Organic Phosphate Binding Inhibits High pH t-Isomerization of the â-Chain in Straw-coloured Fruit Bat (Eidolon helvum) Haemoglobin

Understanding the systematic structural changes accompanying allosteric effector binding to haemoglobin should provide some clues to the understanding of structure–function relationship in other multimeric enzymes. The affinities of the CysF9[93]â sulfhydryl group of oxy-, carbomonoxyand aquomet-derivatives of straw-coloured fruit bat (Eidolon helvum) haemoglobin (SCFB-Hb) for 5,5’-dithiobis(2-nitrobenzoate) (DTNB) were measured in the range 5.6 £ pH £ 9.0 using stripped and inositol hexakisphosphate (inosito-P6) bound haemoglobin. The data were analyzed on the basis of findings that the tertiary structure of the product of the reaction of DTNB with haemoglobin CysF9[93]â sulfhydryl group exists in two conformations; r and t. The result shows that the affinity of DTNB for SCFB-Hb in both r and t conformations are coupled to the ionizations of two ionizable groups, HisH21[143]â and HisFG4[97]â. In the r conformation, the presence of inositol-P6 reduces the pKa of HisH21[143]â by 1.24 units and that of HisFG4[97]â by 2.74. In the t conformation, inositol-P6 raises the pKa of HisH21[143]â by 1.10 pKa units whereas that of HisFG4[97]â was increased by 0.78 pKa units. Change in pKa of ionization of the ionizable groups and isomerization of the tertiary conformations are important modulators of organic phosphate binding.


Introduction
][3] This reaction occurs reversibly.Therefore, the r↔t transition which occurs at the tertiary level is also in dynamic equilibrium when DTNB is bound to the CysF9[93]â sulfhydryl group.It must be stated that this is different from R (relaxed) and T (tensed) quaternary conformations, formed, respectively, in the absence and presence of salt bridges.In the r state, the sulfhydryl group is cis to the terminal amino group, but cis to the carbonyl end in the t state.Transitions between r and t tertiary conformations are accompanied by change in the pK a s of ionization of the ionizable groups that are linked to the reaction of DTNB with the CysF9[93]â haemoglobin sulfhydryl group. 4he affinity of human haemoglobin for oxygen has been shown to be regulated by the organic phosphate, 2,3-bisphosphoglycerate (2,3-BPG). 5-82,3-BPG is a small molecule with a high density of negative charge.It is an allosteric effector which can bind to ligated and unligated haemoglobin in a process that is believed to be proton driven and involves oxygen uptake. 70] ValNA1 [1]â, HisNA2 [2]â, LysEF6[82]â and HisH21[143]â have been implicated in the binding of 2,3-BPG to haemoglobin. 9The same residues have been reported to bind inositol-P 6 , an organic phosphate, which like 2,3-BPG, has a high density of negative charge. 6,11Increase in pH has been reported to reduce the binding affinity of 2,3-BPG.This finding has been attributed to ionization of the cationic groups that become neutral side chains.
Bats are the only known flying mammals.Though the primary structure of the haemoglobin of a number of bat species has been determined, the primary structure of SCFB-Hb has not been characterized.It is expected that like human haemoglobin, inositol-P 6 should also alter the tertiary structure of SCFB-Hb.However, no experimental data have been reported on the magnitude and nature of change in tertiary or quaternary structure of any species of bat haemoglobin due to organic phosphate binding.
While organic phosphate binding is known to alter the pK a of the ionizable groups of a specific tertiary structure (r or t) due to quaternary structure changes, the pK a changes of either stripped or organic phosphate bound haemoglobin should arise from r↔t transition (tertiary level changes) only.If however, binding of organic phosphate results in significant tertiary conformation changes in the haemoglobin structure, the change in pK a resulting from the r↔t transition in stripped haemoglobin should be significantly different from that in inositol-P 6 bound haemoglobin.Value of K rt , the equilibrium constant of isomerization given by K rt = (Hb 4 ) t /(Hb 4 ) r in organic phosphate bound haemoglobin should also be significantly different from its value in the stripped haemoglobin.
Hitherto, the way by which organic phosphates perform the function of regulating oxygen binding has not been fully understood.It is with the aim to quantifying the magnitude of tertiary level conformation changes and understanding the sequence of processes culminating in the binding to, and dissociation of organic phosphate from haemoglobin that we embarked on the study reported herein.

