The Molecular Ecology of Fundulus heteroclitus Hemoglobin-Oxygen Affinity 1

Transcription

1 AMER. ZOOL., 26: (1986) The Molecular Ecology of Fundulus heteroclitus Hemoglobin-Oxygen Affinity 1 DENNIS A. POWERS, PAULA M. DALESSIO, 2 EDWARD LEE, AND LEONARD DIMICHELE Department of Biology and The Chesapeake Bay Institute, The Johns Hopkins University, Baltimore, Maryland SYNOPSIS. Animals faced with environmental perturbations must adapt or face extinction. The respiratory complex, specifically hemoglobins, is perhaps the best system to study such adaptation because it exists at the organism environment interface. Fish are particularly useful models because they respond directly to such environmental variables as temperature, oxygen, ph, carbon dioxide, and salinity. Our experiments have addressed the molecular, cellular, and physiological mechanisms employed by fish to maintain respiratory homeostasis in the wake of changing temperature and oxygen. Immediate, intermediate, and long-term adaptation can only be understood when the hemoglobin's ligand binding properties and the cellular and hormonal regulation of various ligands are considered simultaneously. We describe a detailed thermodynamic model for the binding of oxygen, protons, and organic phosphates to hemoglobin; discuss the role of multiple hemoglobins; and present evidence for physiological and genetic regulation of hemoglobin's major allosteric modifiers in response to environmental stress in the mummichog, Fundulus heteroclitus. INTRODUCTION Comparative biochemists and physiologists have found that the blood-oxygen affinities of various fish species are compatible with the physical and chemical parameters of their environments. For example, fish that live in low oxygen environments have high oxygen affinities while those that live in high oxygen environments have lower oxygen affinities. Moreover, fish that live in environments where physical parameters periodically change have the necessary molecular machinery required for adaptation. This machinery includes species-specific hemoglobins and/ or the regulation of various modifier ligands. The oxygen transport system of the fish Fundulus heteroclitus is an ideal model in which to study these adaptative mechanisms. The primary function of hemoglobin is to carry oxygen from the "organism-environment interface" (e.g., the gills, lungs, etc.) to the respiring tissues. This physio- 1 From the Symposium on The Biology of Fundulus heteroclitus presented at the Annual Meeting of the American Society of Zoologists, December 1983, at Philadelphia, Pennsylvania. 8 Paula M. Dalessio's home institution is: Department of Biology, The George Washington University, Washington, D.C. 235 logical function depends on the hemoglobin's ability to form a reversible complex between oxygen and the ferrous iron in the hematoporphyrin. Under constant physical conditions and in the absence of modifying molecules, the affinity of the hemoglobin for oxygen depends primarily on the hydrophobic nature of the heme pocket and the roles of specific amino acid residues which may directly or indirectly affect hemoglobin-oxygen (Hb-O 2 ) affinity. Since these amino acids are coded for by the globin genes, this intrinsic Hb-O 2 affinity is genetically determined. While intrinsic Hb-O 2 affinity is genetically determined, physiological plasticity is provided by regulating intracellular concentrations of allosteric modifiers, such as HCO 3 ", Cl~, H +, and organic phosphates. The mathematical concept that describes the interactions between hemoglobin and its various ligands was formalized by Wyman (1948, 1964). THE LINKAGE BETWEEN HEMOGLOBIN, OXYGEN, PROTONS, AND ORGANIC PHOSPHATE When hemoglobin (M) binds four oxygen molecules (X) and one molecule of an organic phosphate (D) (Scheme 1), the simultaneous equilibria can be depicted by the following linkage scheme:

2 236 D. A. POWERS ET AL.»K «k "k >*k 4»k M» MX MX, MX, MX, D K D K, K IX K 5x j K 4X MD- 'MXD MX,D- *MX,D- 'MX.D XU SXI. 3x1. 4x1. IVQ IVp KQ Kp (Scheme 1) where the k's are the microconstants for individual oxygenation steps, and the K's are macroconstants. The presuperscripts on all constants indicate the number and type of ligands being bound; while the subscripts indicate the number and type of ligands already bound to the hemoglobin. If temperature, ph, etc. are constant, and the hemoglobin does not dissociate or polymerize, Scheme 1 can be described by 9 of the 13 equilibrium constants. However, when proton concentration is allowed to vary, it affects oxygen binding as well as the binding of organic phosphate. At relatively high ph (e.g., ph 10) there is little or no Bohr or organic phosphate effects on Hb-O 2 affinity; conversely, at low ph there can be a reversal or inhibition of both the Bohr and organic phosphate effects on Hb-O 2 affinity. Thus, the Bohr and organic phosphate effects are most pronounced over the intermediate, physiologically important ph range. Assuming that the protons that affect the binding of organic phosphate are between zero and z for deoxyhemoglobin and between zero and r for oxyhemoglobin, the wiacro-interaction between hemoglobin, oxygen, organic phosphate, and protons can be described by Figure 1. When the hemoglobin is at the extremes of oxygenation (Fig. 1), the apparent binding of organic phosphate to oxy- or deoxyhemoglobin can be measured directly (Powers et al., 1981). The macroconstants for the other faces of Figure 1 are obtainable from oxygen binding measurements carried out as a function of ph and organic phosphate concentrations. For deoxyhemoglobin (M) the apparent binding constant for organic phosphate ( D Kgg xy ) is defined by: D K(1 + H K D [H] + * H K D [H]' +... * H K D [H]') (1 + H K[H] + 2H K[Hf + SH K[H] mh K[H] ) (eq. 1) and the equation for oxyhemoglobin is: "Kg* _ K 4x (l + H K 4xD [H] + '"K 4xD [H]' +... rh K 4xD [H]Q (1 + H K 4x [H] + 2H K[H]! +... " H K 4x [H]") (eq. 2) The derivation for these equations has been published by Hobish and Powers (submitted). The various D K a PP values can be determined for oxy- and deoxyhemoglobin at defined ph values by the method of Powers et al. (1981). These data cannot be fitted to eqs. 1 and 2 easily because there are an infinite number of possible constants. Since we know only a few protons are involved (Greaney et al., 1980a), the data can be fitted by a non-linear least squares analysis to simplified versions of these equations. These (eqs. 3 and 4) are functionally similar to a Hill equation. _ D K(1 + ZH K D [H] Z ) (1 + MH K[H] m ) (eq- 3) -K 4X [H]") Examination of Figure 1 indicates several possibilities for Hb-O 2 affinity regulation. For example, it can be differentially affected by changing intracellular concentrations of ligands, modifying one or more of the ligand-protein binding constants, a combination of these, or other strategies. Moreover, the dissociation of all of the complexes (Fig. 1) is increased by increasing temperature; providing an additional mode of physiological adjustment to environmental temperature. MULTIPLE FISH HEMOGLOBINS Possessing a number of hemoglobins, some of which have unique functional characteristics, is one strategy for adapting to a changing environment. While the presence of multiple hemoglobins in lower vertebrates is a well-established phenomenon, there are discrepancies regarding the actual number, proportion, and structural and functional properties of species-spe-

