Nano Gel Filtration Reveals How Fish Hemoglobins Release Oxygen: The Root Effect

Abstract

The Root Effect is to many species of fish what the Bohr Effect is to humans regarding the release of O₂ from their hemoglobins at low pH. However, Root Effect hemoglobins accomplish this more extensively than human adult hemoglobin in order to satisfy the diverse oxygen requirements in fish. To understand this difference between fish and human hemoglobins, we studied their subunit interface strengths using very low ( nanomolar) concentrations, referred to as nano gel filtration. Root Effect hemoglobins in their CO form dissociate in a tetramer-monomer equilibrium. In contrast, tetramers and dimers but no monomers are found for adult human hemoglobin consistent with its well known tetramer-dimer equilibrium. By analogy to the human variant Hb Kansas and a similar recombinant Hb, both of which readily release oxygen due to an unstable oxygenated structure, the mechanism proposed is that oxygenated Root Effect tetramers release their oxygen to form energetically stable deoxygenated tetramers rather than dissociate to energetically unfavorable oxygenated dimers with labile interfaces. In contrast, the strong binding of CO permits observation of dissociation to monomers, thus revealing an intrinsic property of Root Effect fish hemoglobins enabling it to function as an oxygen pump.

Graphical Abstract

1. Introduction

The binding of protons to proteins has important physiological consequences when the effects are propagated to other parts of the protein assembly and thereby influence a particular function; this is referred to as an allosteric effect [1]. For example, as a consequence of increased CO2-bicarbonate-carbonic acid formation during respiration, the pH decreases and protons bind to the N-terminus of the alpha subunit and the C-terminus of the beta subunit in the deoxy conformation of human adult Hb thus promoting release of oxygen from the heme at a distant location [2]. This allosteric relationship is known as the alkaline Bohr Effect [3] and it provides an important supply of oxygen to tissues; it is illustrated in Figure 1A (solid lines) by the decrease in O₂ binding at pH 7 compared to that at pH 8 [1–3]. At pH 6, however, the process is reversed for human adult hemoglobin and the oxygen affinity increases [4] (Fig. 1A dashed line). This property is referred to as the acid or reverse Bohr Effect and likely results in part from tetramer dissociation to stable dimers which have a higher oxygen affinity than tetramers; it is not of physiological relevance in humans whose blood pH is not normally less than 7. Full cooperativity is retained between pH 6–8 as indicated by 66the sigmoidal shapes of the curves in Fig. 1A with Hill coefficients close to 3 as shown in the inset.

Figure 1.

Typical oxygen binding curves of Bohr Effect hemoglobins (A) and Root Effect hemoglobins (B) at different pH values. The slopes of the Hill plots known as the Hill coefficients (n) indicate the degree of cooperativity and are shown as a function of pH in the insets. From Ref. 12.

Fish hemoglobins are more complex than human hemoglobin, i.e., there are multiple types, classified as anodic and cathodic, whose responses to oxygen binding vary upon changes in pH [5–8]. For example, although the oxygen binding curve of the anodic hemoglobin of teleost fish resembles that of human hemoglobin at pH 8 ( Fig. 1B ), this hemoglobin releases O₂ at pH 6 (Fig. 1B, dashed line); fish are normally exposed to pH <7. This property discovered nearly a century ago is known as the Root Effect [9] and it has profound physiological consequences for fish since it enables them to attain the high O₂ pressure necessary to fill their swim bladders in order to control buoyancy and to oxygenate other tissues including the retina [10–13]. Such hemoglobins attain only partial oxygen saturation at low pH and atmospheric pressure ( Fig. 1B) [12]; the molecular mechanism of the Root Effect has defied explanation for many years although it is generally accepted that stabilization of the deoxy conformation and an unstable oxy conformation are important factors [12,13] but how this difference arises is unclear. Besides a low O₂-binding capacity at low pH, Root Effect hemoglobins display little cooperativity compared to Bohr Effect hemoglobins (Fig. 1B, inset), a property that is also unexplained.

