muscle. Much of the difference in buffering may be due to different amounts of histidine-containing dipeptide buffers in the tissues. However. comparisons of muscle-type (M4) and heart-type (H4) LDHs show that the M, isozyme has approximately twice the number of histidine residues per tetramer of the H, isozyme. This difference between M, and H, isozymcs which occur predominantly in anaerobicallyand aerobically-poised tissues, respectively, suggests that selection for buffering capacity may be a heretofore unappreciated facet of protein evolution.

INTRODUCTION

1982). Freshwater teleosts have total osmotic contents roughly one-third that of seawater, while marine elasmobranchs may have body fluids slightly more concentrated than seawater. Fishes also differ markedly in life style, for example, in their feeding strategies and abilities for powerful bursts of locomotion. Tunas, for instance, have enormously developed locomotory abilities and they must swim continuously to satisfy the requirements of ram ventilation. At the other extreme, many deep-sea fishes are sluggish, sit-and-wait or float-and-wait feeders who have only very limited capacities for locomotion (Sullivan and Somero, 1980). These differences in feeding and locomotory habits are reflected in widely different capacities for muscle glycolysis and varying needs for coping with acidic

No other vertebrate group experiences the wide environmental variations noted for fishes. In the case of temperature, many high-latitude fishes survive continuously at temperatures near the freezing point of seawater (approx - 1.8 ‘C). Desert pupfishes (genus Cyprinodon) of the American Southwest may inhabit waters with temperatures of 40-43”C (Brown and Feldmeth, 1971). Deep-sea fishes typically have body temperatures near 24’C, and in addition these species may live at pressures up to several hundred atmospheres. Fishes also are unusual among vertebrates in their wide variety of osmotic strategies, i.e. in the total osmotic content and types of osmolytes used in osmotic regulation (reviewed in Yancey et al.. <-BP 76/h

I,

621

622

GEORGE N. SOMERO

metabolic end-products (Castellini and Somero, 1981). Because of the extreme diversity in environmental and biological attributes of fishes, this group of vertebrates has been an especially suitable collection of organisms for studying the physiology and biochemistry of adaptation. Indeed, each of the characteristics listed above-temperature, hydrostatic pressure, osmotic concentration, osmolyte composition and locomotory capacity-will be expected to be manifested in the design of the biochemical systems of fishes, especially in the properties of the enzymic and contractile protein systems. The focus of this review is on these two classes of proteins, with emphasis being placed on the nature of the major adaptive strategies that facilitate the conservation of the most critical protein traits in the face of widely different environmental and physiological conditions. We will consider, first, what traits of enzymic and contractile proteins must be conserved under most, if not all, conditions. Secondly, we will examine the types of molecular adaptations that effect these conservations of critical functional and structural traits. We will consider each environmental or physiological factor individually, beginning with an examination effects. However, it will be of temperature apparent from our analysis that, regardless of the environmental/physiological factor being considered, only three basic strategies of adaptation seem to be involved in maintaining key protein traits within optimal ranges. First, changes in protein primary structure may adaptively modify the inherent properties of the enzymic or contractile protein. Second, in some situations, notably in the case of activity- and depth-related changes in enzymic activity, alterations in protein concentration may be a critical type of adaptation. Third, in many circumstances the types and concentrations of proteins remain unchanged, but the comnosition of the milieu in which the proteins fun&ion is adaptively altered. Prior to considerina the details of these different adaptive mechanisms,-it is appropriate to comment on the choice of experimental tissues and organs in which these phenomena have been studied. Despite the fact that this symposium is focused on the fish heart, most of the data discussed in this review are from studies of fish skeletal muscle. To a biochemist, the emphasis on fish muscle over fish hearts is understandable, since skeletal muscle is present in abundance in a fish, whereas the heart represents a minor fraction of body mass. Thus, for studies in which substantial quantities of purified proteins are needed, fish skeletal muscle is a most convenient starting material. Fortunately, the apparent absence of information about fish heart proteins and their adaptations is more apparent than real. All of the environmental factors we discuss impinge on all tissues of a fish’s body, so the fundamental adaptations to temperature, hydrostatic pressure and osmolytes noted for skeletal muscle proteins almost certainly exist in similar or identical form in proteins of other tissues, including heart. One exception to the similarity of heart and muscle will be dwelt on, however. This exception involves the metabolism of acidic end-products and hydrogen ions. We will show that heart muscle, and aerobically-poised red loco-

motory muscle, differ significantly from white skeletal muscle in metabolic and buffering properties. And we will show data on LDHs that are suggestive of major evolutionary changes in amino acid composition directed towards enhancing buffering capacity in muscles where high rates of acidic end-product formation occur. TEMPERATURE

The greatest amount of information on adaptation of fish proteins comes from studies of homologous proteins from species adapted to different temperatures. The emphasis given to temperature in studies of protein adaptation reflects the fact that all of the key properties of proteins, namely their structural integrities, their catalytic functions, and their regulatory properties, are sharply influenced by temperature changes. The differences noted among homologues of a given type of protein from differently-adapted species often are large and provide a clear image of the types of adaptations needed to conserve the critical protein traits serving as the focus of this review.

The initiation of a cycle of catalytic activity and the regulation of catalysis demand proper ligand (substrate, cofactor and allosteric modulator) binding abilities by enzymes. In the case of substrate binding in particular, there are ample data to support the generalization that proper binding necessitates that the enzyme not only be able to form highly stable complexes with its substrate(s), but that the binding capacity is such that the enzyme does not become saturated with substrate. That is, enzymes must be able to bind their substrates tightly, but not too tightly. The gist of the argument here is that an enzyme must retain a certain amount of reserve capacity, i.e. an ability to increase its rate of function as substrate levels rise, as they might during a vigorous bout of locomotory activity. Were an enzyme to bind substrate(s) extremely tightly and always be saturated with substrate(s), an increase in substrate supply could not lead to an immediate rise in catalytic rate. As a consequence of this inability of the enzyme to increase its rate of function in response to rising substrate concentration, demands for elevated flux through the pathway will not be met, and enormous increases in the concentrations of highly reactive intermediates may occur (Atkinson, 1969). The foregoing argument in favor of an enzyme design in which a reserve capacity is maintained even though the enzyme is able to bind substrates effectively at the low substrate concentrations present in the cell (typically 10-j M and lower; Fersht, 1977) is supported by the data presented in Fig. 1 and Table 1. For skeletal muscle-type (M,,, A.,) isozymes of lactate dehydrogenase (LDH), the apparent Michaelis constant (K,) of pyruvate is strongly conserved among species at their normal body temperatures, and the K,,, of pyruvate is held in a range near the pyruvate concentrations found in resting muscle (Table 1). Thus, under quiescent conditions, pyruvate levels in invertebrate skeletal muscle tend to be slightly below Km values, which means that the

Environmental

adaptation of proteins

623

Fig. l.(A) The effect of temperature on intracellular pH and the pH of the imidazole/HCl buffer system used in the studies shown in frame B. The lighter lines indicate the range of measured intracellular pH values determined in ectotherms with different body temperatures (see Somero, 1981). The darker line connecting the closed circles shows the pH vs temperature relationship used in the determination of the K,,, of pyruvate values shown in B. (B) The influence of temperature on the K,,, of pyruvate for M,-LDHs purified from vertebrates having different body temperatures. Species studied include a mammal (the rabbit), two reptiles (alligator lizard and the desert iguana) and a variety of fishes (all other species). The open symbols refer to deep-sea fishes. The dark regions of the lines connecting the points (measured K,,, values) indicate the body temperatures of the species. Ninety-five per cent confidence intervals around the K,,, values are shown. Data are from Yancey and Somero (I 978a), Somero et al. (I 983) and Donahue and Somero (unpublished). Figure is from Somero et al (1983).

Table I. Pyruvate concentrations in skeletal muscle of different vertebrates Animal

[Pyruvate] (mmol/kg tissue)

Dog (37°C) Rat (37°C) Frog (20°C) Trout (12°C) Carp (8-14°C) Eel (15°C) Goldfish (5°C)

0.33 0.18 0.11 0.1 I 0.12 0.14 0.06

Goldfish (25°C) Goby fish (15°C) Goby fish (25°C) Goby fish (25°C;

0.33 0.02 0.04 0.33

swum to exhaustion) Unless otherwise indicated, values were obtained from resting muscle. Data for the goby fish, Gillichthys mirabilis, are from Walsh and Somero (1982); other data are compiled from references given in Yancey and Somero (1978b).

enzyme can approximately double its rate of pyruvate reductase activity if pyruvate concentrations increase rapidly, as might occur during a vigorous bout of locomotory activity. It is important to emphasize that both the K,,, of pyruvate of M,-LDHs and muscle concentrations are conserved among pyruvate different vertebrates. This conservation of substrate concentration and substrate binding abilities (as ap-

proximated by apparent Km values) is an evolutionary trend that has been noted in all enzyme systems so examined (Somero, 1978). In view of the fact that K,,, values typically are altered by changes in temperature (Fig. 1), the conservation of K,,, among species adapted to widely different temperatures suggests that a substantial amount of modification in protein primary structure is involved in K,,, conservation. Thus, K,, conservation relies heavily on the first adaptation strategy we discuss, namely changes in protein amino acid sequence, but we will later see that adjustments in the milieu in which enzymes function also can contribute importantly to the stabilization of K,,, values within optimal ranges. Most studies of enzymic adaptation to temperature have examined diverse species adapted to widely different temperatures. That polar and tropical fishes should have enzymes differing in their thermal responses is perhaps not surprising, but are similar types of differences, albeit on a reduced scale, also found in cases where adaptation temperatures ( = average habitat temperatures) differ by only a few degrees Celsius? In other words, what is the threshold level of temperature difference that can select for adaptive differences in enzymic traits like K,,, values? To approach this question about minimal temperature differences underlying protein evolution, it is appropriate to examine congeneric species that are otherwise closely similar except for their average body temperatures. To this end we studied the M,-LDHs of four species of eastern Pacific bar-

624

GEORGE N. SOMERO

Fig. 2. The effect of temperature on the apparent Michaelis constant (K,,,) of pyruvate for the M,-LDHs of three eastern Pacific barracudas (genus Sphymm). The K”: values for the south temperate species. S. ~~j~.~~~~.~, (not shown) do nol differ statistically from the values for the north temperate species, S. ar,gen&~~. Error bars represent standard deviations. Each point represents an average of K,, values determined with at least three different purified LDH preparations from three different individuals of each species. The pH regime shown in the inset to Fig. I wasused. See legend to Fig. 1 for experimental details. The solid portions of the lines connecting the k;, values indicate the habitat temperature ranges of the species. Figure from Graves and Somero (1982).

racudas (genus Spk,wwnu) (Graves and Somero, 1982). As illustrated in Fig. 2, even differences in average body temperature of 5-8X are apparently adequate to favor selection for adaptive differences in tu,, of pyruvate. As a result of amino acid substitutions leading to the displacements of K,, vs tetnperature curves as shown in Fig. I. the M,-LDHs of the north and south temperature species, S. urgrntea and .S. idiastes, respectively, have K,,l of pyruvate values at 18°C that are virtually the same as those of the M,-LDHs of the subtropical species (5. ~z~~u.~ff~ff) at 23 C and the tropical species (S. u&) at 26’C (Table 2). Thus, we conclude that differences in average habitat ( = body) temperature of only a few degrees Celsius are of strong enough influence on enzyme function to favor over evolutionary time spans adaptive changes in protein primary structure.

The properties

of the M,-LDH

homologues

of the

barracuda congeners also offer us insights into the manners in which the catalytic powers of enzymes, as measured by substrate turnover numbers (k,,, values) are adapted for function under different thermal conditions (Table 2). Studies of a variety of different enzymes from fishes adapted to different temperatures have shown that substrate turnover numbers or k,,, values of enzymes from cold-adapted species are higher than those of the homologous enzymes from warm-adapted species (Low et al., 1973; Borgmann and Moon. 1975; Low and Somero, 1976; Johnston and Walesby, 1977; Somero and Siebena~ler. 1979). The finding that the rate at which an enzyme molecule can convert substrate to product under a standard set of assay conditions is inversely related to the organisms’ adaptation temperatures has been interpreted as a mechanism for metabolic rate compensation to temperature. As shown for the M,-LDH reactions of the temperate, sub-tropical and tropical barracudas. there is indeed a high conservation in the k,,, value at average body temperatures (Table 2). These fine-scale differences among closeiyrelated congeners living in slightly different thermal regimes mirror the more broad-scale differences among M,-LDH k,,, values noted in comparisons of widely different species having body temperatures ranging from - 2 to 39 ‘C (Somero and Siebenaller, 1979).

Structurul

fruits

Proteins characteristically have four levels of structure. Primary structure refers to the sequence of amino acids in the individual polypeptide chains; secondary structure entails the manners in which the polypeptide chain assumes a-helical and /?-pleatedsheet configurations; tertiary structure involves the complex folding of the polypeptide into a compact, usually globular. conformation; and quaternary structure refers to the assembly of two or more po~ypeptide subunits (monomers) into a multisubunit protein. Each level of structure may reflect the effects of temperature adaptation. It has been known for decades that the resistance of proteins to heat denaturation (loss of quaternary, tertiary and secondary structure) reflects the adaptation temperatures of species (reviewed by Alexandrov. 1977). These gross denaturation studies, while suggestive of adaptive trends, suffer from the shortcoming that gross denaturation of proteins al-

Table 2. Kinetic parameters of M,-LDHs of three eastern Pacific barracudas measured 25 C and at the approximate midrange temperatures (TM) of each species ti,,, of pyruvate k,,, @ 25’ C

TM K*, of pyruvate k,,, ia’ TM

((I 25 C

(a~ TM

0.34 f 0.03 mM 893 i 54,kec

0.26 2 0.02 mM 730 5 37:‘sec

at

0.20 t 0.02 mM 658 * t9:sec

18 c

23 c

26 c

0.24 mM 667isec

0.24 mM 682isec

0.23 mM 700jsec

The K,,, values for the south temperate species, Sp~,v~~enu idiusies (not shown). did not differ significantly from values for the north temperate species. S. urgmlru. The K,, estimates arc based on measurements made with five separate LDH preparations from five individuals of each species; the k,,, estimates are based on three LDH purifications from each species. The values of kc,<, and K,,, at TMs are interpolations from Arrhenius plots and K,,, vs temperature plots (Fig. 2), respectively. Data are from Graves and Somero (1982).