3 MOLECULAR ECOLOGY OF FUNDULUS HEMOGLOBIN 237 -±- MDHz* MX«DH,-*- FIG. 1. A generalized schematic of the linkage equilibria relating the binding of protons (H), organic phosphate (D), and oxygen (X) to hemoglobin tetramers (M). The macrobinding constants (K) are those described in the text with the presuperscripts indicating the number and kind of ligands being bound and the postsubscripts, the numbers and kind of ligands bound to the macromolecules (M). cific isohemoglobins (reviewed by Riggs, 1970; Fyhn and Sullivan, 1975; Powers, 1980). Such differences can arise from thermoacclimatory variation, developmental changes, polymerization, dissociation, and susceptibility to autooxidation. One of the most interesting sources of electrophoretic variation arises from genetically unique globin chains that combine to give a series of unique hemoglobin tetramers. An analysis of the subunit structure of multiple fish hemoglobins has only rarely been attempted. In those instances, two adaptive strategies have become evident. First, there are multiple hemoglobins that are structurally different but functionally equivalent. Second, there are structurally different hemoglobins, some of which (but not all) are functionally unique (see Powers, 1980 and references therein). between ph 5-9. These fish hemoglobins show very little tetramer-dimer dissociation in low salt (~10~ 8 M heme) and only slightly more in 1.0 M NaCl (~10" 7 M heme). Moreover, the presence of up to 4 M urea has little effect on the dissociation. Each hemoglobin is composed of two a and two /3 subunits. There are four different globin chains: a 3, a b, /3 a and j3 b. The Hbl is a homotetramer composed of two a b and two /3 b chains; HblV is a homotetramer consisting of two a* and two fit* subunits. Isohemoglobins II and III are heterotetramers consisting of all four chains. A model of these combinations can be seen in Figure 2. While the amino acid compositions of the chains are significantly different the end-groups of homologous chains (i.e., a*:a h and /3 a :/3 b ) are identical (Mied and Powers, 1978). The NH 2 -termini of the a chains are blocked with an acetyl group, while the /3 chains have free NH 2 -terminal Val. The COOH-termini are -Tyr-Arg and -Tyr-His for the a and /3 chains, respectively. FUNCTION OF PURIFIED FUNDULUS HEMOGLOBINS The effect of ph on the P 50 of the unfractionated hemoglobin and of the isolated components I, II, III and IV is illustrated in Figure 3. All of the components are similarly affected by ph. The data in Figure 3 suggest that components II and IV have slightly higher oxygen affinity at ph values greater than 7.5. Although subtle differences may exist among the components, these differences in general appear to lie within the range of experimental variation. Finally, oxygen equilibria studies done at high ph as a function of temperature indicate significant thermal sensitivity of the hemoglobin components. The enthalpy calculated from van't Hoff plot is AH = ± 0.5 kcal/mole. STRUCTURAL ASPECTS OF FUNDULUS HEMOGLOBINS One of the better analyzed hemoglobin systems is that of Fundulus heteroclitus (Mied and Powers, 1978). Fundulus has four isohemoglobins, each a tetramer of about 64,000 daltons (Fig. 2, Table 1). They have different isoelectric points (Table 1); all are THE ORGANIC PHOSPHATE EFFECT The major organic phosphate in fish erythrocytes is either adenosine triphosphate (ATP) or guanosine triphosphate (GTP), while in mammalian red blood cells it is 2,3-diphosphoglycerate (2,3 DPG). The effects of ATP and GTP on the oxygen binding properties offish hemoglobins are

4 238 D. A. POWERS ET AL. f 2 a 2 p? b Hb HI oa a, P, PI HbET "CD i a N O b O (VI a az R a b o b I I I 1 1 ijsjsaj! i i i '1 Q b -8, b 2 Hb E O CVI «2 Pi FIG. 2. A model for the subunit compositions of F. heteroclitus hemoglobins. The four different dimers that can be formed, a"/3", a*/9 b, a b 0*, and a'/s 1 ", have been entered at the left as a,(3, dimers and above each column as a,f} 2 dimers. For association into tetramers, the oy? 2 dimers have been written 8 2 a 2 to more accurately depict the actual spatial relationship to the 0,18, dimer that exists in the tetrameric hemoglobin molecule and to emphasize the a,/3 2 and a^, contacts between dimers. To aid in visualizing the association-dissociation process, the integrity of the dimers has been maintained within each tetramer. The upper portion of the matrix, which contained tetramers identical with those in the lower half, has been omitted for the sake of simplicity. The four tetramers shown with dotted lines between the subunits contain double a-b type contacts (a, 8 2 and Qfj/9, contacts) and/or four different polypeptide chains, and therefore are assumed to be unstable (see "Discussion"). The six remaining stable hemoglobins constitute the four electrophoretically distinguishable components, I, II, III, and IV, of F. heteroclitus hemoglobin. Hbl ob» 2 b similar to those of 2,3 DPG on mammalian hemoglobins. Furthermore, the influence of ATP on both the Bohr effect and on the ph dependence of subunit cooperativity indicates that the reactions of many fish hemoglobins with oxygen, organic phosphates, and protons are functionally linked. In addition to its allosteric effect on oxygen affinity, organic phosphate has an important influence on the Donnan distribution of protons across the erythrocyte membrane (Wood andjohansen, 1972). Several workers have demonstrated that adaptive changes in fish blood hemoglobin-oxygen