Root Effect and Bohr Effect hemoglobins have similar overall architectures (7,12) but differences in amino acid sequences; their oxygen binding curves at low pH bear little resemblance to each other (Fig. 1). The familiar sigmoidal oxygen binding curve of tetrameric mammalian hemoglobin has low oxygen binding in the early stage followed by more extensive binding in the mid-range due to cooperative subunit interactions. The very different oxygen binding profile of Root Effect hemoglobins at low pH (Fig. 1B) displays a plateau after about 50% saturation which may reflect a subunit assembly with intrinsically different properties than those of mammalian Hb but details are lacking. Another major difference between these two types of hemoglobins as reported by Fago et al [8 ] is that the multiple species of Root Effect rainbow trout hemoglobin IV tetramers consist of four different subunits, e.g., α^Aα^Bβ^Cβ^D, whereas mammalian Hb A tetramers consist of two different subunits arranged as pairs of the same heterodimers, e.g., α^Eβ^F α^Eβ^F. For these reasons, we decided to study fish Hb assembly properties in detail and compare them to adult mammalian Hb.

2. Materials and Methods

2.1. Preparation of Root Effect Fish Hemoglobins-

Fresh blood samples from several species of fish (rainbow trout, brook trout, brown trout, Atlantic salmon) were obtained from Dr. Mark Tisa of the Massachusetts Division of Fisheries and Wildlife as arranged by Dr. Gwilym Jones. The blood samples were drawn by venipuncture into tubes containing anticoagulant (heparin or EDTA, which gave the same results), packed in ice onsite and transported immediately to Northeastern University where they were processed to isolate the hemolysates, which were frozen at −80⁰C in their oxygenated form until use. Steven Spina of the New England Aquarium in Boston also generously provided us with freshly drawn blood samples from brook trout. The absorption spectrum of each oxygenated hemolysate at pH 6.3 showed spectral peak at 415 nm for the oxygenated form with a pronounced shoulder at 430 nm indicating the presence of a significant amount of hemoglobin in the deoxy state; this is a manifestation of the Root Effect. When the hemolysates were saturated with CO, which binds to heme more strongly than O2, both the 415 nm band and the 430 nm band completely shifted to 420 nm indicative of CO-Hb; the 430 nm band of deoxy Hb was no longer present. These are well known properties of Root Effect hemoglobins.

We separated the anodic and cathodic components from hemolysates using a Mono Q cation exchange column on an FPLC system that we described previously [14]. The presence of a Root Effect in the oxygenated forms of the anodic fraction of rainbow trout hemoglobin (also referred to as component IV [6, 8]) was demonstrated at pH 6.3 by the spontaneous appearance of both a pronounced spectral band of deoxy Hb at 430 nm and the partial transition of the 541 nm and 577 nm twin peaks characteristic of oxy Hb to the 555 nm band of deoxy Hb. The magnitude of these spectral changes indicated ~50% oxygen saturation consistent with other reports [15]. Component IV from mono Q showed the presence of 4–5 close bands by the isoelectric focusing Hb Resolve system in agreement with the results of Fago, Weber, and colleagues [8 ] who found that the Hb in each band was tetrameric with subunit masses between 15,000–16,000; there was no evidence for separate dimers or monomers. Differences in masses of 1,000 are outside the fractionation range of our system but would not affect the tetramer-monomer distribution reported here.

