The effects of inositol hexaphosphate on the allosteric properties of two beta-99-substituted abnormal hemoglobins, hemoglobin Yakima and hemoglobin Kempsey.

Hemoglobins (Hb) Yakima and Kempsey were purified from patients' blood with diethylaminoethyl cellulose column chromatography. The oxygen equilibrium curves of the two hemoglobins and the effects of organic phosphates on the function were investigated. In 0.1 M phosphate buffer, Hill's constants n for Hb Yakima and Hb Kempsey were 1.0 to 1.1 at the pH range for 6.5 to 8.0 and the oxygen affinities of both the mutant hemoglobins were about 15 to 20 times that of Hb A at pH 7.0. The Bohr effect was normal in Hb Yakima and one-fourth normal in Hb Kempsey. In the presence of inositol hexaphosphate, the oxygen affinities to Hb Yakima and Hb Kempsey were greatly decreased, and an interesting result revealed that these hemoglobins showed clear cooperativity in oxygen binding. Hill's constant n in the presence of inositol hexaphosphate was 1.9 for Hb Kempsey and 2.3 for Hb Yakima at pH 7.0. The cooperativities of these mutant hemoglobins were pH-dependent, and Hb Kempsey showed high cooperativity at low pH (n equal 2.1 at pH 6.6) and low cooperativity at high pH (n equal 1.0 at pH 8.0). Hb Yakima showed similar pH dependence in cooperativity. In the presence of inositol hexaphosphate, Hb A showed a pH-dependent cooperativity different from those of Hb Yakima and Hb Kempsey, namely, Hill's n was the highest in alkaline pH (n equal 3.0 at pH 8.0) and decreased at lower pH (n equal 1.5 at pH 6.5). 2,3Diphosphoglycerate bound with the deoxygenated Hb Yakima and Hb Kempsey, however, had no effect on the oxygen binding of these abnormal hemoglobin. The pH-dependent cooperativity of alpha1beta2 contact anomalous hemoglobin and normal hemoglobin was explained by the shifts in the equilibrium between the high and low ligand affinity forms.

Hemoglobins (Hb) Yakima and Kempsey were purified from patients' blood with diethylaminoethyl cellulose column chromatography. The oxygen equilibrium curves of the two hemoglobins and the effects of organic phosphates on the function were investigated. In 0.1 M phosphate buffer, Hill's constants n for Hb Yakima and Hb Kempsey were 1.0 to 1.1 at the pH range for 6.5 to 8.0 and the oxygen affinities of both the mutant hemoglobins were about 15 to 20 times that of Hb A at pH 7.0. The Bohr effect was normal in Hb Yakima and one-fourth normal in Hb Kempsey. In the presence of inositol hexaphosphate, the oxygen affinities of Hb Yakima and Hb Kempsey were greatly decreased, and an interesting result revealed that these hemoglobins showed clear cooperativity in oxygen binding. Hill's constant n in the presence of inositol hexaphosphate was 1.9 for Hb Kempsey and 2.3 for Hb Yakima at pH 7.0. The cooperativities of these mutant hemoglobins were pH-dependent, and Hb Kempsey showed high cooperativity at low pH (n = 2.1 at pH 6.6) and low cooperativity at high pH (n = 1.0 at pH 8.0). Hb Yakima showed similar pH dependence in cooperativity. In the presence of inositol hexaphosphate, Hb A showed a pH-dependent cooperativity different from those of Hb Yakima and Hb Kempsey, namely, Hill's n was the highest in alkaline pH (n = 3.0 at pH 8.0) and decreased at lower pH (n = 1.5 at pH 6.5). 2,8Diphosphoglycerate bound with the deoxygenated Hb Yakima and Hb Kempsey, however, had no effect on the oxygen binding of these abnormal hemoglobins. The pH-dependent cooperativity of a,fiI contact anomalous hemoglobin and normal hemoglobin was explained by the shifts in the equilibrium between the high and low Xgand affinity forms.