5 MOLECULAR ECOLOGY OF FUNDULUS HEMOGLOBIN 239 TABLE 1. Selected physiochemical properties of the hemoglobins from Fundulus heteroclitus. Parameter Proportions (%)* Isoelectric point S»ow b K.i'lmMheme) K 4, d (mmheme) M r '(0.1 M NaCl) Hb I 10.2 ± ± < ,000 Electrophoretic component Hb II Hb III 37.8 ± ± , ± ± ,000 Hb IV 11.5 ± ± < ,000 * Determined by gel scanning of 16 gels. b At several protein concentrations between 1 mg/ml and 10 mg/ml in 0.1 M phosphate, 0.1 M NaCl and ph 7.0. c Tetramer-dimer dissociation constants in mm heme determined in 0.1 M phosphate buffer, ph 7.0, 0.1 Af NaCl and 1 mm EDTA. d Same as c but performed in 1.0 M NaCl at three different protein concentrations and three rotor speeds. e Determined by gel filtration (see: Mied and Powers, 1978). affinity during acclimation to low oxygen or increased temperature are inversely related to changes in red blood cell organic phosphate concentrations (Wood and Johansen, 1972; Wood et al., 1975; Greaney and Powers, 1978). ATP is the major allosteric modifier of Fundulus hemoglobins. ATP binds to F. heteroclitus deoxyhemoglobin with a stoichiometry of one major binding site per hemoglobin tetramer. The affinity of the deoxyhemoglobins for ATP is elevated as temperature decreases. At 11.5 C, its dissociation constant is 4.0 ± 1.2 x 10~ 6 M. However, at 22.5 C, it increases to 1.18 ± 0.8 x 10" 5 M. The van't Hoff enthalpy (AH), calculated from the equilibrium association constants, is approximately 15.0 kcal/mole of tetramer. As illustrated in Figure 3, the presence of ATP dramatically decreases the hemoglobin-oxygen affinity of Fundulus hemoglobins and the effect is amplified at low ph. A number of studies have shown that the NH 2 -termini of the 0 chains are involved in ionic interactions with the organic phosphate. Organic phosphate binds to certain basic residues in the central cavity of the human deoxytetramer (Arnone, 1972). Presumably, salt bridges are formed with the NH? -termini of the 0 chains (Val, 01), and the imidazole side chains (His: 02, 0143) of both 0 chains, as well as the e-amino group of Lys, /382, from one of the 0 chains. The homologous residues in Fundulus and most other fish hemoglobins are: Val (01), Glu (/32), Lys (082), and Arg (0143) (reviewed by Powers, 1980). Assuming that the same amino acid residues are involved in the binding of ATP by Fundulus hemoglobins, the major residues that would be titrated over a physiological ph range would be the amino termini of the 0 chains. ADAPTATIONS TO TEMPERATURE CHANGES As temperature increases, the availability of oxygen in water decreases because the solubility of oxygen decreases with increasing temperature and elevated biological activity may reduce dissolved oxygen to below saturation. Consequently, at higher temperatures fish require more oxygen, but less is available. Fish respond to this dilemma by increasing ventilation volume, heart rate, and oxygen carrying capacity of the blood. The effect of temperature on the oxygen equilibria of fish hemoglobin has been reviewed (Riggs, 1970; Johansen and Lenfant, 1972; Johansen and Weber, 1976). Although the number of species examined is limited, the functional properties of teleost hemoglobins can be divided into three major categories as suggested by Weber etal. (1976). Class I contains species with one or more hemoglobins all of which are sensitive to both temperature and ph (Gillen and Riggs, 1971; Gillen and Riggs, 1972; Bonaventura et al, 1974; Weber, 1975; Mied and Powers, 1978). Class II has species with multiple components, some of which are functionally similar to the Class

6 240 D. A. POWERS ET AL. FIG. 3. A plot of variation of log P 50 with ph for F. heteroclitus hemoglobins. Data were obtained using the Aminco Hem-O-Scan oxygen dissociation analyzer. Open symbols represent stripped Hb; solid symbols represent Hb in the presence of 10 ATP/Hb. O and, unfractionated Hb; V, HB I; 0. Hb II; O, HB III; A, Hb IV. I hemoglobins, while other components are not strongly affected by temperature and ph (Hashimoto et al., 1963; Binotti et al., 1971; Powers and Edmundson, 1972a; Powers, 1972, 1974; Wyman et al., 1977). Class III fishes have hemoglobins that are ph sensitive, but temperature insensitive (Rossi-Fanelli and Antonini, 1961; Anderson et al., 1973). The overall enthalpy of bood oxygen equilibria represents both the intrinsic heat of oxygen binding to hemoglobin and contributions due to other ligand-linked processes (see earlier discussion). One such process is associated with proton equilibria of amino acid side chains (i.e., the Root and Bohr effects). At alkaline ph (i.e., ph 9-10) the Root and Bohr effects are essentially inoperative. Thus, data collected on "stripped" hemoglobins at alkaline ph, over a range of temperatures, provide information primarily on the intrinsic thermodynamic parameters of the hemoglobins per se. Under such conditions, enthalpies are very similar for a number of fish hemoglobins from a variety of thermal environments (Powers et al., 19796). Powers et al. (1979a) have clearly shown that if evolution has favored a decrease in temperature sensitivity of blood-oxygen affinity in fish experiencing large fluctuations, as suggested by Johansen and Lenfant (1972), then it must be primarily associated with the regulation of intracellular ph, the levels and types of organic phosphates, and other ligand-linked phenomena rather than selection of hemoglobins per se. Thus, blood-oxygen affinity adaptation to different thermal regimes must be primarily at the erythrocyte level rather than in the intrinsic enthalpy of hemoglobin oxygen binding. Greaney and Powers (1977, 1978) demonstrated that Fundulus heteroclitus adapt to thermal changes by changing ATP concentrations within erythrocytes. This prompted us to ask if this response was triggered by reduced oxygen (due to the increased temperature), increased temperature, or both of these variables. Therefore, fish were acclimated to various temperatures (10 C, 22 C, and 30 C), but with oxygen concentration maintained constant (about 7 ppm) at each temperature. These animals also decreased their erythrocyte ATP with increased temperature. Moreover, fish acclimated to 10 C but in air saturated water (12.5 ppm) showed the same red blood cell ATP levels as those maintained at 10 C but with 7 ppm oxygen. These data demonstrate that the ATP response was elicited by increased temperature alone and was independent of dissolved oxygen levels in the range ppm. However, studies on other fishes suggest the strategy employed by Fundulus is not universal (Powers, unpublished). STRATEGIES OF ADAPTING TO HYPOXIA There are numerous strategies by which fish are able to maintain respiratory homeostasis when environmental oxygen is reduced. Perhaps the most common is to seek out a more favorable environment. Species that remain in oxygen poor environments have: (1) immediate, (2) intermediate, and (3) long-term adaptation strategies. Immediate response to hypoxia The immediate response to an oxygen poor environment is an increase in heart rate and ventilation volume (Prosser, 1973).