Preparation of Haemoglobin
2][3] The major difference here is in the use of 11.5 g dm -3 saline solution rather than 9.5 g dm -3 at 5 °C for washing the red blood cells.2][3] Full conversion to the carbonmonoxy derivative was confirmed with previously obtained spectrum of carbonmonoxyhaemoglobin. 4 Conversion to aquomethaemoglobin was carried out by oxidation of oxyhaemoglobin with two-fold molar excess K 3 Fe(CN) 6 .Each derivative was passed through a Dintzis ion exchange column to remove the endogenous ions. 3,12The concentration of the haemoglobin derivatives were determined as described earlier. 3

Determination of p-(Hydroxymercuri)benzoic Acid (p-MB) Reactive Sulfhydryl Group
The number of p-MB reactive sulfhydryl groups was determined four times by titration of 5 µmol dm -3 carbonmonoxyhaemoglobin tetramer in phosphate buffer pH 7.6, ionic strength 50 mmol dm -3 (NaCl) with 420 µmol dm -3 p-MB as described by Boyer, et.al. 13

Determination of DTNB Reactive Sulfhydryl Group
The number of sulfhydryl groups reacting with DTNB was determined by titration of the 2.5 µmol dm -3 (tetramer) carbonmonoxyhaemoglobin, four times in phosphate buffer pH 7.6, ionic strength 50 mmol dm -3 (NaCl) with 0.5 mmol dm -3 DTNB stock solution as described elsewhere. 1,14

Determination of Equilibrium Constant of Reaction of Stripped Haemoglobin with DTNB
Stock DTNB used for equilibrium studies was prepared as described earlier. 2 Method used for the determination of equilibrium constant had been previously described. 2The mean of at least six replicate equilibrium constant determinations at each pH of the experiments was calculated.The pH of the reaction solutions was determined at the end of each experiment.

Determination of the Equilibrium Constant in the
Presence of Inositol-P 6  2.5 × 10 -6 dm 3 of 50 mmol dm -3 stock inositol-P 6 which had been titrated to pH 6.7 with 1.0 mol dm -3 HCl was added to haemoglobin to make inositol-P 6 final concentration 50 × 10 -6 mol dm -3 in a 25 cm 3 volumetric flask.The equilibrium reaction with DTNB were carried out as described earlier. 2 The mean equilibrium constant values of at least six replicate experiments at each pH of reaction was determined.

Calculation of Equilibrium Constant from Experimental Data
Various equilibria accompanying the reaction between haemo-globin sulfhydryl group and DTNB are described by Equation 1below: K EQ is related to the other quantities in Equation 1 by: A detailed derivation of Equation 2 has been described before. 2n Equations 1 and 2 PSH is the haemoglobin with CysF9[93]â in its protonated form which does not react with non-mercurial sulfhydryl reagents; [14][15][16] PS -is the corresponding thiol anion form which reacts with DTNB; PS.TNB is the mixed disulfide, the product of reaction of PS -with DTNB; TNB -is the chromophoric product of DTNB reaction with haemoglobin, whose concentration is determined spectrophotometrically at 412 nm.The protonated form of TNB -is TNBH; Q SH and Q TNB are the ionization constants of CysF9[93]â and TNBH, respectively; K EQ is the equilibrium constant for the formation of the mixed disulfide (PS.TNB) from the reaction of DTNB with haemoglobin sulfhydryl group.
7][18] We therefore, assumed a pQ SH value of 8.3.These Q SH and Q TNB values together with the TNB -concentration measured at 412 nm from the absorbance of TNB -at specific pH values were substituted into Equation 2 to obtain K EQ .It is assumed that the absorbance of TNBH at 412 nm is insignificant. 2 The standard error in the determination of K EQ is about 10 %.