2.2. Nano Gel Filtration-

Hemoglobin tetramer dissociation was measured on an FPLC column using Superose-12 and the software data analysis system described in detail previously ( 16, 17). Standard hemoglobins used to calibrate the column were: for the tetramer position--a crosslinked Hb with a covalent bond between the N-terminals of the two alpha subunits rendering it undissociable; for the dimer position—the natural Hb Rothschild with an Arg to Trp substitution at position 37 of the beta-subunit precluding tetramer formation; for the monomer—the hydroxymercuribenzoate derivative on the alpha subunit preventing tetramerization (16, 17). Replicate FPLC analysis of each standard showed that peak positions did not vary. Hb A was used to evaluate reproducibility of peak position and peak width; reproducibility was 0.003% and 1%, respectively, for 6 different samples [16, 17]. Kd values were read directly from double reciprocal plots of the data as described earlier [16,17]. Some samples and elution buffers were saturated with CO by bubbling with 100% CO for several minutes prior to application to the column; other samples were maintained in the oxy state as they were isolated from the fish blood. Samples were analyzed by high resolution gel filtration in 150mM Tris-Ac, pH 6.3, using very dilute hemoglobin concentrations in order to expose the nature of the equilibria between the monomeric, dimeric, and tetrameric constituents of their assemblies. We refer to this technique as nano gel filtration and it was coupled to a mathematical analysis of the data [16,17]. Avoiding the high haemoglobin concentrations usually employed, which greatly exceed their equilibrium dissociation constants, enables observation of all assembly intermediates since it shifts the equilibrium away from the predominant tetramer species and towards its constituent dimer and monomer components. This rationale is analogous to the principle used a century ago to accurately determine gravity-induced bending of light by measuring it during a solar eclipse when the intensity of the sun’s intense glare, which would ordinarily obscure such observations, was minimized [18 ]; this experiment was an important proof of Einstein’s theory of general relativity.

3. Results

3.1. Properties of Hemoglobin Assemblies -

The nature of the equilibria involving the assembly of the adult human hemoglobin tetramer from its subunits is well known [2,19]; the corresponding properties for Root Effect fish hemoglobins have been tacitly assumed to reflect those of human adult hemoglobin. Hence, explanations for the Root Effect in fish hemoglobin usually focus on their primary sequence differences compared to adult human hemoglobin and how these are construed to affect its oxygen-binding properties. This approach referred to as homology modeling assumes that the underlying structural foundations are the same for both hemoglobin types. It can provide useful information in some cases but has not been successful in explaining the Root Effect. Hence, we have chosen an alternative approach described below.

Adult hemoglobin exists in two stable tetrameric states---oxy and deoxy in equilibrium with their respective dimers, which have very stable interfaces and do not appreciably dissociate to monomers under physiological conditions [2, 19]. We studied the corresponding assembly equilibria for Root Effect hemoglobins using nanomolar hemoglobin concentrations and a very high resolution gel filtration method [16,17] described in detail in section 2.2. We compared the subunit bonding strengths between monomers in the dimer and between dimers in the tetramer for Bohr Effect and Root Effect hemoglobins and we find significant differences between them consistent with their vastly different oxygen binding properties [20, 21].

3.2. Complete Dissociation of Rainbow Trout Hemoglobin in the CO-State-

The dissociation of anodic rainbow trout hemoglobin in the CO-form as a function of concentration is shown in Fig. 2; it displays a smooth profile from tetramer to monomer. Very low concentrations (3–164 nM present on the column) were needed to observe complete dissociation, indicative of a strong tetramer, a conclusion also reached by Edelstein, Gibson and colleagues for several types of fish hemoglobin [20]. Tetrameric, dimeric and monomeric species are readily separable on the system we used as described in section 2.2. Importantly, there was no distinct dimeric species ( Fig. 3B) which contrasts sharply with the dissociation profile of human adult hemoglobin in non-dissociating buffers where the dissociation stops at the dimer stage ( Fig. 3A, red line and Fig. 3C, red dot).

Figure 2.

Complete dissociation of rainbow trout hemoglobin in the CO-Form. Samples and elution buffers were saturated with 100% CO before application to the column. The concentrations of hemoglobin on the column were 3–164 nM. The buffer was 150mM Tris-Ac, pH 6.3. Tetramer, dimer, and monomer standards are described in the text.

The two important conclusions are that tetramers are capable of complete dissociation to monomers and that tetramer formation is extremely efficient (about 300 more efficient than adult human hemoglobin}. These properties enable fish containing Root Effect Hb both to readily bind the low amounts of oxygen in its environment (due to the monomer component} and to release it to tissues that utilize it ( due to the ease of tetramer formation). The Graphical Abstract was constructed based on these two conclusions.

Figure 3.