7 MOLECULAR ECOLOGY OF FUNDULUS HEMOGLOBIN E 15 o 10 I0" 1,-7 10 Epinephrine (M) IO"! FIG. 4. The effect of varying concentrations of epinephrine on the P 50 values of washed erythrocytes from Fundulus heteroclitus. Bars indicate ± standard error of the mean (SEM). In addition, fish will often "gulp" air and/ or utilize the water at the air-interface which has the greatest oxygen content. Furthermore, fish are able to rapidly increase or decrease blood oxygen affinity by employing hormonal and/or other factors in the serum. Circulating catecholamine concentrations increase markedly in fish subjected to stress such as hypoxia, surgery, physical disturbance, and confinement (Nakano and Tomlinson, 1967; Butlers al., 1978,1979; Wahlqvist and Nilsson, 1980; LeBras, 1982). Nikinmaa (1982, 1983) assessed the role of epinephrine in controlling the blood-oxygen affinity in rainbow trout. Both in vivo and in vitro, epinephrine increased blood-oxygen affinity and erythrocyte volume, decreased mean erythrocyte hemoglobin and ATP concentrations, and decreased the extracellular to intracellular proton gradient. All of which are similar responses seen in hypoxic trout. Fundulus heteroclitus whole blood has a much higher oxygen affinity than washed Fundulus erythrocytes a phenomenon also observed in trout and flounder (Soivio et al, 1980; Nikinmaa, 1983; Dalessio et al., 1984,1985 manuscript in preparation). Since removal of catecholamines from the 120 FIG. 5. The effect of epinephrine on the P M values of washed erythrocytes obtained from Fundulus heteroclitus. The solid circles ( ) represent P 50 values of untreated erythrocytes. The open triangles (A) represent P 50 values of epinephrine treated erythrocytes. medium of washed erythrocytes results in a general instability of these cells (Bourne and Cossins, 1982), then perhaps the decreased oxygen affinity of washed fish erythrocytes is also due to the absence of catecholamines. To test this hypothesis, epinephrine was added to washed Fundulus erythrocytes and the oxygen affinity of red cells was measured (Fig. 4). The P 50 of washed cells decreased in a dose-dependent fashion in concentrations ranging from 10~ 7 M to 10" 6 M epinephrine. Further decreases in P 50 did not occur with epinephrine in concentrations exceeding 10~ 6 M. The P 50 of washed cells decreased within 10 min following addition of epinephrine to the medium and remained constant for at least 110 min (Fig. 5). Although the P 50 of washed cells decreased from 18.0 mm Hg to 7.5 mm Hg following treatment with \0~ 6 M epinephrine, the P 50 value of whole blood, 4.5 mm Hg, was never attained. In vivo, epinephrine is generally degraded very rapidly. It is not known if washed erythrocytes maintained in an artificial medium are capable of degrading epinephrine and hence, if the decreased P 50 seen over 2 hr is due to the continued presence of epinephrine. Washed cells incubated for 5 min in medium containing epi-