Hemoglobins Yakima and Kempsey are abnormal hemoglobins which show increased oxygen affinity and decreased heme-heme interaction (l-3). The /399(Gl) aspartic acid in Hb' A is replaced by histidine in Hb Yakima and by asparagine in Hb Kempsey. The substitution is located at the a#, contact region of the protein (1,3). According to the x-ray crystallographic analysis by Perutz  from those of normal hemoglobin. An oxygen equilibrium study of Hb Yakima was carried out mainly with hemolysate and partially with purified Hb Yakima (2), whereas that of Hb Kempsey was concerned only with unfractionated hemolysate (3). The oxygen equilibria of purified samples and their detailed analysis have not been reported.
In this paper, we describe the oxygen equilibrium curves and the effects of organic phosphates on the function by using purified Hb Yakima and Hb Kempsey and we discuss the relationship between oxygen affinity and heme-heme interactions.
with Dowex 50 resin, and neutralized to pH 7 with a sodium hydroxide solution. IHP and bis-tris were purchased from Sigma. Separation of Hb Yakima and Hb Kempsey from Hb A-Hemolysate from heterozygotes contained approximately 40% Hb Yakima and 60% Hb A. These two hemoglobins were separated by chromatography on DEAE-cellulose (Whatman DE32) as follows. Glycerinated erythrocytes containing Hb Yakima (stored at -80" before use) were hemolyzed by dialysis against deionized water. After removing the ghosts by centrifugation, the hemoglobin solution was equilibrated with 0.01 M Tris-HCl buffer, pH 8.3, by passage through a Sephadex column, and applied to a DE32 column (1.5 x 40 cm) which had been previously equilibrated with the same buffer. Elution was performed bv a linear eradient from 0 to 0.1 M NaCl in 0.01 M Tris-HCl buffer. nH 8:3. A typical elution pattern is shown in Fig. 1. Hb Kempsey 'was separated from Hb A by the same procedure as that used for Hb Yakima. The completeness of separation was verified by acrylamide gel electrophoresis in the discontinuous system of Poulik (5). No trace of Hb A was found in any of the preparations used.
Stripped hemoglobin was prepared according to the method of Benesch et al. (6) but with a slight modification. A Sephadex G-25 column was used which was equilibrated with 0.1 M NaCl in 0.05 M bistris buffer, pH 7.0, instead of 0.1 M NaCl solution.

Oxygen
Equilibrium-Oxygen equilibrium curves were determined by a spectrophotometric method according to Sugita and Yoneyama (7). Concentrations of hemoglobin were between 4 and 6 x 10mB M (in heme) in 0.1 M phosphate buffer. The effects of organic phosphates on the oxygen equilibrium curves were measured in 0.05 M bis-tris buffer containing 0.1 M NaCl. The pH of the hemoglobin solution was determined at the end of each experiment. Deoxygenation was carried out by repeating alternate evacuation and flushing with Q gas (helium-isobutane, 99.05:0.95) in a Thunberg type cell with a l-cm light path and then small quantities of sodium borohydride were added to ensure the complete deoxygenation of the sample. As Hb A in the presence of IHP could not be completely oxygenated by air at pH below 7.0, 100% oxygenation of hemoglobin was performed by using pure oxygen.

Measurements of 2,3-P,-glycerate
and IHP Binding-The amount of 2,3-P,-glycerate bound to hemoglobin was measured by equilibrium dialysis according to De Bruin and Janssen (8) and the number of moles bound per hemoglobin tetramer calculated. The gel filtration method of Gray and Gibson (9) was adopted for the determination of IHP bound to hemoglobin because IHP did not equilibrate across a Visking dialysis membrane. The amount of hemoglobin was determined spectrophotometrically by the absorption of the oxygenated form at 580 nm, and that of phosphate was determined by the method of Gray and Gibson (9). This latter method is suitable for the samples containing large amounts of organic materials.

Oxygen Equilibrium
of Hb Yakima and Hb Kempsey-The oxygen affinities of Hb Yakima and Hb Kempsey as a function of pH are shown in Fig. 2. Oxygen affinities of these abnormal hemoglobins were much higher than that of Hb A. The psO values, which correspond to the oxygen pressure at which one-half of the hemoglobin is in the oxygenated forms, were 10.5 mmHg for Hb A, 0.7 mmHg for Hb Yakima, and 0.