Representative nano gel filtration patterns of hemoglobins. Panel A. Trout Hb (blue line) and human HbA (red line) were analyzed at pH 6.3 as described in (16,17). Different concentrations were used for each Hb because trout Hb is a stronger tetramer than human HbA (see text and Panel B). The anodic fractions of Root Effect hemoglobins from brook trout, brown trout, rainbow trout, and Atlantic salmon gave similar patterns.

Panel B. Rainbow trout Hb complete dissociation profile over a range of concentrations. The blue point represents the lowest concentration tested (3nM). The samples were saturated with CO prior to application to the column. Panel C. Adult human HbA dissociation profile over a range of concentrations. The red point represents the lowest concentration tested (10 nM). T = tetramer, D = dimer, M = monomer. The complete dissociation of tetramers to monomers in Panel B is defined as 100%; the dissociation of tetramers to dimers in Panel C is defined as 50%.

The tetramer-monomer dissociation constant (Kd) of rainbow trout Hb in the CO-form at pH 6.3 was calculated from Fig, 2 to be 7 nM, which is 300 times lower than the tetramer-dimer Kd of human adult Hb, which is 2100 nM at pH 6.5 [16]. The same profile was noted for brook trout hemoglobin [ 23 ] in the CO-form with a Kd of 12 nM in good agreement with results in Fig. 2 for rainbow trout Hb. Such a low Kd ensures that tetramerization of subunits occurs very efficiently at very low hemoglobin and oxygen concentrations, which likely suits the development of most fish. Anodic hemoglobins from brown trout and from Atlantic salmon also dissociate to monomers. Sometimes tetramers were present in addition to monomer peaks but discrete dimers were never observed during dissociation ( Figs. 3A, blue line). We attribute this to the partial displacement of CO by O2 when CO amounts are limiting; such oxygenated subunits would convert to deoxygenated tetramers ( see also section 3.4). Dimers are undoubtedly present but their existence is transient due to rapid dissociation to monomers.

3.3. Peak Widths During Dissociation-

The dissociation of rainbow trout Hb shown in Fig. 2 was determined from changes in peak positions during the nano gel filtration procedure described in detail in section 2.2. Because of the high resolution capability of the system, peak widths are an independent measure of the number of species involved in the dissociation process with the maximum width occurring close to the Kd value [17,24]. Only dissociable hemoglobins but not covalently crosslinked undissociable hemoglobin (used as the tetramer standard in calibrating the column) displayed this expansion and contractionof peak width during dissociation [17, 24]. Adult and fetal human hemoglobins have peak widths of 0.65–0.70 ml occurring when the highest relative concentrations of tetramers and dimers ( no monomers detected) were present during dissociation [24]. However, during the dissociation of the CO-form of tetrameric rainbow trout Hb shown in the inset of Fig. 2, the peak widths increased to 1.5 ml, as tetramers dissociated and then contracted to their original width when the dissociation was complete, i.e., when predominantly monomers were present. The maximum peak at 1.5 ml occurred when all three species--tetramer, dimer, and monomer were present at their highest relative concentrations consistent with a tetramer-monomer equilibrium. In a separate study on the dissociation of brook trout Hb, the maximum peak width was 1.2 ml [23]. Thus, this independent measurement of peak width due to the extent of tetramer dissociation shows a significant difference between Bohr Effect and Root Effect hemoglobins confirming the complete dissociation of rainbow trout Hb. Rainbow trout hemoglobin in the cathodic fraction does not have a Root Effect and remains tetrameric at all concentrations tested.

3.4. Functional Significance of Complete Dissociation of Root Effect Trout Hemoglobin in the CO-Form-

Two major conclusions can be drawn from the results in Figure 2---first, tetramer formation is highly efficient at low Hb concentrations and much more so than human adult hemoglobin and second, the equilibrium system always contains some monomer regardless of the Hb concentration. The implications of these properties are described in section 4.8 as it relates to Root Effect hemoglobin functioning as an oxygen pump. Since tetramers and monomers reflect low (favorable) and high (unfavorable) energy states, respectively, there is a thermodynamic preference for tetramer formation in the deoxy conformation as studies with adult human hemoglobin have shown ( 19 ). The most important difference between the two hemoglobin types is an unstable liganded tetramer in the case of the Root Effect hemoglobins (in contrast to adult human Hb) as reported here ( see Graphical Abstract) and by other studies ( see section 4.3) leading to the presence of deoxy tetramers in equilibrium with monomers ( Figures 2. 3A, 3B). Such monomers would be expected to bind O2 preferentially.