8 242 D. A. POWERS ET AL. In vitro and in vivo erythrocyte-oxygen affin- TABLE 2. ities. Treatment In vitro (washed erythrocytes)" Control Epinephrine treated Control Epinephrine treated (washed) Control Epinephrine treated (washed, 4 hr later) In vivo (whole blood) 6 Non-injected control Bull saline injected control Epinephrine 10" 6 M Epinephrine 10" s M Epinephrine 10~*Af P,o (mm Hg) 13.4 ± ± ± ± ± ± ± ± ± ± ± 0.3 Following P 50 measurements of washed cells in presence of epinephrine, cells were washed with and resuspended in epinephrine free medium, and the P 50 followed for 4 hr. Errors are reported as ±SEM. b Each fish was injected with 0.01 ml/g body wt of one of the above substances. Each group consisted of 10 fish. Errors are reported as ±SEM. nephrine and then washed with and resuspended in fresh medium without epinephrine have a decreased P 50 that was maintained over a 4 hr period (Table 2). Stadel etal. (1983) have shown that washed frog erythrocytes incubated for 3 hr in medium containing the /?-adrenergic agonist, isoproterenol, results in desensitization of the /3-adrenergic receptor-coupled adenylate cyclase in the plasma membrane. Furthermore, they show that prolonged stimulation of adenylate cyclase by isoproterenol results in a 50-60% decrease in receptor sites. Perhaps epinephrine causes a shift in erythrocyte metabolism resulting in a new steady-state that is temporarily independent of the presence of the hormone. At low concentrations (10~ 9 M and 10~ 8 M) epinephrine has little effect on the rate of oxygen consumption of washed erythrocytes. However, 5 min after the addition of 10~ 6 M epinephrine, oxygen consumption increases markedly and remains elevated for 5-10 min (Fig. 6). Oxygen consumption decreases slightly below basal levels for a short time, after which it returns to basal levels for an indefinite period. However, it is not known whether this shift in metabolism lasts for 5 min after which it returns to a pre-epinephrine state or if it reaches a new steadystate. In Fundulus injected with various concentrations of epinephrine, P 50 values of blood increase slightly; this response does not appear to be dose-dependent (Table 2). The effects of epinephrine on blood in vivo appear to be opposite from those of washed erythrocytes. Epinephrine added to whole blood in vitro yielded P 50 values similar to those obtained in epinephrine injected fish. Intermediate response to hypoxia Intermediate responses are generally activated after several hours of hypoxic conditions and last many days or until longterm compensation is achieved. These responses include increasing hematocrit by retaining serum in muscle tissues (Cameron, 1970), or releasing stored erythrocytes from the spleen, reducing intraerythrocyte organophosphate concentrations (Wood and Johansen, 1972; Greaney and Powers, 1977, 1978), changing ph, and changing the ionic micro-environments of erythrocytes (Houston and Mearow, 1979). Such attempts to maintain oxygen delivery to respiring tissues is usually accompanied by large fluctuations in various enzyme activities during metabolic readjustment (Greaney etal, 19806). During the intermediate responses described above, fish synthesize new erythrocytes so that the total oxygen carrying capacity of the blood is increased. Eventually, a new steady-state between new cells, enzyme levels, organophosphates and hemoglobin function is achieved. This level and the balance is different for each fish species. When exposed to low oxygen environments for several days, both mammals and fish (Krogh and Leitch, 1919; Johansen, 1970) increase the oxygen carrying capacity of their blood. This is accomplished by a number of strategies. The common mechanisms between mammals and fish are: increased hematocrit, increased hemoglobin content, and increased blood buffering capacity. On the

9 MOLECULAR ECOLOGY OF FUNDULUS HEMOGLOBIN Control A Epinephrine (IO" 6 M) il 2 E MINUTES FIG. 6. The effect of epinephrine on the oxygen consumption (Vo 2 ) of washed erythrocytes obtained from Fundulus heteroclitus. The solid circles ( ) represent Vo 2 values of untreated erythrocytes. The open triangles (A) represent Vo 2 values of treated erythrocytes. The arrow indicates the time at which epinephrine was added to erythrocytes. other hand, mammals and fish differ considerably in other aspects of their response to low oxygen. Mammals decrease the affinity of their hemoglobin for oxygen. Hypoxic fish, in contrast, increase hemoglobin-oxygen affinity as characterized by a decrease in the P 50 of their corresponding oxygen-saturation curves (Wood and Johansen, 1972). For example, mammals acclimated to high altitudes increase levels of 2,3 DPG in their erythrocytes (Lenfant et al., 1968). Since mammals typically live in an oxygen-rich environment, it has been suggested that this response might be best suited to the more commonly encountered forms of low altitude hypoxia, such as chronic hypoxemia (Gerlach and Duhm, 1972; Eaton, 1974). The response of waterbreathing vertebrates to chronic hypoxia is quite different (Powers, 1974). Eels have been shown to decrease erythrocyte ATP and increase Hb-O 2 affinity when acclimated to low environmental oxygen (Wood and Johansen, 1972). In addition, Fundulus heteroclitus acclimated to hypoxic conditions lower red cell ATP by as much as 40% and increase the percent hematocrit which presumably increases the oxygen carrying capacity of the blood (Fig. 7) (Greaney and Powers, 1978). Moreover, there were concomitant increases in serum lactate and a decrease in blood ph. We predicted that if the control mechanism was directed at the erythrocyte level, then fish red blood cells should decrease ATP, in vitro, under anoxic conditions. Consistent with our in vivo observation, we found that anaerobic F. heteroclitus red cells significantly lowered their ATP levels in vitro (Greaney and Powers, 1977). Since

10 244 D. A. POWERS ET AL. acclimation time (days) FIG. 7. Fundulus heteroclitus hypoxia studies: A. Time course of % blood hematocrit of hypoxic ( ) and normoxic ( ) fish. B. Time course of acclimation of F. heteroclitus to hypoxic ( ) and normoxic (O) conditions at 22 C. Dissolved oxygen values were parts per million (ppm) for hypoxic and ppm for control fish. Red cell ATP for this experiment and all others described in this paper were determined using the firefly luciferase assay. All points represent averages of 6-7 fish. Bars indicate ±standard error of the mean (SEM) (from Greaney and Powers, 1978). the nucleated erythrocytes of fish possess mitochondria, we reasoned that this response may be mediated by way of a decrease in oxidative phosphorylation. This hypothesis was supported when aerated cells were incubated in the presence of low concentrations of cyanide. This inhibitor of aerobic respiration reduced intracellular ATP to levels similar to those found in the anoxic cells (Greaney and Powers, 1977, 1978). These data have been confirmed by red cell oxygen consumption studies (Powers, 1983). Thus it seems that one reason fish have retained a functional oxygen-consuming electron transport system in the mitochondria of their red blood cells is to control the levels of hemoglobin allosteric effector. PHYSIOLOGICAL CORRELATION BETWEEN LACTATE DEHYDROGENASE GENOTYPE AND HEMOGLOBIN FUNCTION IN FUNDULUS During our hypoxic studies, we observed significant variability in red cell ATP concentrations among individual fish. We questioned whether this variability could be the result of a genetic component. Consequently, we analyzed F. heteroclitus for red cell ATP, concomitantly screening for a series of enzyme variants. Not only were ATP levels under genetic control, but they were also highly correlated with lactate dehydrogenase (LDH) phenotype (Powers etal., 1979c). There are three major electrophoretically distinguishable phenotypes of the heart-type LDH (the LDH-B locus) in F. heteroclitus (Place and Powers, 1978). The polymorphism is due to two co-dominant alleles which exhibit a dramatic north-south cline in gene frequency along the Atlantic coast of the U.S. (Powers and Powers, 1975; Powers and Place, 1978). The phenotypes are LDH-B a B a, LDH-B a B b and LDH-B b B b. The B locus is the only LDH expressed in the red blood cells of this fish. When intraerythrocyte ATP levels were compared for individuals of different LDH-B phenotypes, there was a significant association. As the intraerythrocyte ATP/Hb ratio is different for each of the LDH-B phenotypes, we expected to find differences in hemoglobin-oxygen affinity. In agreement with this expectation, LDH-B a B a homozygotes with the lowest ATP/Hb ratio had the highest oxygen affinity and LDH-B b B b homozygotes with the highest ATP/Hb value have the lowest oxygen affinity (Powers et al., 1979c). This was found to be an important factor in differential developmental rates and swimming abilities (DiMichele and Powers, 1982a, b). Since the differential developmental rates of LDH-B genotypes are being addressed elsewhere in this volume (DiMichele et al., 1986 [this vol.]), we shall restrict our comments to swimming performance. Our analysis of purified LDH-B allelic isozymes (Place and Powers, 1979, 1984a, b), indicated that the greatest catalytic differences between LDH-B a B a and LDH-B b B b existed at low temperature (10 C) while no significant difference exists at 25 C. Thus,