Fitting of Experimental Data to Various Equilibria Steps in the Theoretical Model
The number and pK a of the ionizable groups linked to the pH dependence of the K EQ of the reaction of DTNB with the haemoglobin were analyzed based on previous findings that: (i) CysF9[93]â sulfhydryl group of liganded haemoglobin exist in two tertiary conformations; cis to the terminal amino group (r-conformation) and cis to the carbonyl group (t-conformation), [20][21] and (ii) that these two sulfhydryl conformations are coupled to transitions between the two tertiary structures in dynamic equilibrium. 4,22In the experiment reported here, the reaction of DTNB with three derivatives of liganded stripped haemoglobin (haemoglobin that is free of organic phosphate) and three derivatives of inositol-P 6 bound haemoglobin were studied.][25] In determining the number and the nature of the DTNBlinked ionizable groups, we employed Scheme 1. 3,[25][26][27][28] Scheme 1 is more detailed form of Equation 1 in which the ionization of groups on the haemoglobin species are shown.In Scheme 1, n is the total number of ionizable groups linked to the equilibrium reaction of DTNB with the haemoglobin sulfhydryl group.H n-i+1 PSH (i = 1, 2 …, n+1) denotes the various protonated forms of the haemoglobin which do not react with DTNB and are therefore omitted from Scheme 1.The protons arising from the ionization of the ionizable groups linked to the reaction of DTNB with the haemoglobin sulfhydryl group are also omitted for clarity.H n-i+1 PS -(i = 1, 2 …, n+1) are species in which the sulfhydryl group is in its thiolate anion form, the form that reacts with DTNB; Species marked with subscripts 'r ' and 't' are those in which the haemoglobin sulfhydryl is in the r-and t-tertiary
The relationship between K EQ and the parameters in Scheme 1 is given by: 4 Equation 3 was used to fit the dependence of -logK EQ on pH.The value of n (in Equation 3) that gives the best fit of the calculated curve through the experimental data points is the number of the ionizable groups linked to the equilibrium reaction of haemoglobin sulfhydryl group with DTNB.The best fit values of pQ ir and pQ it give the ionization constants of the ionizable groups in the 'r' and 't' tertiary conformations, respectively.K Ei {i = 1, 2 …(n+1)} are the equilibrium constants of each reacting species.Fitting of the experimental data to Equation 3 were performed using Micromath Scientist software (Salt Lake City, Utah) with a script of the relationship between K EQ and pH according to Equation 3.

Number of Reactive Sulfhydryl Groups
The amino acid sequence of SCFB-Hb has not been previously reported.However, the haemoglobin amino acid sequences of the bat species which had been previously characterized showed that: Egyptian fruit bat (Rousettus aegyptiacus) haemoglobin has sulfhydryl groups at positions G11[104]á, F9[93]â and G14[112]â; 29  (3) G14[112]â, and F9[93]â. 32Of these sulfhydryl groups, only CysG11[104]á and CysG14 [112]â are known to be masked to both DTNB, and p-MB.We therefore expect SCFB-Hb, being a member of the same family, to have reactive sulfhydryl group at position F9[93]â, in addition to one or two other positions.Surprisingly, Boyer titration of the haemoglobin with p-MB repeated three times gave a mean value of 2.05 ± 0.12 sulfhydryl group per tetramer (figure not shown).
Also, DTNB titrations of SCFB-Hb repeated three times (figure not shown) gave a mean value of 2.02 ± 0.06 titratable sulfhydryl groups.This shows that bat haemoglobin has only two reactive sulfhydryl groups per tetramer.6,27 We therefore assigned the only pair of p-MB and DTNB reactive sulfhydryl group in the bat haemoglobin to position F9[93]â.