3.5. Dissociation of Oxygenated Root Effect Rainbow Trout Hemoglobin-

In contrast to the smooth and complete dissociation of the CO form of rainbow trout Hb described in Figure 2, in the presence of O₂ both tetramers and monomers are found (Fig. 3A, blue line). This result is consistent with the absorption spectra of the oxygenated hemolysates, described in section 2.1, showing the presence of some of the deoxygenated tetramer. This conclusion also agrees with the findings of Mazzarella, di Prisco and colleagues [25] who could not obtain crystals of the oxy form of trout Hb IV at pH 6.2 because it deoxygenated spontaneously. The differences in the dissociation profiles of the CO-form compared to the oxy form of Root Effect hemoglobin suggests a possible mechanism (see below) to explain both its inability to attain greater than 50% O2 saturation at atmospheric pressure ( Fig. 1) and its partial spectral conversion to the deoxy form described in section 2.1.

3.6. Dissociation of Bohr Effect Hemoglobins-

Very dilute concentrations of human adult HbA were dimeric with no evidence of dissociation to monomers (Fig. 3C ); this behavior is well known for the dissociation profile of adult Hb A [2,4,16,19] and is due to a dimer with a strong interface. This property ensures that the oxygen affinity of adult Hb will not increase as it does for the embryonic hemoglobins due to their weaker dimer interfaces [24–27], i.e. the oxygen-binding property of the overall assembly is influenced by contributions of all equilibria-- from monomers with very high oxygen affinity to tetramers with very low affinity and intermediate assembly components in between these extrema.

4. Discussion

4.1. Equilibria Differences Between Root Effect and Bohr Effect Hemoglobins -

Our results can be explained by two different equilibria for Root Effect and Bohr Effect hemoglobins (Fig. 4); for both systems, the energetic stability increases from right to left with deoxy tetramers being energetically the most stable and the oxygenated monomers being energetically the least stable (19). It is the overall equilibria balance of each system that determines its oxygen carrying properties. The dashed green arrows indicate which steps are unfavorable, which differ for the two types of hemoglobin. For human hemoglobin, the alkaline Bohr Effect involves the binding of protons to the deoxy tetramer (T) resulting in the release of O₂ as shown in the top equation [3]. Further lowering of the pH promotes the dissociation of tetramers to dimers [4 ] as shown by a shift to the right from T (O₂) to D, which does not dissociate further to monomers for mammalian Hb (green dashed arrow). A small percentage of dimers is always present in dynamic and rapid equilibrium with tetramers at any Hb concentration.

Figure 4.

Equilibria differences between Root Effect trout Hb and Bohr Effect adult human HbA. The green dashed arrows represent unfavorable steps in each equilibrium, e.g. in the oxy conformation, adult Bohr Effect Hb has a strong dimer mwhich does not dissociate to monomers but fish Root Effect Hb has a very weak dimer so the oxygenated tetramer readily rearranges to the deoxy conformation.

Root Effect tetramers can be observed to dissociate beyond the dimer stage to monomers in the presence of a strong ligand such as CO; hence, their dimer interfaces are labile. However, with the weaker O₂ ligand, this dissociation is energetically unfavorable (Fig. 4, lower equation, green dashed arrow). Hence, oxygenated tetramers of Root Effect hemoglobin are unstable as shown here and by other investigators ( see section 4.3) so the equilibrium shifts to the left favoring the stable deoxy conformation and thereby releasing O₂. Perhaps the presence of an acetyl group at the N-terminus of Root Effect hemoglobins reported by Fago et al. [8] and others, a modification that we demonstrated weakens the Hb F tetramer interfaces in Hb F₁[14], favors dimer to monomer dissociation in fish Hb. This equilibrium shift is reminiscent of that for the natural hemoglobin variant Hb Kansas (N102T beta), as described in section 4.2.