11 MOLECULAR ECOLOGY OF FUNDULUS HEMOGLOBIN 245 if the LDH-B enzyme has a direct influence on erythrocyte ATP concentration, differences in ATP and blood-oxygen affinity should only exist at acclimation temperatures below 25 C. Furthermore, since organic phosphate amplifies the Bohr effect of Fundulus heteroclitus hemoglobins (see Fig. 3), these phenomena should be exaggerated at low blood ph values, like those produced during swimming performance experiments. DiMichele and Powers (19826) have reported that swimming performance is highly correlated with genetic variation at the LDH-B locus for Fundulus acclimated to 10 C while no such differences exist for 25 C acclimated fish. After an acclimation period, fish of each of the two homozygous LDH-B phenotypes were swum to exhaustion in a closed water tunnel. The exhausted fish were sacrificed immediately and appropriate biochemical and physiological parameters determined. Among resting fish acclimated to 10 C, hematocrit, blood ph, blood-oxygen affinity, serum lactate, liver lactate, and muscle lactate were not significantly different between the two LDH-B homozygous phenotypes (DiMichele and Powers, 19826). Exercising fish, acclimated to 10 C, to the point of fatigue caused a significant change in all of these parameters. The LDH-B b B b phenotype was able to sustain a swimming speed 20% higher than that of LDH-B a B a phenotype. Blood-oxygen affinity, serum lactate and muscle lactate also differed between the phenotypes. In an analysis of the binding of ATP to carp deoxyhemoglobin, Greaney et al. (1980a) have shown that the organophosphate-hemoglobin affinity constants change by two orders of magnitude between ph 8 and ph 7. The same general phenomenon appears to be true for Fundulus heteroclitus hemoglobins. In resting Fundulus at 10 C, the blood ph was about 7.9. At this ph, the difference in erythrocyte ATP between LDH-B phenotypes (ATP/Hb were 1.65 ± 0.12 and 2.11 ± 0.22 for LDH-B a B a and LDH-B b B b respectively) is not reflected as a significant difference in blood-oxygen affinity. However, as blood ph fell with increasing exercise, the organophosphate hemoglo- 100 FIG. 8. Oxygen equilibrium curves for whole blood of Fundulus heteroclitus acclimated to 10 C, as determined with an oxygen dissociation analyzer (Aminco). The ordinate is the percentage saturation of hemoglobin by oxygen, and the abscissa is the partial pressure of oxygen (po 2 ). Oxygen equilibrium curves of blood from (a) resting fish of both LDH-B phenotypes, (b) LDH-B'B" swum to exhaustion, and (c) LDH- B b B b swum to exhaustion. The intraerythrocyte ratio of ATP to hemoglobin (ATP/Hb) resting fish is 1.65 ±0.12 for LDH-B>B* and 2.11 ± 0.22 for LDH-B"B b (from DiMichele and Powers, 19826). bin affinity constant increased, and differences in oxygen affinity between homozygous LDH-B genotypes became apparent (Fig. 8). As blood ph is lowered, ATP amplifies the dissociation of oxygen from Fundulus heteroclitus hemoglobin (i.e., P 50 increases as ph decreases; see Fig. 3); the more ATP, the greater the effect. This difference is translated into a differential ability to deliver oxygen to muscle tissue which in turn affects swimming performance. This phenomenon is particularly important because it illustrates how an enzyme that is not involved in respiration can indirectly affect the availability of oxygen and thus the ability to perform work. ACKNOWLEDGMENTS We thank the National Fishery Center, Kearneysville, WV for providing us with