Reaction of DTNB with SCFB CysF9[93]â of Oxyhaemoglobin
The strong pH dependences of the negative logarithm to base 10 of the equilibrium constant (-log 10 K EQ ) of the reaction of the stripped oxyhaemoglobin and the inositol-P 6 bound SCFB oxyhaemoglobin with DTNB are presented in Fig. 1.Each data point is obtained from the mean value of at least six replicate experiments.The curves through the experimental data points were best fit curves using Equation 3 (cf.Scheme 1) with n = 2, together with the fitting parameters reported in Table 1 column 2. The fitting parameters of Table 2 column 2 using Equation 3(cf.Scheme 1) with n = 2 report the data of inositol-P 6 bound oxyhaemoglobin.The equilibrium constant values of both the stripped and the inositol-P 6 bound haemoglobin are quite similar in the range 5.6 £ pH £ 7.0.Above pH 7.0, however, the equilibrium constant of the reaction of stripped haemoglobin gets increasingly greater than that of the inositol-P 6 bound haemoglobin with increasing pH.

Reaction of DTNB with SCFB CysF9[93]â of Carbonmonoxyhaemoglobin
The experimental data points in Fig. 2 were fitted as described in section 4.2 above, using the fitting parameters reported in Table 1 column 3 for the stripped carbonmonoxyhaemoglobin and that reported in Table 2 column 3 for the inositol-P 6 bound carbonmonoxyhaemoglobin.The curve of the pH dependences of the -log 10 K EQ of the reaction of DTNB with stripped and inositol-P 6 bound SCFB carbonmonoxyhaemoglobin are both sigmoidal (Fig. 2).Below pH 7.3, the K EQ values of the stripped haemoglobin were lower than that of the inositol-P 6 bound carbonmonoxyhaemoglobin at equivalent pHs.In the range 7.3 £ pH £ 7.7, the values of the K EQ of both stripped and inositol-P 6 bound haemoglobin were quite comparable at identical pHs.Above pH 8.0, the K EQ values of the stripped haemoglobin are somewhat higher than that of the organic phosphate bound haemoglobin at the corresponding pHs.

Reaction of DTNB with SCFB CysF9[93]â Aquomethaemoglobin
The experimental data of the pH dependences of -log 10 K EQ for the reaction of DTNB with stripped aquomethaemoglobin are plotted along with that of reaction of DTNB with inositol-P 6 bound haemoglobin for comparison (Fig. 3).The K EQ value is strongly dependent on pH.The data were well fitted with n = 2 as described in sections 4.2 and 4.3.The curve through the experimental data points of stripped haemoglobin reaction with DTNB was fitted with parameters reported in Table 1 column 4. That of inositol-P 6 bound aquomethaemoglobin reaction was fitted with the parameters reported in Table 2 column 4. It is evident that the K EQ values for the reaction of the stripped aquomethaemoglobin with DTNB are greater than that of the reaction of inositol-P 6 bound haemoglobin with DTNB at equivalent pH, over the entire pH range of the experiment.However, the differences in equilibrium constant of the reaction were minimized in the range 7.3 £ pH £ 7.5.

pK a Values and K rt , the Equilibrium Constant of r↔t Transition
In order to calculate the equilibrium constant for r↔t transition, K rt , we made use of Scheme 1.According to which at equilibrium: -log 10 K rt2 = pQ 2r + (-log 10 K rt3 ) -pQ 2t (5)   and -log 10 K rt1 = pQ 1r + (-log 10 K rt2 ) -pQ 1t (6)   For the stripped haemoglobin, all the parameters on the right side of Equation 5were obtained from the mean value of the fitting parameters of the experimental data in Table 1.For inositol-P 6 bound haemoglobin, the value of the parameters on the right side of Equation 5were obtained from the mean values of the fitting parameters in Table 2. Once K rt2 was determined using Equation 5, its value and Equation 6with appropriate parameter from the mean values reported in Tables 1 or 2 (as applicable) was used to calculate K rt1 .The values of K rti (i = 1, 2 and 3) so calculated for stripped haemoglobin were presented in Table 3.Similar data for inositol-P 6 bound haemoglobin are summarized in Table 4.