for both Hb Yakima and Hb
Kempsey and 2.6 to 2.8 for Hb A, and the results showed the lack of cooperative interactions in Hb Yakima and Hb Kempsey. Plots of Y/l -Y uersus po, were linear over the pH range examined.
The Alog p,,/ApH values calculated between pH 7.3 and 7.8, are 0.65, 0.50, and 0.18 for Hb A, Hb Yakima, and Hb Kempsey, respectively. The Bohr effect was almost normal for Hb Yakima and one-fourth for Hb Kempsey compared with that for Hb A.
Effects of Organic Phosphates on Oxygen Equilibrium- Fig.   3 shows the oxygen equilibrium curves of Hb Kempsey and Hb A in the presence and absence of organic phosphates. The oxygen affinity of stripped Hb A was decreased to about 2 times and 10 times in pso value by the addition of 2,3-Pp-glycerate and IHP, respectively. However, as already reported by Benesch and Benesch (lo), the values of Hill's n for stripped Hb A did not change in the presence and absence of these organic phosphates at neutral pH. When Hb Kempsey was stripped, the oxygen affinity increased to about twice that of nonstripped hemoglobin. 2,3-Pz-glycerate had no effect on the oxygen affinity and heme-heme interaction of stripped Hb Kempsey, but in the presence of IHP the oxygen affinity of Hb Kempsey showed a 5-fold increase in psO. The Hill's coefficient n was increased from 1.0 to 1.9 at pH 7. This result shows that  Fig. 4 shows the effects of 2,3-PI-glycerate and IHP on stripped Hb Yakima. 2,3-PI-glycerate had no effect on the oxygen affinity and heme-heme interaction of Hb Yakima. IHP reduced the oxygen affinity of Hb Yakima and the Hill's coefficient n was increased from 1.0 to 2.3 at pH 7. These results indicate that IHP has a great effect on the cooperativity of these cu& contact anomalous hemoglobins. The changes in psO of Hb Kempsey, Hb Yakima, and Hb A as a function of pH in the presence of IHP are shown in Fig. 5 together with those of nonstripped hemoglobins in 0.1 M phosphate buffer solution for comparison. As shown in the previous section, the Bohr effect of Hb Kempsey in the absence of IHP was one-fourth that for Hb A, and that of Hb Yakima was nearly normal. By the addition of IHP, changes in psO with pH of both Hb Kempsey and Hb Yakima were greatly increased, whereas the Bohr effect of Hb A was scarcely changed. The curves of log plo uersus pH for Hb A in the presence of IHP were shifted to the upper region, indicating that the oxygen affinity was greatly decreased by binding with IHP. The Bohr effect (Alog p,,/ApH) calculated between pH 7.3 and 7.8 are given in Table I. Hb A showed about the same magnitude of the Bohr effect both in the presence and absence of IHP, whereas the two abnormal hemoglobins exhibited larger Bohr effects in the presence of IHP. However, the Hill's n in the presence of IHP was greatly changed with pH in both normal and abnormal hemoglobins. Therefore, Alog pJApH of each hemoglobin in Table I is an apparent Bohr effect.
By measuring 2,3-PI-glycerate binding to hemoglobin by equilibrium dialysis (8) and IHP binding to hemoglobin by gel filtration (9), we confirmed that under the present experimental conditions, both the organic phosphate compounds were stoichiometrically bound to Hb Yakima and Hb Kempsey in the deoxygenated state.
Change of Hill Coefficient n at Function ofpH-The Hill's n of Hb A, Hb Yakima, and Hb Kempsey in the presence of IHP varied with the change in pH. The relationship between pH and the Hill coefficient of Hb A and Hb Kempsey are shown in Fig. 6. The cooperativity of Hb A was enhanced by increasing pH (n = 3.0 at pH 8.0) and was dropped at a lower pH range (n = 1.5 at pH 6.5). On the contrary, the n values of Hb Kempsey were increased by lowering pH (n = 2.1 at pH 6.5) and were decreased at higher pH (n = 1.0 at pH 8.0). IHP was most effective in restoring the cooperativity of Hb Kempsey at the acidic pH. The Hill's n of Hb Yakima changed from 1.6 at pH 7.9 to 2.3 at pH 6.7. Accordingly, the effects of pH on the Hill coefficient of abnormal hemoglobins Yakima and Kempsey were quite different from that of Hb A in the presence of IHP.  Fig. 3.