4.2. Mammalian Mutant Hemoglobins With Partially Destabilized Oxygenated Tetramers-

The model proposed here for Root Effect hemoglobins, which invokes a shift in the overall equilibrium (Fig. 4, lower equation), has a precedent in the natural human variant, Hb Kansas, which has a very low oxygen affinity. This human mutant Hb, which was extensively characterized by Bonaventura and Riggs [28 ], has an Asn→Thr substitution at beta-102, an important contact at the tetramer-dimer interface of the oxy conformation tetramer rendering it energetically unstable and thus shifting the tetrameric conformational equilibrium partially towards the deoxy state. In addition to a reduced oxygen affinity, this natural Hb mutant also dissociates to dimers about 200 times more than does Hb A. Hence, with respect to the equilibrium in the upper equation of Fig. 4, Hb Kansas undergoes conformational shifts in both directions-- away from the oxygenated destabilized tetramer -- towards both stable deoxygenated tetramers and also towards oxygenated dimers. Since its oxygen binding curve is shifted significantly to the right, the shift towards deoxygenated tetramers predominates. We extended these results with a study of the recombinant Hb N102A-beta [ 29 ]. Like Hb Kansas, this hemoglobin also had a very low oxygen affinity (P50=42 mm Hg compared to a P50 = 5mmHg for Hb A) with cooperativity decreased to 1.9 compared to 2.4 for Hb A. Furthermore, it dissociated to dimers about 400 times more than did Hb A. The above findings on the two partially destabilized mammalian hemoglobin tetramers support the mechanism proposed for Root Effect hemoglobins regarding how they attain their unique very low oxygen binding properties; since oxygenated dimers have not been detected for Root Effect hemoglobins, the right shift to dimers mentioned above for partially destabilized mammalian tetramers, Hb Kansas and Hb N102A-beta, is absent in Root Effect hemoglobins thereby forcing the equilibrium only to the left in favor of the deoxygenated conformation ( see Graphical Abstract).

4.3. Proton Uptake in Root Effect and Bohr Effect Hemoglobins-

Analogous to the original experiments of Bohr et al. [3] on the release of oxygen from human Hb, Root’s experiments on fish Hb were performed with added CO₂. It was subsequently realized that protons were the active agent due to their presence in the equilibrium of CO₂ in H₂O involving formation of carbonic acid, protons and bicarbonate. Whereas about 2 protons per tetramer are absorbed for the deoxy state of human Hb as a result of the alkaline Bohr Effect [3], the uptake of protons by Root Effect fish Hb is about 4 depending on the species [30] [Fig. 4, both equations]. Candidate sites have been suggested but none yet confirmed although there is general agreement among most investigators that the deoxy conformation is greatly stabilized compared to the oxy conformation in Root Effect hemoglobins [31–36]. We suggest that this enhanced stabilization of the deoxy conformation by binding additional protons further drives the release of oxygen from the oxy conformation rather than dissociating to dimers and monomers, an energetically unfavorable event.

4.4. Comparison With Human Embryonic Hemoglobins-

In previous studies using nano gel filtration we reported the importance of the assembly equilibria of human embryonic hemoglobins in determining the mechanism of their important physiological property of increased O₂ affinity that matches the low O₂ concentration in the embryo [21]. This was found to be due to moderately weak subunit interactions for the embryonic hemoglobins as demonstrated by a significant monomer population during nano gel filtration, i.e., an attenuation (loosening) of the tetrameric structure, a profile not found for fetal and adult human hemoglobins. This feature correlated with the increased O₂ affinity of the embryonic hemoglobins and is consistent with the allosteric model of Monod, Wyman, and Changeux [1] although in this case association of embryonic subunits generates weaker tetramers than for fetal and adult hemoglobins. These weaker subunit interfaces also correlated with the decreased longevity of these hemoglobins in vivo, suggesting a role for subunit interface strength in the duration of their gene expressions [24, 26, 27].

4.5. Multiplicity in Root Effect Hemoglobins-

The results presented here on the complete dissociation of Root Effect hemoglobins are supported by the reports of Fago et al. [8] on multiple rainbow trout hemoglobins each containing 4 different subunits; mammalian hemoglobins have 2 different subunits arranged as heterodimer pairs. We interpret the results of Fago et al (8) as perhaps due to dimers with weak interfaces that easily dissociate to monomers and re-associate to different tetramers. In such a tetramer-monomer equilibrium, a finite concentration of monomers would be present at all concentrations. Hence, for rainbow trout hemoglobins random re-assembly of a variety of such monomers would generate multiple hemoglobin tetramers each comprised of 4 different subunits in contrast to re-assembly of identical dimers comprised of only 2 different subunits as in mammalian hemoglobins. If the multiple hemoglobins in fish hemoglobins had different oxygen binding affinities, this would be advantageous since it would extend their range into water depths containing varying O₂ levels, analogous to the weak subunit interfaces of human embryonic hemoglobins leading to their increased oxygen affinity that matches the low oxygen levels in utero (21).

4.6. Linkage Between Labile Dimers and Loss of Cooperativity-

The basis for the loss of cooperativity of Root Effect hemoglobins at low pH shown in Fig. 1B (inset) has been known for many years but has never been explained. We suggest that labile dimers could contribute to this behavior since cooperativity in mammalian Hb A tetramers arises when the dimers rearrange their subunit contacts during the transition between the deoxy and oxy tetrameric conformations; in order for this transition to occur the dimers must be stable. A weak dimer prone to dissociation to monomers would not be functional in this process.

4.7. Additional Roles of Root Effect Hemoglobins-

In addition to the known functions of the Root Effect hemoglobins enabling teleost fish to attain the high oxygen pressure to fill their swim bladders and to oxygenate their retinas ( see section 1 ), Rummer and Brauner and colleagues ( 37,38 ) have recently reported other ramifications of the Root Effect in fish hemoglobin. They found a high degree of oxygenation in muscle cells of rainbow trout in the presence of carbonic anhydrase during periods of stress generating excess CO₂ and high acidity. The increase in muscle oxygenation was as high as 65% under such conditions pointing to a more generalized role of the Root Effect in tissue oxygenation than had been previously realized.

4.8. Root Effect Hemoglobin as an Oxygen Pump-

The tetramer-monomer equilibrium described here may be an important aspect of the functional ramifications of the Root Effect as a dispenser of oxygen to a variety of tissues in fish, as described above. The dynamics of the scheme shown in the Graphical Abstract describing the interrelationships among the tetramers and monomers suggest a mechanism whereby Root Effect hemoglobin could act as an oxygen pump capable of operating against high pressure gradients ( 39 ). Monomers with their very high oxygen affinity (19). would enable the binding of the very low amounts of oxygen available in the waters where rainbow trout live. Rapid tetramerization of oxy monomers at low Hb concentrations ( due to the very low Kd of 7 nM calculated from the data in Fig. 2) would form an oxygenated tetramer which, because of its instability, would rapidly switch to the deoxy tetameric state and release oxygen. These events would occur at low oxygen tension without complete saturation. The driving force in this model is the significant free energy difference between monomers and tetramers, a well known property of hemoglobins ( 19 ). Because the system is in equilibrium ( see Graphical Abstract), deoxygenated monomers would then bind additional oxygen to complete the cycle, which represents a self-sustaining system whereby the energy difference involved in protein subunit association / dissociation would be transduced to metabolic energy generated by the transfer of oxygen to tissues that require it; such a system was envisaged by Brunori and Wyman and colleagues ( 39 ).

HIGHLIGHTS.

• --The Root Effect in fish hemoglobins promotes oxygen release at pH less than 7.
• --We studied human adult and fish hemoglobins by nano gel fitration, which permits detection of protein assembly intermediates.
• --Root Effect hemoglobins have a tetramer-monomer equilibrium
• --Human adult hemoglobin has a tetramer-dimer equilibrium.
• --As a result of a dynamic tetramer-monomer equilibrium, Root Effect hemoglobins act as an oxygen pump.