How does Inositol-P 6 affect the Equilibrium Constant of Isomerization, K rt ?
The reaction of all the three SCFB haemoglobin derivatives were carried out under identical pH conditions and were analyzed based on previous finding by Shaanan, 21 that human oxyhaemoglobin exists as a mixture of two tertiary conformations (rand t-) which are in dynamic equilibrium in solution. 21We suggest that under a given condition, the pK a of the ionizable groups linked to these two haemoglobin conformations will be different if the isomerization process results in change in the environment of the ionizable groups.By similar hypothesis, the pK a of ionizable groups linked to the equilibrium reaction of

DTNB with CysF9
[93]â sulfhydryl group in stripped haemoglobin should be significantly different from that of inositol-P 6 bound haemoglobin.Inositol-P 6 , a molecule with high density of negative charge should create an environment which is different from that in the stripped haemoglobin around the ionizable groups.
It must be noted that whereas, the pK a values obtained for the two ionizable groups of the stripped haemoglobin were raised as a result of r→t isomerization in the kinetic experiment carried out earlier with SCFB haemoglobin, 33 only the first ionizable group has its pK a raised in the equilibrium constant experiment.In addition, the K rt3 value obtained in the equilibrium experiment at high pH is about 40 times greater than that determined using kinetic data. 33This latter discrepancy can be the consequence of incomplete binding of DTNB to haemoglobin in kinetic reaction compared to equilibrium experiment where DTNB binding to haemoglobin is maximum.We therefore, submit that at high pH, DTNB binding favours conversion of stripped haemoglobin from r-isomer to t-isomer.The result also suggests that as the DTNB gets bound to the haemoglobin, the â-chains might get progressively converted from r conformation to the t conformation.This discovery is consistent with the observation that the mean K rt3 (the K rt at high pH) of inositol-P 6 bound haemoglobin is ca.three orders of magnitude lower than the K rt3 of stripped haemoglobin (compare the mean K rt3 in Tables 3 & 4).This suggests that organic phosphate binding might be acting by strongly favouring r-isomer formation as opposed to t-isomer at high pH.The following observations from the comparison of Tables 3 and 4 are also noteworthy: (i) low pH had a drastic effect of converting the haemoglobin almost entirely to the t-isomer in both stripped and inositol-P 6 bound haemoglobin; (ii) at intermediate pH, in the stripped haemoglobin, the t-isomer is about 39 % which is reduced to about 30 % t-isomer in organic phosphate bound haemoglobin; (iii) at high pH, whereas, the population of t-isomer in stripped haemoglobin is as high as ca.80 %, in organic phosphate bound haemoglobin, the population of t-isomer is less than 0.1 %.This suggests that inositol-P 6 acts by strongly favouring the t-isomer at low pH, and progressively favouring r-isomer with increasing pH.This is indicative of pH dependent organic phosphate binding for physiological function.It also shows that the organic phosphate acts by altering the equilibrium between r and t tertiary conformation of the haemoglobin chains.

Assignment of the First Ionizable Group
The values of the mean pQ 1r (in the r-isomer) and pQ 1t (in the t-isomer) are 6.56 and 9.36, respectively, for the first ionizable group of the stripped haemoglobin.The corresponding values of pQ 1r and pQ 1t for the inositol-P 6 bound haemoglobin are 5.32 and 10.46, respectively.Based on previous assignment in human carbonmonoxyhaemoglobin, 34 the ionizable group was assigned to HisH21[143]â.The assignment made in this report was based on the presumption that the bat haemoglobin also possess Histidine at H21[143]â position.The mean pK a of this ionizable group in human carbonmonoxyhaemoglobin as determined using 1 H NMR technique was 5.65. 34n the bat stripped haemoglobin, the implication of these results is that isomerization from r-to t-conformation raises the pK a of the first ionizable group by about 2.80 pK a units.Similar isomerization of the inositol-P 6 bound haemoglobin leads to 5.14 pK a unit increase in the same ionizable group.This is almost double the change in pK a of the ionizable group resulting from isomerization from r-to t-tertiary conformation observed in the stripped haemoglobin.This suggests that organic phosphate binding increase the change in pK a of isomerization.This may not be unconnected with the strong binding of organic phosphate to HisH21[143]â site.The increased change in pK a in presence of organic phosphate was achieved by lowering the pK a of ionization of HisH21[143]â in the r-isomer and raising it (making it more difficult for it to ionize) in the t-isomer, compared to the stripped haemoglobin.The values of the pK a s of this ionizable group in the two isomeric forms of both stripped and inositol-P 6 bound haemoglobin, suggest that in t-isomer, the ionizable group might be brought nearer to an anionic side chain which stabilizes the protonated form of the histidine as opposed to the deprotonated form.On the other hand, in the r-isomer conformation, HisH21[143]â might be brought nearer a cationic side chain or a neutral residue.This can at least semi quantitatively account for the higher value of the pK a of this ionizable group in t-isomer of the organic phoshphate bound haemoglobin compared the stripped haemoglobin.Judging from close examination of the â-chain three-dimensional structure of haemoglobin, the HisH21[143[â] of SCFB haemoglobin, should be located near AspG1[99]â and Glu3[101]â.HisH21[143[â]  should be brought closer to these negatively charged groups when salt bridge is formed between HisHC3[146]â and AspFG1[94]â in the presence of inositol-P 6 .This is supposed to significantly raise the pK a of the HisH21[143[â in the presence of organic phosphate in agreement with our findings.We therefore suggest that in the presence of inositol-P 6 , increase in pK a change on isomerization from r-to t-isomer compared to stripped haemoglobin arises from HisH21[143[â being brought nearer to

Assignment of the Second Ionizable Group
The mean pK a s of the second ionizable group in the stripped haemoglobin described by pQ 2r and pQ 2t are 8.42 and 7.61, respectively.This suggests that r-to t-isomerization results in 0.81 pK a unit reduction.In the presence of saturating amount of inositol-P 6 on the other hand, isomerization from r-to t-conformation results in change in pK a of the second ionizable from 5.68 to 8.39.This translates to 2.71 pK a unit increase.These values show that pK a change in the second ionizable group arising from r↔t isomerization in inositol-P 6 bound haemoglobin is about 3.3 times greater than that of stripped haemoglobin.Again, we see here that inositol-P 6 significantly increases the change in pK a of r→t isomerization compared to stripped haemoglobin.
Based on previous assignment, 36 we apportioned the ionizable group characterized by pQ 2r and pQ 2t to the Histidine group at position FG4[97]â.0][31][32] These residues should raise the pK a of histidine above the typical value of 6.0 for an isolate histidine residue.This is consistent with the pK a values assigned to FG4[97]â in the stripped haemoglobin.Also, the mean pK a value of 5.68 obtained in the presence of organic phosphate can be rationalized based on salt bridge formation between HisHC3[146]â and AspFG1[94]â.Organic phosphate favour r-isomer at high pH and is expected to bring HisFG4[97]â, a group that must have been completely ionized at high pH, nearer to LysFG2[95]â, a positively charged amino residue with pK a value of 10.53 for the isolated group.This should lower the pK a of HisFG4[97]â and stabilize the r-isomer as observed in this experiment.The three-dimensional structure of haemoglobin reveals that FG4[97]â position is within ionic atmosphere of F9[93]â position.Its reaction should therefore be linked to FG4[97]â position.The pK a of HisFG4[97]â previously reported for liganded mammalian haemoglobin is ca.7.75. 36This value is in good agreement with pQ 2t value of 7.61 obtained in our experiment.