Conditions are the same as in Fig. 5.

DISCUSSION
Structural integrity at a& contact region is of great importance for normal hemoglobin function, because most of the amino acid residues in this region are invariants in all of the known hemoglobins of vertebrate and almost all mutants 1  with replacements at the contact region show impaired hemeheme interaction (11). Such mutants also possess abnormal affinity for ligands whereas most of them retain a normal Bohr effect.
The oxygen equilibrium curves of purified Hb Yakima and Hb Kempsey in 0.1 M phosphate buffer are hyperbolic, indicating significant impairment of heme-heme interaction. Hill's n over the pH range examined in the present experiment are 1.0 to 1.1 for both Hb Yakima and Hb Kempsey. These values are similar to those of Novy et al. (2) for Hb Yakima (n = 1.0) and of Reed et al. (3) for Hb Kempsey (n = 1.1) which were obtained from the oxygen equilibrium curves of mixed hemolysate. They also showed the normal Bohr effect for Hb Yakima and the decreased Bohr effect for Hb Kempsey by comparing the changes of pH during oxygenation and deoxygenation. From the present measurements of oxygen equilibrium with spectrophotometric method, the Bohr effect for Hb Yakima was approximately the same as that for normal hemoglobin and that for Hb Kempsey was one-fourth that for Hb A. 2,3-P,-glycerate was bound to both Hb Yakima and Hb Kempsey in the deoxygenated state, but it had no effect on the ligand binding properties of two /399-substituted hemoglobins. When IHP was bound to these abnormal hemoglobins, it had great effects on the ligand affinity, the Bohr effect, and the cooperativity. The deoxygenated Hb Yakima and Hb Kempsey do not seem to convert to a constrained form by binding with 2,3-P,-glycerate but by binding with a more potent allosteric effector, IHP, the conformations of the abnormal hemoglobins may change and their oxygen affinity decreases. In the presence of IHP, the oxygen affinity of Hb Yakima and Hb Kempsey were greatly decreased (pe,, at pH 7.0; 0.28(-IHP) and 1.4( +IHP) for Hb Kempsey, 0.36( -1HP) and 2.25( +IHP) for Hb Yakima) and the Hill's n at pH 7.0 was increased from 1.0 to 2.3 in Hb Yakima and to 1.9 in Hb Kempsey. Hb Kempsey and Hb Yakima showed the greatest cooperativity at acidic pH. In the presence of IHP, the cooperativity in Hb A was the greatest at alkaline pH. The result concerning Hb A is similar to those obtained by Bunn and Guidotti (12).
The ligand affinities and the Hill's n for Hb Yakima, H.b Kempsey, and Hb A in various conditions are plotted on the same figure as shown in Fig. 7, and the over-all dependence of cooperativity on ligand affinity is found to be hyperbolic with n being decreased at both the extremely high and low values of pso. In the range between 2 and 50 mmHg of psO, the n values of where Y is the fractional saturation and a is the normalized ligand concentration. Rubin and Changeux (14) and Edelstein (15) had shown through numerical analysis that the dependence of Hill's constant on L follows a bell-shaped curve, that is, the degree of cooperativity is largest at an optimum value of L and decreases on either side. They had also shown that L is directly related to a. Bunn and Guidotti (12) and Minton (16) also found a dependence of n on pso. They showed, however, that no single value of c could fit the calculated curve to all of the data and suggested the change in the intrinsic affinity of the T state.