12 246 D. A. POWERS ET AL. rainbow trout. We are also grateful for the flounder blood supplied by Jeffrey Price. This work was supported by NSF Grants DEB , DEB , DCM and NIH Grant HL to DAP. LDM was also supported by an NIH Fellowship (F32-GM ). This is contribution No from the Department of Biology and the McCollum-Pratt Institute of The Johns Hopkins University. REFERENCES Anderson, M. E., J. S. Olson, and Q. H. Gibson Studies on ligand binding to hemoglobins from teleosts and elasmobranchs. J. Biol. Chem. 248: Arnone, A X-ray diffraction study of binding of 2,3-diphosphate glycerate to human deoxyhaemoglobin. Nature 237: Binotti, S., B. Giovenco, J. Giardina, E. Antonini, M. Brunori, and J. Wyman Studies of the functional properties offish hemoglobins. II. The oxygen equilibrium of the isolated hemoglobin components from trout blood. Arch. Biochem. Biophys. 142: Bonaventura.J., R. Gillen, and A. Riggs The hemoglobin of the Crossopterygian fish Latimeria chalumnae (Smith) Subunit structure and oxygen equilibrium. Arch. Biochem. Biophys. 163: Bourne, P. K. and A. R. Cossins On the instability of K + influx in erythrocytes of the rainbow trout, Salmo gairdneri, and the role of catecholamine hormones in maintaining in vivo influx activity. J. Exp. Biol. 101: Butler, P. J., E. W. Taylor, M. E. Capra, and W. Davison The effect of hypoxia on the levels of circulating catecholamines in the dogfish Scyliorhinuscanicula.]. Comp. Physiol. 132: Butler, P. J., E. W. Taylor, and W. Davison The effect of long-term moderate hypoxia on acid-base balance, plasma catacholamines and possible anaerobic end products in the unrestrained dogfish Scyliorhinus canicula. J. Comp. Physiol. 132: Cameron, J. N The influence of environmental variables on the haematology of the pinfish (Lagodon rhomboides) and striped mullet (Mugil cephalus). Comp. Biochem. Physiol. 32: Dalessio, P., L. DiMichele, and D. A. Powers Extracellular control of erythrocyte oxygen affinity in teleosts. Amer. Zool. 24:120A. (Abstr.) DiMichele, L. and D. A. Powers. 1982a. LDH-B genotype specific hatching times of Fundulus heteroclitus embryos. Nature 260: DiMichele, L. and D. A. Powers The physiological basis for swimming endurance differences between LDH-B genotypes of Fundulus heteroditus. Science 216: DiMichele, L. and D. A. Powers. 1984a. The relationship between oxygen consumption rate and hatching in Fundulus heteroditus. Physiol. Zool. 57: DiMichele, L. and D. A. Powers Developmental and oxygen consumption differences between LDH-B genotypes of Fundulus heteroditus and their effect on hatching time. Physiol. Zool. 57: DiMichele, L., D. A. Powers, and J. A. DiMichele Developmental and physiological consequences of genetic variation at enzyme synthesizing loci in Fundulus heteroditus. Amer. Zool. 26: Eaton, J. W Oxygen affinity and environmental adaptation. Ann. N.Y. Acad. Sci. 241: Fyhn, U. and B.Sullivan Elasmobranch hemoglobins: Dimerization and polymerization in various species. Comp. Biochem. Physiol. 50B:l Gerlach, E.andJ. Duhm ,3 DPG Metabolism of red cells: Regulation and adaptative changes during hypoxia. In P. Astrop and M. Roth (eds.), Oxygen affinity of hemoglobin and red cell acid base status. Academic Press, Munksgaard, Copenhagen, Denmark. Gillen, R. G. and A. Riggs The hemoglobins of a freshwater teleost Cichlasoma cyanoguttatum: The effects of phosphorylated organic compounds upon oxygen equilibria. Comp. Biochem. Physiol. 38B: Gillen, R. and A. Riggs Structure and function of the hemoglobins of the carp, Cyprinus carpio. J. Biol. Chem. 247: Greaney, G. S., M. K. Hobish, and D. A. Powers. 1980a. The effects of temperature and ph on the binding of ATP to carp (Cyprinus carpio) deoxyhemoglobin (Hbl). J. Biol. Chem. 255: Greaney, G. S., A. R. Place, R. E. Cashon, G. Smith, and D. A. Powers Time course of changes in enzyme activities and blood respiratory properties of killifish during long-term acclimation to hypoxia. Physiol. Zool. 53: Greaney, G. S. and D. A. Powers Cellular regulation of an allosteric modifier of fish hemoglobin. Nature 270: Greaney, G. S. and D. A. Powers Allosteric modifiers of fish hemoglobins: In vitro and in vivo studies of the effect of ambient oxygen and ph on erythrocyte ATP concentrations. J. Exp. Zool. 203: Hashimoto, K., Y. Yamaguchi, and F. Matsuura Comparative studies on two hemoglobins of salmon IV. Oxygen dissociation curve. Bull. Japan. Soc. Sci. Fish. 26:827. Houston, A. H. and K. M. Mearow Thermoacclimatory modification of red cell ionic composition in rainbow trout, Salmo gairdneri: Possible relationship with (Na + /K + )- and (HCO,-)- stimulated ATPase activities. Comp. Biochem. Physiol. 70A: Johansen, K Air breathing in fishes. In W. S.