Environmental adaptation of proteins most invariably occurs only at temperatures well above the upper lethal limits of the organism itself (proteins of thermophilic bacteria may be an exception). We must, therefore, examine more subtle aspects of temperature effects on protein structure if we are to discern the types of structural adaptations that lead to an optimization of protein properties at different temperatures. An appropriate starting point for this analysis is to inquire how the flexibility of protein structure relates to the constraints on life arising from differences in temperature. A most important element in this analysis is the fact that flexibility in protein structure appears to be a necessary concomitant for efficient protein function. During the catalytic cycle of enzymes, for example, the tertiary structure of portions of the enzyme may undergo rapid and reversible changes. Many regulatory events involve modulatorinduced changes in enzyme conformation and, in some cases, in subunit assembly equilibria. The need to maintain an ability to alter structure in a rapid and reversible fashion carries with it the cost of some temperature sensitivity, however, since the weak bonds which are broken or formed during these structural changes can also be disrupted by small inputs of thermal energy. cases of will examine now two We structure-function relationships in proteins to illustrate how compromises are reached between the demands for efficient function, on the one hand, and for adequately stable structures, on the other hand. We first will return to a consideration of the kinetic adaptations discussed in the previous two subsections of this essay. We noted that M,-LDHs display a strong conservation of K,,, and k,,, values at normal body temperatures. At any single temperature, an M4-LDH from a high-body-temperature species has lower K,,, and k,,, values than an M,-LDH from a low-body-temperature species (Figs 1 and 2; Table 2). Is this covariation in K, and k,,, among differently-adapted species a reflection of a causal relationship? And can the interspecific variation in these two kinetic parameters be linked with the differences in thermal stability noted among M,-LDHs (Borgmann and Moon, 1975)? Both of these questions can be given an affirmative answer. From the studies of Borgmann and Moon (1975) it appears that the differences in the kinetic and structural characteristics of M,-LDHs are in part a reflection of different amounts of structural change during the catalytic cycle. For example, during the reaction catalysed by an M,-LDH of a high-body-temperature species, more weak bonds may be formed to stabilize the enzyme-substrate-cofactor complex than in the case of an M,-LDH from a cold-adapted species. These additional weak bonds lead to a tighter enzyme-ligand complex, so that Km values can be held within the appropriate range even at higher temperatures which tend to perturb enzyme-ligand binding. Thus, the formation of a more stable enzyme-ligand complex through relatively extensive weak bond formation may be the basis for Km conservation in enzymes of high-body-temperature organisms. However, since an enzyme must return to its initial state before it can initiate another round of catalysis, these weak bonds stabilizing the

625

enzyme-ligand complex (enzyme-product complex) must be ruptured. To break these bonds energy is required, and the additional bond-breaking energy demanded in enzymes of high-body-temperature species may increase the activation free energy barrier to catalysis (Somero, 1978). In the case of LDH reaction in the direction of lactate formation, the rate-limiting step is the dissociation of oxidized cofactor (NAD) (Everse and Kaplan, 1973). For LDHs having the enzyme-NAD complex stabilized by a relatively large number of bonds, the increased energy cost of NAD dissociation will be translated directly into a higher free energy of activation for the reaction, i.e. into lower k,,, value. To at least some extent, then, the differences found in protein thermal stability can be correlated with the variations in Km and k,,, noted for these same proteins. Additionally, differences in structural stability may be related to needs for maintaining correct active site geometries under different temperature conditions, and for establishing the appropriate energy changes involved in subunit binding. We now turn to a consideration of the latter facet of protein adaptation to temperature. The assembly of filamentous (F) actin from globular, monomeric (G) actin provides a clear illustration of the manners in which assembly energetics are modified in temperature-adaptive fashion (Swezey and Somero, 1982). F-actin is the principal component of muscle thin filaments, and its assembly from G-actin subunits is strongly affected by changes in temperature. The G- to F-actin transformation has, in fact, two temperature-sensitive events. The first is the change in G-actin tertiary structure which is required to make the monomer capable of being added to a growing F-actin chain (Rich and Estes, 1976). As in the case of an enzymic protein, the G-actin monomer must be flexible in order to achieve its biological function, the formation of a polymerized state. The second temperature-sensitive event in actin assembly involves the formation of interactions between subunit contact sites. These subunit-subunit interactions form with the displacement of water from the contact sites, a common event in the assembly of multi-subunit proteins. These water displacement events are strongly endothermic (thermal energy is needed to “melt” the water from the protein interfaces that must bond together), but the overall free energy change for the assembly reaction may be negative due to the large increase in system (protein + water) entropy that results from the disruption of this organized water. These considerations suggest that the actin assembly reaction should be a revealing study system for examining temperature adaptation, since adaptive differences in G-actin flexibility and in the amount of enthalpy needed to “melt” water from the subunit contact sites may be required to insure proper actin assembly in differently-adapted species. These expectations are fully realized, as shown by a recent study of actin assembly energetics of skeletal muscle actins from vertebrates adapted to widely different temperatures (Swezey and Somero, 1982). Figure 3 illustrates the striking variation in the enthalpy (AH) and entropy (AS) changes accompanying actin polymerization in these species. Both

624

GEORGE N. SOMERO

AS

( e. u 1

poilmerll.ltOl

Fig. 3. A plot of the enthalpy (AH) and entropy (AS) changes accompanying the G-to-F transformation of actins from vertebrates adauted to different temoeratures and hydrostatic pressures. ‘This type of plot is t&ned a “compensation plot” to emphasize that the covariation in AHand AS leads to only a small amount of variation in AG between species. Symbols denote the different types of organisms: terrestrial vertebrates (a), elasmobranch fish (O), temperate water teleosts (0). Antarctic fishes (A), and very deep-living fishes (a). The numbers next to the points refer to the following species: (I) rabbit; (2) chicken; (3) Dipsosaurus dorsalis (desert iguana); (4) Sebasrolobus afascanus; (5)Seba.~rQlubus altiveli.~; (6) Caran.x hippos; (7) ~~p~nodon macularius; (8) ~hinobatus productus; (9) Thunnus alalunga (white muscle); (10) Thunnus alalunga (red mu&e); (11) G,ymnodraco acuticeps: (I 2) Pagothenia borchgrevinki; (13) Coryphaenoides armatus; (14) Coryphaenoides acrolepis; (1.5) Halosauropsis macrochir. Figure from Swezey and Somero

The interspecific differences in AH and AS of assembly have been interpreted to involve both of the actin polyaccompanying structural changes merization: alterations in the conformation of Gactin and subunit-subunit interactions (Swezey and Somero, 1982). Figure 4 illustrates the relationship between the AH of assembly and the resistance of the actins to heat denaturation (incubation of purified G-actin at 37’C). With the exceptions of the actins from the two deepest-living fishes studied, actins which also display unusually low AH and AS characteristics (Fig. 2), there is significant increase in heat stability associated with increasing AH of polymerization. Thus, at least a portion of the differences in assembly energy changes noted among species seem interpretable in the context of different energy requirements for effecting changes in the conformation of G-actin during the first event in the polymerization reaction. The more structurally-stable G-actins of high-body-temperature species may require a greater input of energy (enthalpy) to modify their conformations prior to polymerization than the actins of cold-adapted species. In addition, part of the variation in the AH and AS of polymerization may derive from differences in the amounts of water which must be displaced from the subunit contact sites. In low-buy-temperature species, reduced amounts of water displacement are suggested by the very low AH and AS changes accompanying polymerization (Swezey and Somero, 1982). Regardless of how the ditrerences in the AH and AS of assembly among interspecific homologues of actin are partitioned between differences in structural ~exibility and variatjons in water displacement, the end result of these

f

20.0 i

“I

of actin assembly strongly conserved.

at physiological temperatures is Conservation of subunit assem-

bly abilities at physiological another

critical

feature

temperatures

of protein

evolution.

is, then,

”

’

f

’

I

’

8

9

IO

I,

3

lB

(1982).

AH and AS were found to increase with rising adaptation temperature, such that polymerization of the actin from the desert iguana (Dipsosaurus dors&s) occurred with almost a IO-fold higher AH than polymerization of actins from the two Antarctic fishes, Pugot~en~a borc~gre~i~k~ and ~~~nodruco acuticeps (Fig. 3). However, the concomitant rise in AS with rising adaptation temperature “compensates” for the rise in entropy, and the net free energy change accompanying polymerization varies little among sDecies (Swezev and Somero. 1982). That is. despite large interspecific differences in thk AH and AS of actin polymerization, the equilibrium constant

’

1234567

I2

13

Heat Stablllty ( K, x t O”mln”) Fig. 4. The correlation between thermal stability of G-actin and the enthalpy of polymerization. The solid line is a linear regression of the data from species I to 12; the rz value is 0.78. The numbering of the different species is the same as in the legend to Fig. 3. Figure from Swezey and Somero (1982).

Environmental

adaptation of proteins

adaptations is a strong conservation of assembly capacity under a diverse set of temperatures and, as we show below, widely varying pressures. PRESSURE

We can develop arguments that are analogous to those made for temperature adaptations in the case of pressure effects on proteins. The same basic requirements for protein traits to be conserved within narrow limits apply for pressure as well as for temperature, and many of the basic properties of protein systems that establish their temperature sensitivities also make these systems sensitive to changes in hydrostatic pressure (reviewed in Somero et al., 1983). Binding

properties

Pressure like temperature can perturb enzyme-ligand interactions, although pressure effects seem less general than those of temperature (Somero et al., 1983). Whenever a binding event occurs with a change in system (protein + ligand + water) volume, pressure will affect the binding equilibrium. For example, if ligand binding involves the displacement of highly organized (=dense) water from around the ligand and the complementary region of the binding sites, then a volume increase due to water expansion is likely to accompany the reaction. In such a system, increases in pressure will perturb binding, and this will be reflected in K,,, values that increase with rising pressure.

621

The M,-LDHs of fishes again provide an interesting study system for analyzing adaptive trends (Fig. 5). For all of the shallow-living fishes examined, increases in measurement pressure led to increases in the K, of pyruvate and the K, of cofactor (NADH) (Siebenaller and Somero, 1979). The increases in K, of NADH are especially large (note the logarithmic scale on the ordinate of Fig. 5) and would lead to maladaptively low cofactor binding abilities at high pressures. Studies of NADH binding suggest that for M,-LDHs and other dehydrogenases, NADH(NAD) binding sites should remain cofactor-saturated, such that the direction of dehydrogenase function is established by the redox state, the NADH/NAD ratio, of the cell (Greaney and Somero, 1980). For the M,-LDHs of all of the deep-living species studied, the K, of pyruvate is completely insensitive to pressure, and the K,,, of NADH exhibits only a small increase at low pressures, and then remains stable up to the highest pressures examined. K,,, conservation is again seen to be an important feature of protein evolution. As in the case of temperature effects, we can ask about the minimal or threshold pressure at which selection will favor the development of pressureinsensitive enzymes. To address this question we examined the pressure sensitivities of the M,-LDHs of two congeners of the genus Sebastolobus (Siebenaller and Somero, 1978). Sebastolobus alascanus has a distribution range of approximately 180-44Om as adults; S. altivelis adults occur mostly between 550 and 1300 m (Siebenaller and Somero, 1978). The differences between the M,-LDHs of the Sebastolobus congeners mirror the differences noted between other shallow- and deep-living fishes (Fig. 5). This finding shows that pressure increases of only some 5&100atm are adequate to select for pressureadaptive differences in protein function. Catalytic functions

2oul I

68

204

340

476

t 6

s

,I--

5

--+---~-____~

p_____;

oz.4

2

-____

+~-.--~

.

-6 E .3

. t_---_m

.

.

Y

tPressure (otrn) Fig. 5. The effects of hydrostatic pressure on the apparent K,,, of NADH (upper panel) and pyruvate (lower panel) for

M,-LDHs of several shallow- and deep-living marine teleost fishes. Shallow-living species: Pagotheniu borchgrevinki (e), Scorpaena guttata (A), and Sebastolobus alascanus (0). Deep-living species: Coryphaenoides acrolepis (m), Antimora rosirala (A), Halosauropsis macrochir (+), and Sebustolobus altiuelis (0). Figure from Siebenaller and Somero (1979).

Deep-sea temperatures are typically very low, averaging only 24’C through much of the deep ocean. One might expect that enzymes of deep-sea fishes would resemble the homologous enzymes from coldadapted shallow-living fishes by having relatively high (= temperature-compensatory) k,,, values. However, in the case of M,-LDHs, the only set of enzyme homologues for which data are available at this time, this prediction is not realized (Somero and Siebenaller, 1979). The M,-LDHs of the several deep-sea fishes examined (listed in the legend to Fig. 5) had k,,, values of only one-half to two-thirds those of M,-LDHs from cold-adapted, shallow-living species like the Antarctic fish, Pagothenia borchgreuinki. There would appear to be two possible explanations for this seeming “failure” of M,-LDHs of deep-sea fishes to become cold-adapted in terms of k,,, values. On the one hand, the finding that the metabolic rates of deep-sea fishes are extremely low (Smith and Hessler, 1974; Smith, 1978) suggests that selection has favored low rates of enzymic activity in these organisms via use of low efficiency enzymes. However, this hypothesized basis for low k,,, values for enzymes of deep-living forms appears untenable in view of the energy available in the deep-sea, Except for organisms of the hydrothermal vent communities,