The two-state model postulates that the conformation of hemoglobin with high psO is almost frozen in the T state and with low psO in the R state. In accordance with the model, hemoglobin variants with pso of less than 2 mmHg, such as Hb Chesapeake (pso = 0.4 mmHg at pH 7.2, n = 1.3) (17) and Hb Rainier (pso = 0.5 mmHg at pH 7.1, n = 1.3) (18), usually show decreased heme-heme interaction. Hemoglobin variants with very low oxygen affinity, such as Hb M Iwate (plo = 55 mmHg at pH 7.0, n = 1.0) (19) and Hb Kansas (p6,, = 40 mmHg at pH 7.0, n = 1.3) (20), have also decreased cooperativity. Hb A and other hemoglobin variants which show normal ligand affinity are considered to take a suitable equilibrium constant, L for high n value.
In Fig. 7, each set of points for Hb A, Hb Yakima, and Hb Kempsey at various pH values is on a calculated curve based on two-state model with a fixed c value according to Bunn and Guidotti (12), and the results can be interpreted to mean that proton produces a change in L without affecting the value of c. IHP appeared similarly to shift the equilibrium substantially towards the T state, but it seems also to decrease c slightly, namely the ligand affinity in the T state was a little lowered. Both the left ends of the curves formed by the data for Hb Yakima and Hb Kempsey join with the straight line representing n = 1 at the point corresponding to the mean pso value for the isolated (Y and /3 chains of Hb A (plo = 0.4 mmHg). As shown in Fig. 7, the data for the abnormal hemoglobins fit fairly well with the curve obtained by assuming pso of the R state as 0.4 mmHg and c as 0.04. The curve formed by the data for Hb A in the presence of IHP joins with the line representing n = 1 at pso of 100 mmHg which is considered to be the ligand affinity of the T state. If we assume that the pso of the R state is the same as that of isolated chains (0.4 mmHg), then the value for c becomes 0.004. The calculated curve using these parameters agrees well with the data for Hb A in the presence of IHP. Similarly, the data for Hb A in the absence of IHP can be explained by a simple two-state model assuming the same pso for the R state as above and a little smaller pso value for the T state than that in the presence of IHP (c = 0.01). The assumption that normal hemoglobin and abnormal hemo-Hill's n were observed only when the value for L was changed considerably by the concurrent effect of proton and IHP.
globins which have amino acid substitutions at (Y& contact region, all have the same oxygen affinity in the R state as that for isolated chains is consistent with the models proposed by Monod et al. (13) and Perutz (21) that the subunits in the R state suffer the least constraints. The result that the effect of proton on ligand affinity is only to change the allosteric equilibrium constant L, agrees with the mechanism proposed by Perutz (21). IHP is supposed to bind to hemoglobin in the same way as does 2,3-PI-glycerate and it actually appears to stabilize the T state. However, IHP seems to decrease the intrinsic ligand affinity in the T state, and the data for Hb A in the presence and absence of IHP cannot be fitted with a single c value as pointed out by Bunn and Guidotti (12).
Amino acid substitutions in Hb Yakima and Hb Kempsey are in the position of fl9 Asp, which is located at (Y& contact region and forms an important hydrogen bond with (~42 Tyr in the deoxygenated form and stabilizes deoxy conformation. Consequently substitution of the acidic aspartic acid residue by basic histidine in Hb Yakima and asparagine in Hb Kempsey, breaks this hydrogen bond and the allosteric equilibrium is shifted extremely toward the R state. In the presence of a potent allosteric effector, IHP, the equilibrium is shifted toward the T state and the shift is further strengthened by hydrogen ions. So, the cooperativity of Hb Kempsey and Hb Yakima is increased in the presence of IHP at low pH. According to Edelstein, L is proportional to the fourth power of cu( i/2), and allosteric equilibrium constant for normal hemoglobin is calculated as L = 4 x 10". This value is larger than that originally reported by Monod et al. (L = 9 x lOa) (13) and is nearly equal to that proposed by Edelstein (L = 3 x 10') (15).
As reported above, the abnormal function of (Y& contact anomalous hemoglobins is not a fixed one but in suitable conditions the lost cooperativity by amino acid substitution is restored. The cooperativity of Hb A is lost in certain conditions, where the allosteric equilibrium is shifted extremely. In both normal and abnormal hemoglobins, these changes in