Effect of Equilibrium Constant of Isomerization, K rt values on Organic Phosphate Binding
The results reported in this paper shows that greater increase in the pK a of the fist ionizable group occurred when the â-chains changes from r-to t-conformation in the presence of inositol-P 6 , than in the stripped haemoglobin.This suggests that the isomerization process is important in regulating the binding affinity of organic phosphate.In stripped haemoglobin, at low pH, the proportion of t-isomer is ca.99.75 %.At high pH, it is barely reduced to about 80.5 % (see Table 3).Since the proportion of t-isomer in the haemoglobin â-chain of stripped haemoglobin at both low and high pH is high, and the pK a of HisH21[143]â in the t-isomer is high (ca.9.36) and HisH21[143]â remains essentially protonated over the entire pH of the experiment.The consequence of this is that the affinity of the stripped haemoglobin for organic phosphate remains essentially high at both low and high pH.
In the presence of organic phosphate however, at low pH, the proportion of t-isomer was ca. 100 %.This proportion reduces drastically with increasing pH, with the r-isomer rising to ca. 99.92 % at high pH.This ensures that HisH21[143]â is almost entirely protonated (positively charged) at low pH, but mostly deprotonated (neutral) at high pH.The consequence of which is high affinity of the haemoglobin for organic phosphate (reduced affinity for oxygen) at low pH.Protonation of histidine residue of the haemoglobin at low pH, increased affinity of haemoglobin for organic phosphate and the attendant reduction in the affinity of haemoglobin for oxygen leads to increased physiological pH.Increase in the proportion of r-isomer (the low pK a form of HisH21[143]â) accompanying increase in pH in the presence of organic phosphate, ensures that increase in pH results in the ionization of HisH21[143]â.The consequence of this is lowering of the affinity of HisH21[143] for the organic phosphate with increasing pH.This may be used to account for the reported cooperativity of oxygen binding to haemoglobin.Previous findings have shown that HisH21[143] makes the most contribution (ca.71 %) to the acid Bohr effect at pH 5.1. 34

Equilibrium Constant of Isomerization, K rt Values and HisFG4[97]â
As noted in the previous section for the stripped haemoglobin, the t-isomer dominates at both low and high pH.The pK a of HisFG4[97]â in the t-isomer of the stripped haemoglobin is ca.7.61.This value is barely lower than the pK a value of 8.42 for the same ionizable group in the r-isomer.Therefore, HisFG4[97]â remains essentially protonated in the physiological pH range (pH 6.9-7.2).Based on our data, it can only be deprotonated (converted to the neutral form) above pH 7.62.The implication of this is higher affinity of the ionizable group for organic phosphate at acidic pH and over the entire physiological pH range.Again, this shows that in the absence of organic phosphate, HisFG4[97]â contribute to higher affinity of the haemoglobin for organic phosphate at physiological pH.
However, in the presence of inositol-P 6 , as observed earlier, the t-isomer population of ca. 100 % at low pH reduces sharply to a population of less than 0.1 % at high pH (the proportion of risomer rises to about 99.9 %).It should again be noted that while r-isomer of this ionizable group is characterized by low pK a value of 5.68 in the inositol-P 6 bound haemoglobin, the t-isomer has a pK a value of 8.39.This shows that at low pH (below pH 5.68), the protonated form of the residue with high affinity for organic phosphate dominate.Whereas, at high pH, the deprotonated form of the ionizable group which has low affinity for organic phosphate dominates, if the dominating conformation is the r-isomer.It should be emphasized from this findings that the stripped haemoglobin possesses high affinity for organic phosphate at low and high pH.Lowering of the affinity of HisFG4[97]â at high pH, in the presence of organic phosphate is achieved because r-conformation with low pK a is favoured.This is a further indication that isomerization is critical for the regulation of organic phosphate binding to haemoglobin.

Conclusion
This study has shown quantitatively that isomerization process is important in regulating organic phosphate binding, and by extension oxygen binding to haemoglobin under different pH conditions.We demonstrated that the pK a change on isomerization from r-to t-conformation for the first ionizable group is 1.84 times greater in the presence of organic phosphate bound haemoglobin than the stripped haemoglobin.In the second ionizable group, the pK a change is 3.34 times greater in organic phosphate bound haemoglobin than in stripped haemoglobin.This is a strong indication that organic phosphate regulates the function of haemoglobin by altering both the pK a of the ionizable groups and altering the equilibrium constant of isomerization.
The good quality of the theoretical fit to the experimental data with only two ionizable groups; (HisH21[143]â and HisFG6[97]â) in isomerizing â-chain tertiary conformations suggests that HisEF1[77]â the third ionizable group which was present in human haemoglobin might be either absent or unable to ionize in SCFB-Hb.1][32] At present, amino acid sequences of SCFB haemoglobin has yet to be characterized for any categorical statement about this to be made.It would therefore be necessary to carryout similar experiment using other mammalian haemoglobin that are known to possess histidine at position EF1[77]â for a more definite conclusion to be made.

Figure 1
Figure 1 Reaction of 5,5'-dithiobis(2-nitrobenzoate), DTNB, with the CysF9[93]â sulfhydryl group of SCFB oxyhaemoglobin.Dependence of the equilibrium constant on pH: stripped oxyhaemoglobin, open circle and broken curve; inositol-P 6 bound oxyhaemoglobin, filled circle and full curve.Conditions: 30 °C; phosphate buffers, pH 5.6-7.8;borate buffers, pH 8.0-9.0;ionic strength, 50 mmol dm -3 ; haemoglobin concentration, 25 µmol dm -3 in reacting sulfhydryl groups, that is 50 µmol (heme) dm -3 .Each data point is the mean of at least six replicate experiments and is subject to a standard error of about ± 0.1 in the logarithm unit.The curves through the data points are the best-fit theoretical lines calculated with Equation 3 in the text with n = 2 (cf.Scheme 1)

Figure 2
Figure 2 Reaction of 5,5'-dithiobis(2-nitrobenzoate), DTNB, with the CysF9[93]â sulfhydryl group of SCFB carbonmonoxyhaemoglobin. Dependence of the equilibrium constant on pH: stripped carbonmooxyhaemoglobin, open circle and broken curve; inositol-P 6 bound carbonmonoxyhaemoglobin, filled circle and full curve.Conditions were as described for Fig. 1.Each data point is the mean of at least six replicate experiments and is subject to a standard error of about ± 0.1 in the logarithm unit.The curves through the data points are the best-fit theoretical lines calculated with Equation 3 in the text with n = 2 (cf.Scheme 1).

Figure 3
Figure 3 Reaction of 5,5'-dithiobis(2-nitrobenzoate), DTNB, with the CysF9[93]â sulfhydryl group of SCFB aquomethaemoglobin.Dependence of the equilibrium constant on pH: stripped aquomethaemoglobin, open circle and broken curve; inositol-P 6 bound aquomethaemoglobin, filled circle and full curve.Conditions were as described for Fig. 1.Each data point is the mean of at least six replicate experiments and is subject to a standard error of about ± 0.1 in the logarithm unit.The curves through the data points are the best-fit theoretical lines calculated with Equation 3 in the text for n = 2 (cf.Scheme 1).

Table 1
Reaction of DTNB with stripped haemoglobin: parameters of the best-fit curves through the open circles in Figs.1-3 using Scheme 1 and Equation 3 in the text with n = 2.
The last column gives the values of the parameter ± standard deviation (S.D.).

Table 2
Reaction of DTNB with haemoglobin in the presence of inositol-P 6 : parameters of the best-fit curves through the filled circles in Figs.1-3using Scheme 1 and Equation 3 in the text with n = 2.
The last column gives the values of the parameter ± standard deviation (S.D.).

Table 3
Mean isomerization data of the different ionization species of stripped SCFB haemoglobin.Proportion of t-isomer in the various ionization states are shown in brackets.

Table 4
Mean isomerization data of the different ionization species of SCFB haemoglobin in the presence of inositol-P 6 .Proportion of t-isomer in the various ionization states are shown in brackets.