628

GEORGEN. SOMERO

deep-sea species typically encounter a low input of food into their environment (Somero et al., 1983). Because of the energy limitations facing deep-sea animals, it seems maladaptive to reduce metabolic demands by synthesizing relatively large numbers of inefficient enzyme molecules. The same reduction in metabolic rate could be achieved by synthesizing fewer copies of more efficient enzymes. In fact, the reduced metabolic rates of deep-sea fishes appear to be the result of greatly reduced enzyme concentrations; comparisons of highly active, shallow-living species like tunas, with sluggish deep-sea fishes have shown up to three order of magnitude decreases in white skeletal muscle enzymic activity per gram fresh weight of muscle (Childress and Somero, 1979; Sullivan and Somero, 1980; Castellini and Somero, 1981; Siebenaller et al., 1982). Enzymic activities in heart and brain typically show no marked differences related to depth of occurrence, however (Childress and Somero. 1979; Sullivan and Somero, 1980). This intertissue difference suggests that the reductions in skeletal (locomotory) muscle enzymic activity are specific adaptations involving selection for low metabolic rate and a reduced locomotory energy expenditure. And these reduced potentials for muscle metabolism are primarily the result of vastly lower enzyme concentrations in muscle of deep-living fishes. How, then, can we explain the low k,., values noted for M,-LDHs of deep-sea fishes? One explanation mirrors the arguments made earlier about the basis for low k,,, values of enzymes from hightemperature-adapted organisms. High pressure, like high temperature, may perturb the structures of enzymes, including those aspects of structure that insure proper ligand binding. Thus, the requirement that K,, values remain within a narrow range may dictate that more stable enzyme sructures and enzyme-ligand complexes exist in enzymes of deepsea fishes. And, as in the case of temperature adaptations, the increments in structural stability needed to offset pressure perturbations may have as their concomitant a decrease in k,,, values. Structural

truits

The conjecture that proteins of deep-living fishes may require more stable structures for successful function at high in situ pressures is supported by the discovery that skeletal muscle actins of the deepestliving fishes examined by Somero and Swezey (1982) were found to have extremely high heat stabilities relative to actins from other fishes (Fig. 4). In fact, the actins of Coryphaenoides armatus and Halosauropsis macrochir had heat stabilities comparable to actin of the desert iguana, a species which experiences body temperatures ranging up to 47°C (Norris, 1953). Thus, both high temperature and high pressure may lead to similar selective pressures for enhanced protein structural stability. Despite the fact that the actins of C. armatus and H. macrochir are so structurally stable, the enthalpy and entropy changes accompanying the G-to-F transformation are very small (Fig. 3). These low values for AH and AS suggest that, whereas the initial conversion of the G-monomer into the conformation needed for polymerization to occur may require a

considerable energy input, the amount of energy needed to “melt” water from the subunit contact sites is extremely low for these actins. This hypothesis is also reasonable from the standpoint of pressure effects on actin assembly. Displacement of densely organized water from the subunit contact sites should lead to an increase in system volume, i.e. to a potential for pressure perturbation of the G-to-F transformation. By minimizing the amount of water that must be displaced during polymerization, the volume increase and, therefore, the pressure sensitivity of the assembly process will be reduced. In fact, we have found that the actin assembly reactions for the deep-sea species are markedly less pressuresensitive than those from shallow-living or terrestrial species (Swezey and Somero, in preparation). In summary, pressure adaptation of proteins may involve two distinct types of changes. A strengthening of protein tertiary structure may be needed to maintain the proper geometries for ligand binding and subunit-subunit interactions. A reduction in events which necessitate water displacement (=volume expansion) may also be an important component of the adaptation process. Both types of adaptations appear to be present in the skeletal muscle actins of deep-living fishes. MICROENVIRONMENTAL INFLUENCES THE INTRACELLULAR MILIEU: OSMOLYTE AND pH EFFECTS

OF

Our emphasis up to this point has been on adaptations involving alterations in protein structure and, in the case of metabolic adjustments in deep-sea animals, in enzyme concentration. A properly balanced portrayal of molecular adaptation processes dictates a broader perspective, however. One must also consider the low molecular weight constituents of the cell that can have major influences on the in situ functional and structural properties of enzymic and contractile proteins. Of these microenvironmental factors of the intracellular milieu, osmotic agents (osmolytes) and the hydrogen ion are especially crucial as protein effecters. Osmolyte efsects Because fishes differ so greatly in their total osmotic content and in the types of osmolytes found in their body fluids, it is pertinent to ask how the properties of proteins reflect the osmolyte microenvironment in which they function. Marine elasmobranchs are especially worth our attention because the high osmotic concentrations of their body fluids are due to an unusual combination of nitrogenous solutes. These fishes contain in their intracellular fluids a 2 : 1 concentration ratio of urea and methylamines [trimethylamine-N-oxide (TMAO), betaine and sarcosine] (Pang et al., 1977; Yancey et al., 1982). effects of urea on protein In view of the strong structure and function, effects which often are manifested even at the urea concentrations found in elasmobranch cells, approximately 0.3-0.5 M (Pang et al., 1977; Yancey et al., 1982), the use of urea as a major osmolyte seem paradoxical. How have marine cartilaginous fishes and the coelacanth (Latimeriu chalumnae), also a urea-accumulator, adapted

Environmental

adaptation of proteins

to cope with urea perturbation of the essential enzymic traits (binding, catalysis and structure) that we have argued must be conserved in all species? Two mechanisms for conserving these traits have been discovered. In some cases, albeit probably in the great minority of cases, the proteins of urea-rich fishes appear “urea-adapted (Yancey and Somero, 1978b; Yancey et al., 1982). For example, the M,-LDHs of marine elasmobranchs [but not the homologous enzyme from the Amazon (freshwater) ray, Potamotrygon sp.] display appropriate K,,, values only when physiological concentrations of urea are present in the in tlitro assay medium. In the absence of added urea, the K,, values of the marine elasmobranch M,-LDHs are significantly lower than the values expected at the body temperatures of these fishes. Addition of 0.4 M urea to the assay medium raises the K,s to the values found in other vertebrates having similar body temperatures. In the large majority of cases so far examined, however, the proteins of elasmobranchs display no obvious urea adaptations, but instead are as ureasensitive as the homologous proteins of organisms having no, or only very low, urea concentrations in their body fluids. In these other cases, the conservation of critical protein traits is effected by an interesting solute counteraction mechanism: when urea and methylamines are both present in the physiological concentration ratio noted in cartilaginous fishes, little or no urea perturbation is found (Yancey and Somero, 1979, 1980). This counteracting solute mechanism is, illustrated in Fig. 6. The key feature of this means for reducing or eliminating urea effects is the opposite influence which methylamine solutes like TMAO have on protein properties. Thus, while urea increases the K, of ADP for creatine phosphokinase and pyruvate kinase (Fig. 6) TMAO and other methylamines reduce the K,,s. When combined, the influences of urea and the methylamines display an algebraic additivity, and at the approximately 2 : 1 ratio of [urea] : summed [methylamines], this counteracting effect is maximized. The counteracting effects of urea and methylamines are not restricted to any one absolute concentration of the osmolytes, but instead are noted over a wide range of urea and methylamine concentrations as long as the 2:l ratio of [urea]:summed [methylamines] is maintained (Fig. 6). It is important to realize as well that the counteraction of urea and methylamine effects is not restricted to proteins from urea-rich species, but rather appears to be a general characteristic of protein-solute interactions (Yancey et al., 1982). Homologous proteins from elasmobranches, teleosts and mammals have been shown to display the type of solute counteraction noted in Fig. 6. Thus, the proteins of urea-rich species are not specially adapted to enable the urea : methylamine counteraction to occur. The type of solute counteraction shown in Fig. 6 for K,, values has been found as well in studies of catalytic rates and protein structural stability. The osmolyte microenvironment of proteins therefore is seen to affect all of the traits we have termed critical and demanding of close conservation. It follows that, if one wishes to examine proteins in t&o under experimental conditions that closely mimic the in situ I‘HP 7613~ Y

T

629

DOGFISH

CREATINE KINASE

.I3

‘:

iQ t

6 loo

t

ma

ax

Total Ott-w Sdutes~M 400

600

UREA. mM

400

am

Fig. 6. The effects of different concentrations of nitrogenous osmolytes (urea and methylamines) on the K, of ADP for elasmobranch creatine kinase and pyruvate kinase. Symbols: control (0) (no urea or methylamines present), urea (A), trimethylamine-N-oxide (TMAO) (0) betaine (V), sarcosine (0) urea+TMAO (0). urea + 65mM TMAO + 55 mM sarcosine + 50 mM /I-alanine + 30 mM betaine (I) (the approximate osmolyte composition reported for a skate). Error bars represent 95% confidence intervals around the Km values. Figure from Yancey and Somero (I 980).

microenvironment of the proteins, one must duplicate the osmolyte composition of the cellular fluids in question. This caveat leads us to consider an even more critical and ubiquitous set of protein-solute interactions, those involving the hydrogen ion. pH t$ects Other than water itself, no constituent of the cellular fluids has a more pervasive influence on protein structure and function than the hydrogen ion. Protons may be involved in the catalytic mechanisms of enzymes; protonation of amino acid side-chains may lead to large-scale changes in protein assembly and function; and transitions between reduced and activated metabolic states frequently appear to involve small (0.1-0.4 pH unit) changes in pH (reviewed by Nuccitelli and Heiple, 1982). Moreover, in the case of ectotherms, the pH values of blood and cytosol vary with temperature; each 1°C rise in temperature is accompanied by a 0.015-0.020 fall in pH (Reeves, 1977; see inset to Fig. 1). It behoves us, then to consider how pH changes of the magnitude Found in living systems alter the protein traits seen as critical for physiological function.

630

GEORCEN. SOMERO

Many enzyme-ligand complexes are stabilized by bonds between the ligand and a protonated (or proton-accepting) residue in the active site. The imidazole ring of histidine is especially important in this regard, as illustrated by the inset of Fig. 7 which portrays the active site of LDH. Histidine-195 can interact with pyruvate only when the imidazole ring is protonated. When deprotonated, histidine-195 can bind lactate. Since the ability of LDH to bind pyruvate is thus apt to depend strongly on pH, we must re-examine the arguments made earlier on lu, conservation in organisms having different body temperatures to learn if the conservation of appropriate K,, of pyruvate values is influenced by pHtemperature effects. What we find, in fact, is that the close conservation K,, of pyruvate at physiological temperatures is observed only when biologically realistic pH regimes are used in enzyme studies (Fig. 7). X;, of pyruvate values vary widely among species, and for a given M,-LDH as a function of measurement temperature, when a constant pH assay system is used at all temperatures. The variable pH assay system which simulates the temperature-dependence of intracellular pH yieids a high degree of k=, conservation. The basis of the relatively temperatureindependent K,,, of pyruvate values noted in the imidazole buffer system (Fig. 7) is easily understood. The pK of the imidazole group changes with temperature in parallel with the change in body fluid pH

0

5

i0

15

20

25

30

35

40

TEMPERATURE PC) Fig. 7. The effects of temperature on the apparent Michaelis constant (K,,) of pyruvate of M,-LDHs from vertebrates adapted to different temperatures. Measurements were made using two different pH regimes. The open symbols are values determined at a constant pH of 7.4 at all temperatures, using 66.7 mM phosphate buffer. The closed symbols are data obtained using a temperature-dependent imidazole buffer (80mM imidazole/HCI) having the pH values indicated by the dark line of the insert to Fig. 1. The vertical lines to the right of the data bracket the observed ranges of K,,, of pyruvate found under the two pH regimes for the LDHs shown in this figure and for M,-LDHs from a number of other vertebrates (see Fig. 1). The inset shows a model of part of the active site of LDH, illustrating the key histidine residue (number 195) which must be protonated to bind pyruvate, and deprotonated to bind lactate. Figure from Somero, (1981). based on data in Yancey and Somero (197Ra).

(Fig. 1). Thus, as Reeves (1977) stresses, the observed variation in pH with body temperature insures the conservation of a stable fractional dissociation state of imidazole groups, an effect which Reeves termed “alphastat” regulation (alpha is the fractional dissociation state of imid~oiej. Through this conservation of the fractional dissociation state of imidazole groups, all of the key protein traits that are dependent on either a protonated or deprotonated form of histidine tend to be held highly independent of temperature (Somero, 198 1). This is a key point, and the conservation of Kmof pyruvate is but one example of the benefits of alphastat regulation. Structural e&a

Protein subunit assembly may also be highly dependent on pH. For example, the glycolytic regulatory enzyme phosphofructokinase (PFK) undergoes shifts between tetrameric (fun~tionai) and dimeric (nonfunctional~ states in response to small changes in pH through the biological pH range (Bock and Frieden, 1976). These pH effects on the tetramerto-dimer equilibrium are extremely important in the regulation of glycolytic rate (Bock and Frieden, 1976), and in combination with temperature effects on this equilibrium, the pH effects may be instrumental in controlling large-scale metabolic transactions during entry into, and arousal from, hibernation (Hand and Somero, 1983). For these regulatory effects on PFK to be operative, the histidine residues thought to be responsible for these effects (Bock and Frieden, 1976) must be in a fractionaIly protonated state. In this state, slight shifts in pH can effect large “titrations” of enzymic activity. Thus, alphastat regulation is critical in establishing optimal structural and regulatory properties of proteins, as well as maintaining correct binding characteristics and reaction reversibility (Somero, 1981). Because fishes differ so greatly in their capacities for active swimming, as discussed earlier in comparisons of shallow-living and deep-sea fishes, it is appropriate to consider whether pH regulatory strategies differ among fishes and between different tissues of a single species. How do different species and different tissues cope with varying amounts of acidic end-product formation during intense bursts of locomotion? One important microenvironmental adaptation involved in pH regulation is buffering capacity (Burton, 1978; Castellini and Somero, 1981). In white skeletal muscle of fishes the amount of buffering capacity increases with rising potential for anaerobic glycolysis. The capacity for anaerobic glycolysis is indicated by the amount of LDH activity per gram fresh weight of skeletal muscle (Somero and Childress, 1980). A plot of LDH activity vs buffering capacity (J) as in Fig. 8 shows that sluggish deep-sea fishes with a low glycolytic potential have very low buffering capacities compared to warm-bodied tunas, which have the highest LDH activities and buffering capacities ever observed (Castellini and Somero, 1981). Shallow-living, ectothermic fishes fall in between these two other groups. These adaptive

Environmental

adaptation

of proteins

631

Shallow-water ectothermc

Deep-sea fishes 20

, , ,,,

( , (,

IO

50

100

1 200

500

1000

2m

LDH Actuty (unlk/g wet wsght) Fig. 8. The relationship

activity for white skeletal muscle of capacity is expressed as pmol of the pH of a muscle homogenate by one pH unit, between pH values of weight of muscle. LDH activities are expressed as international units of to lactate/min) per gram wet weight of muscle tissue, at a measurement of lo’-C. Data from Castellini and Somero (1981).

between

buffering

capacity

(j?) and LDH

deep-sea, shallow-living ectothermic, and warm-bodied fishes. Buffering base (NaOH) needed to change approx 6 and 7. per gram wet activity (pmol pyruvate reduced temperature

differences in buffering capacity of white muscle would seem to insure that the optimal pH for muscle function is maintained despite the production of acidic end-products during burst swimming. Before considering the molecular bases of intracellular buffering, it is necessary to examine other tissues and organs to see if the trends noted in white skeletal muscle are characteristic of other systems. In the case of fish red muscle, much lower buffering capacities are found than in white muscle, a difference which reflects the more aerobic poise of red muscle relative to white muscle (Castellini and Somero, 1981). Heart muscle of fishes also has a relatively low buffering capacity (Damm Hansen and Gesser, 1980), and these authors failed to find any correlation between buffering capacity and the abilities of hearts to function under hypercapnic acidosis. As Damm Hansen and Gesser suggest, pH regulatory mechanisms other than simple buffering must account for the differences among these hearts in abilities for function under acidotic conditions. It must be emphasized, however, that heart (and red) muscle face different conditions from those encountered by white muscle in terms of acidic end-product metabolism. Acidic end-products generated during intense periods of swimming originate primarily in the white musculature. This type of muscle, unlike heart and red swimming muscle, is very poorly vascularized, and the removal of lactate from white muscle may take several hours. Lactate entry into heart and red muscle, where lactate is burned aerobically, can be a controlled process, and build-up of acidity may not be a critical problem in these aerobically-functioning muscles. The buffering substances used to absorb protons generated during metabolism are varied, and include the bicarbonate system, phosphates, and histidinecontaining buffers. The latter family of buffers is by far the major contributor to intracellular buffering in tissues having a high buffering capacity (Burton,

1978). The histidine-containing buffers include, in addition to free histidine, dipeptides (carnosine, anserine and ophidine) and protein-bound histidyl residues. Although some fishes have extremely high concentrations of histidine in their muscles, e.g. tunas (Abe, 1981), most of the buffering capacity in highlybuffered tissues appears to be due to the histidinecontaining dipeptides (Crush, 1970; Burton, 1978). These are present in high concentrations in muscles which have an anaerobic poise, but are present in only low amounts in highly aerobic muscles like heart muscle. The role of protein-bound histidine residues in intracellular buffering is not well understood. In particular, there has been no systemic study of homologous proteins from poorly- and well-buffered tissues to see if histidine contents differ in apparently adaptive manners. To examine this possibility, I computed the average histidine contents of M,- and H,- (heart type) LDHs using literature values (Pesce et al., 1967; Eventoff et al., 1977). Most histidines are on or near the surface of the LDH subunits (Eventoff et al., 1977) and so will be able to enter into buffering reactions. As a working hypothesis, I predicted that the average histidine content of M,-LDHs, which typically function in muscles having a high potential for anaerobic glycolysis, would be greater than in H,-LDHs, which typically are the dominant LDH isozyme in aerobically-poised tissues such as heart and brain. The data in Table 3 are consistent with this hypothesis. Averaging all M,-LDHs in one group and all H,-LDHs in another, the average numbers of histidines per tetramer are 5 1 for the M, isozyme and 28 for the H, isozyme. Thus, on the average, an M,-LDH molecule can contribute about twice as much buffering capacity as an H,-LDH molecule. Moreover, the M,-LDHs from muscles that are noted for having an exceptionally high anaerobic capacity and high buffering ability, notably avian breast muscles, have the highest histidine contents. Thus,

GEORGE N. SOMERS

632

Table 3. The histidine contents of M,- and H,-LDHs of different vertebrates Histidine content Species Mammals Pig Beef Rabbit Human Birds Chicken Turkey Pheasant Duck Ostrich Rhea Amphibians Bullfrog Leopard frog Fishes Lamprey Halibut Dogfish Mean + SD (n):

M,

H,

48 33 41

28 26 27 25

63 13 61 58 60 60

30 27 28

29 31 41 49 42 51+ 13 (13)

28t_2 (8)

tion. The conserved traits we have discussed include ligand binding abilities, catalytic rate constants, and subunit assembly capacities. While the conservation of these traits often involves adaptive modifications in protein primary structure which, in turn, lead to appropriate changes in higher orders of structure and in function, many adaptations include alterations in the microenvironment in which the proteins function, i.e. in the intracellular milieu. The latter types of adaptations have received less than their due share of attention from workers concerned with molecular evolution. Moreover, studies of macromolecular function often fall short of establishing in t;itro conditions which can make manifest the actual in situ properties of the system under study. Thus. it is imperative that biochemists pay increased heed to the low molecular weight factors which are such important partners in adaptation processes, and whose abilities to affect the properties of macromolecules must be taken into account in experimentation. AcknoM,I~d~ement-Portions of these studies were supported by National Science Foundation grant PCM 80-01949.

Histidine content is expressed as the number of histidine residues per LDH tetramer. The data for pig LDHs are from Eventoff

et al. (1977); all other data are from Pesce ef al. (I 967).

while the generality of this proposed relationship between histidine contents of proteins and the needs for buffering remains to be established, the data on M,- and H,-LDHs are suggestive of a selective factor in protein evolution that, to my knowledge, has not formerly been considered. The advantages of increasing the histidine-type buffering of a tissue by elevating protein-bound histidine content would include a reduced problem for osmotic concentration. If histidine buffering is achieved via elevating concentrations of free histidine or dipeptides, a substantial increase in cellular osmolarity may result. However, if the increase in histidine buffering is achieved by increasing the histidine content of proteins, no additional osmotic problem is created. Lastly, regardless of the type of histidine buffering employed, the advantages of these types of buffers in the context of temperature-pH relationships will be the same. Since the pK of the imidazole group falls with increasing temperature at the same rate as the pH of the body fluids decreases, the buffering capacity of histidine buffers will be temperatureindependent. This is not true for bicarbonate and phosphate buffers, whose dissociation constants are affected only slightly by temperature changes. CONCLUSlONS

This review has focused on the protein traits that are highly conserved in all organisms, regardless of the physical or chemical attributes of the organisms’ environments. Comparative study of homologous forms of proteins from differently-adapted organisms seems an especially effective means for establishing “what matters most” in protein structure and func-

REFERENCES

Abe H. (1981) Determination of L-histidine-related compounds in fish muscles using high-performance liquid chromatography. Bull. Jap. Sci. Fish. 41, 139. Alexandrov V. Ya. (1977) Cells, Molecules und Tempertrture. Springer Verlag, New York. Atkinson D. E. (1969) Limitation of metabolite concentrations and the conservation of solvent capacity in the living cell. In Current Topics in Cellular Regulation (Edited by Horecker B. L. and Stadtman E. B.). pp. 29-43. Academic Press. New York. Bock P. E. and Frieden C. (1976) Phosphofructokinase-1. Mechanism of the pH-dependent inactivation and reactivation of the rabbit muscle enzyme. J. biol. Chem. 251, 563&5636. Borgmann U. and Moon T. W. (1975) A comparison of lactate dehydrogenases from an ectothermic and an endothermic animal. Gun. J. Biochrm. 53, 998~-1004. Brown J. H. and Feldmeth C. R. (1971) Evolution in constant fluctuating environments: thermal tolerances of desert pup&h (C.virinodon). Et)oiution 25, 190-398. Burton R. F. (1978) Intracellular buffering. I Resoirution Physiol. 33, 51-58. Castellini M. A. and Somero G. N. (1981) Buffering capacity of vertebrate muscle: correlations with potentials for anaerobic function. J. camp. Ph)siol. 143, 191-198. Childress J. J. and Somero G. N. (1979) Depth-related enzymic activities in muscle, brain and heart of deepliving pelagic marine teleosts. Mar. Biol. 52, 2733283. Crush K. G. (1970) Carnosine and related substances m animal tissues. Camp. Biochem. Physiol. 34, 3-30. Damm Hansen H. and Gesser H. (1980) Relation between non-bicarbonate buffer value and tolerance to cellular acidosis: a comparative study of myocardial tissue. J. up. Biol. 84, 16ll167. Eventoff W.. Rossmann M. G., Taylor S. S.. Torff H.-J.. Meyer H., Keil W. and Kiltz H.-H. (1977) Structural adaptations of lactate dehydrogenase isozymes. Proc natn. Acad. Sci. (U.S.A.) 74. 2677-2681. Everse J. and Kaplan N. d. (1973) Lactate dehydrogenase: structure and function. A&. En;ym. 37, 61-133. Fersht A. (1977) Enzyme Structure and Mechanrsm. p. 256.

Freeman, San Francisco

Environmental

adaptation

Graves J. E. and Somero G. N. (1982) Electrophoretic and functional enzymic evolution in four species of Eastern Pacific barracudas from different thermal environments. Evolution 36, 97- 106. Greaney G. S. and Somero G. N. (1980) Contributions of binding and catalytic rate constants to evolutionary modifications in K, of NADH for muscle-type (M4) lactate dehydrogenases. J. camp. Physiol. 137, 115-121. Hand S. C. and Somero G. N. (1983) Phosphofructokinase of the hibernator, Citellus bee&i: temperature and pH regulation of activity uia influences on the tetramer-dimer equilibrium. Physiol. Zool. (in press). Johnston I. A. and Walesby N. J. (1977) Molecular mechanisms of temperature adaptation in fish myofibrillar adenosine triphosphatases. J. camp. Physiol. 119, 1955206. Low P. S. and Somero G. N. (1976) Adaptation of muscle pyruvate kinases to environmental temperatures and pressures. J. exp. Zool. 198, l-12. Low P. S.. Bada J. L. and Somero G. N. (1973) Temperature adaptation of enzymes: roles of the free energy, the enthalpy and the entropy of activation. Proc. natn. Acad. Sci. (U.S.A.) 70, 43&432. Norris K. S. (1953) The ecology of the desert iguana Dipsosuurus dorsulis. Ecology 34, 265-287. Nuccitelli R. and Heiole J. M. (1982) Summarv of the evidence and discussion concerning the involvement of pH, in the control of cellular functions. In Inrracellular pH: Its Measurement, Regulation. and Utilization in Cellular Function (Edited by Nuccitelli R. and Deamer D. W.), pp. 567.-586. Allan R. Liss, New York. Pang P. K. T., Griffith R. W. and Atz J. W. (1977) Osmoregulation in elasmobranchs. Am. Zool. 17, 365-377. Pesce A.. Fondy T. P., Stolzenbach F., Castillo F. and Kaplan N. 0. (1967) The comparative enzymology of lactic dehydrogenase--III. Properties of the H, and M, enzymes from a number of vertebrates. J. biol. Chem. 242, 2151-2167. Reeves R. B. (1977) The interaction of body temperature and acid-base balance in ectothermic vertebrates. A. Rec. Phjjsiol. 39, 559-586. Rich S. A. and Estes J. E. (1976) Detection of conformational changes in actin by proteolytic digestion: evidence for a new monomeric species. J. molec. Biol. 104, 777-792. Siebenaller J. F. and Somero G. N. (1978) Pressure-adaptive differences in lactate dehydrogenases of congeneric fishes living at different depths. Science N.Y. 201, 255-257. Siebenaller J. F. and Somero G. N. (1979) Pressure-adaptive differences in the binding and catalytic properties of muscle-type (M4) lactate dehydrogenases of shallow- and deep-living marine fishes. J. camp. Phvsiol. 129, 295-300. Siebenaller J. F., Somero G. N. and Haedrich R. L. (1982) Biochemical characteristics of macrourid fishes differing

of proteins

633

in their depths of distribution. Biol. Bull. mar. biol. Lab., Wood.s Hole (in press). Smith K. L. (1978)Metabolism of the abyssopelagic rattail Coryphaenoides armatus measured in situ. Nature, Lond. 214, 362-364. Smith K. L. and Hessler R. R. (1974) Respiration of benthopelagic fishes: in situ measurements at 1230 meters. Science N. Y. 184, 72-73. Somero G. N. (1978) Temperature adaptation of enzymes: Biological optimization through structure-function compromises. A. Rea. Ecol. Svst. 9, l-29. Somero G. N. (1981) pH-temperature interactions on proteins: principles of optimal pH and buffer system design. Mur. Biol. Letf. 2, 163-l 78. Somero G. N. and Childress J. J. (1980) A violation of the metabolism-size scaling paradigm: activities of glycolytic enzymes in muscle increase in larger-size fish. Physiol. 2001. 53, 322-337. Somero G. N. and Siebenaller J. F. (1979) Inefficient lactate dehydrogenases of deep-sea fishes. Nature, Lond. 282, 100-102. Somero G. N., Siebenaller J. F. and Hochachka P. W. (1983) Biochemical and physiological adaptations of deep-sea animals. In The Sea (Edited by Rowe G. T.), Vol. 8. Wiley-Interscience, New York. Sullivan K. M. and Somero G. N. (1980) Enzyme activities of fish skeletal muscle and brain as influenced by depth of occurrence and habits of feeding and locomotion. Mar. Biol. 60, 91-99. Swezey R. R. and Somero G. N. (1982) Polymerization thermodynamics and structural stabilities of skeletal muscle actins from vertebrates adapted to different temperatures and hydrostatic pressures. Biochemisfry (in press). Walsh P. J. and Somero G. N. (1982) Interactions among pyruvate concentration, pH. and K, of pyruvate in determining in ciro Q10 values of the lactate dehydrogenase reaction. Can. J. Zool. 60, 1293-1299. Yancey P. H. and Somero G. N. (1978a) Temperature dependence of intracellular pH: Its role in the conservation of pyruvate K, values of vertebrates lactate dehydrogenases. J. Comp. Physiol. 125, 129-134. Yancey P. H. and Somero G. N. (1978b) Urea-requiring lactate dehydrogenases of marine elasmobranch fishes. J. camp. Phvsiol. 125, 135- 14 1, Yancey P. H. and Somero G. N. (1979) Counteraction of urea destabilization of protein structure by methylamine osmoregulatory solutes of elasmobranchs. Biochem. J. 183, 317-323. Yancey P. H. and Somero G. N. (1980) Methylamine osmoregulatory solutes of elasmobranch fishes counteract urea inhibition of enzymes. J. exp. 2001. 212, 205-213. Yancey P. H., Clark M. E., Hand S. C., Bowlus R. D. and Somero G. N. (1982) Living with water stress: the evolution of osmolyte systems. Science, N. Y. (in press).