13 MOLECULAR ECOLOGY OF FUNDULUS HEMOGLOBIN 247 Hoar and D. J. Randall (eds.), Fish physiology, Vol. IV, pp Academic Press, New York. Johansen, K. and C. Lenfant A comparative approach to the adaptability of O 2 -Hb affinity. In P. Astrup and M. Rorth (eds.), Oxygen affinity of hemoglobin and red cell acid base status. Academic Press, Munksgaard, Copenhagen, Denmark. Johansen, K. and R. E. Weber On the adaptability of haemoglobin function to environmental conditions. In P. S. Davies (ed.), Perspectives in experimental biology, pp Pergamon Press, New York. Krogh, A. and I. Leitch The respiratory function of blood in fishes. J. Physiol. London 52: 288. LeBras, Y. M Effects of anaesthesia and surgery on levels of adrenaline and noradrenaline in blood plasma of the eel (Anguilla anguilla L.). Comp. Biochem. Physiol. 720: Lenfant, C. and K. Johansen Respiration in the African lungfish, Protopterus aethiopicus. I. Respiratory properties of blood and normal patterns of breathing and gas exchange. J. Exp. Biol. 49: Mied, P. and D. A. Powers Hemoglobins of the killifish Fundulus heteroclitus: Separation, characterization and a model for the subunit composition. J. Biol. Chem. 253: Nakano, T. and N. Tomlinson Catecholamines and carbohydrate concentration in rainbow trout (Salmo gairdneri) in relation to physical disturbance. J. Fish. Res. Board Can. 24: Nakinmaa, M Effects of adrenaline on red cell volume and concentration gradient of protons across the red cell membrane in the rainbow trout, Salmo gairdneri. Mol. Physiol. 2: Nikinmaa, M Adrenergic regulation of hemoglobin oxygen affinity in rainbow trout red cell. J. Comp. Physiol. 152: Place, A. R. and D. A. Powers Genetic basis for protein polymorphism in Fundulus heteroclitus. Biochem. Genetics 16: Place, A. R. and D. A. Powers Genetic variation and relative catalytic efficiencies: The LDH-B allozymes of Fundulus heteroclitus. Proc. Natl. Acad. Sci. U.S.A. 76: Place, A. R. and D. A. Powers. 1984a. The lactate dehydrogenase (LDH-B) allozymes of Fundulus heteroclitus (Lin.): I. Purification and characterization. J. Biol. Chem. 259: Place, A. R. and D. A. Powers The lactate dehydrogenase (LDH-B) allozymes of Fundulus heteroclitus (Lin.): II. Kinetic analyses. J. Biol. Chem. 259: Powers, D. A Hemoglobin adaptation for fast and slow water habitats in sympatric catostomid fishes. Science 177: Powers, D. A Structure-function and molecular ecology offish hemoglobins. Ann. N.Y. Acad. Sci. 241: Powers, D. A The molecular ecology of teleost fish hemoglobins: Strategies for adapting to changing environments. Amer. Zool. 20: Powers, D. A Adaptation of erythrocyte function during changes in environmental oxygen and temperature. In A. R. Cossins and P. Sheterline (eds.), Cellular acclimation to environmental change, pp Cambridge Univ. Press. Powers, D. A. and A. B. Edmundson. 1972a. Multiple hemoglobins of catostomid fish. I. Isolation and characterization of the isohaemoglobins from Catostomusclarhii.]. Biol. Chem. 247: Powers, D. A. and A. B. Edmundson Multiple hemoglobins of catostomid fish. II. The amino acid sequence of the major alpha chain from Catostomusclarhii hemoglobins. J. Biol. Chem. 247: Powers, D. A., H.J. Fyhn, U. F. H. Fyhn.J. P. Martin, R. L. Garlick, and S. C. Wood. 1979a. A comparative study of the oxygen equilibria of blood from 40 genera of Amazonian fishes. Comp. Biochem. and Physiol. 62A: Powers, D. A., G. S. Greaney, and A. R. Place. 1979c. Physiological correlation between lactate dehydrogenase genotype and haemoglobin function in killifish. Nature 277: Powers, D. A., M. K. Hobish, and G. S. Greaney Rapid-rate equilibrium analysis of the interactions between organic phosphate and hemoglobins. In Methods in enzymology, Vol. 76, pp Academic Press, New York. Powers, D. A.,J. P. Martin, R. L. Garlick, and H. J. Fyhn The effect of temperature on the oxygen equilibria offish hemoglobins in relation to environmental thermal variability. Comp. Biochem. and Physiol. 62A: Powers, D. A. and A. R. Place Biochemical genetics of Fundulus heteroclitus. I. Temporal and spatial variation in gene frequencies of Ldh-B, Mdh-A, Gpi-B and Pgm-A. Biochemical Genetics 16: Powers, D. A. and D. W. Powers Predicting gene frequencies in a natural population: A testable hypothesis. In C. Markert (ed.), The isozymes, genetics and evolution, Vol. IV, pp Academic Press, New York. Prosser, C. L Comparative animal physiology. 3rd ed. W. B. Saunders Co., Philadelphia, PA. Riggs, A Properties of fish hemoglobins. In W. S. Hoar and D. J. Randall (eds.), Fish physiology, Vol. IV, pp Academic Press, New York. Rossi-Fanelli, A. and E. Antonini Oxygen equilibrium of hemoglobin from Thunnus thynnus. Nature 188: Soivio, A., M. Nikinmaa, and K. Westman The blood oxygen binding properties of hypoxic Salmo gairdneri. J. Comp. Physiol. 136: Stadel, J. M., B. Strubovici, P. Nambi, T. N. Lavin, M. M. Briggs, M. G. Caron, and R. J. Lefkowitz Desensitization of the /3-adrenergic receptor of frog erythrocytes.j. Biol. Chem. 235: Wahlquist, I. and S. Nilsson Adrenergic control of the cardio-vascular system of the Atlantic

14 248 D. A. POWERS ET AL. cod, Gadus morhua, during stress. J. Comp. Physiol. 137: Weber, R. E Respiratory properties of hemoglobin from eunicid polychaetes. J. Comp. Physiol. 99: Weber, R. E., B. Sullivan, J. Bonaventura, and C. Bonaventura The hemoglobin system of the primitive fish Amia calva: Isolation and functional characterization of the individual hemoglobin components. Biochim. Biophys. Acta 434: Wood, S. C. and K. Johansen Adaptation to hypoxia by increased HbO 2 affinity and decreased red cell ATP concentration. Nature 237: Wood, S. C, K. Johansen, and R. E. Weber Effects of ambient po s on O 2 -Hb affinity and red cell ATP concentration in a benthic fish, Pleuronectes platessa. Respir. Physiol. 25: Wyman, J Heme proteins. Adv. Protein Chem. 19: Wyman, J Linked functions and reciprocal effects in hemoglobin: A second look. In C. B. Anfinsen, M. L. Anson, J. T. Edsal, and F. M. Richards (eds.), Advances in protein chemistry, Vol. 19, pp Academic Press, New York. Wyman, J., S. J. Gill, L. Noll, B. Giardina, A. Colosima, and M. Brunori The balance sheet of a hemoglobin: Thermodynamics of CO binding by hemoglobin trout I. J. Mol. Biol. 109: Yamanaka, H., K. Yamaguchi,and F. Matsurra Starch gel electrophoresis of fish hemoglobins. I. Usefulness of cyanmethemoglobin for the electrophoresis. Bull. Japan Soc. Sci. Fish. 31: