Allosteric modifiers of hemoglobin: 2-[4-[[(3,5-disubstituted anilino)carbonyl]methyl]phenoxy]-2-methylpropionic acid derivatives that lower the oxygen affinity of hemoglobin in red cell suspensions, in whole blood, and in vivo in rats.

Two new potent allosteric effectors of hemoglobin, RSR-4 [2-[4-[[(3,5-dichloroanilino)carbonyl]-methyl]phenoxy]-2- methylpropionic acid] and RSR-13 [2-[4-[[(3,5-dimethlanilino)carbonyl]methyl]-phenoxy]-2-methylp rop ionic, are compared to the previously reported compounds L3,5 and L3,4,5 [Lalezari, I., Lalezari, P., Poyart, C., Marden, M., Kister, J., Bohn, B., Fermi, G., & Perutz, M. F. (1990) Biochemistry 29, 1515]. Unlike L3,5 and L3,4,5, RSR-4 and RSR-13 are less impeded by physiological concentrations of serum albumin. RSR-4 has also been shown to be more effective than L3,5 in shifting the allosteric equilibrium of bovine Hb toward the low-affinity T-state. X-ray crystal studies show that both RSR-4 and RSR-13 bind to only one pair of symmetry-related sites in the Hb central water cavity whereas previous studies on L3,5 and L3,4,5 demonstrated a second pair of symmetry-related binding sites near Arg 104 beta. Three major interactions between these allosteric effectors and Hb include the acid group with the guanidinium group of C-terminal Arg 141 alpha, the effector's amide oxygen with the ammonium ion of Lys 99 alpha, and the phi electrons of the halogenated or methylated aromatic ring and Asn 108 beta. No explanation has been found for the difference in number of binding sites observed for RSR-4 and RSR-13 (two sites) compared to L3,5 and L3,4,5 (four sites); also no correlation has been made between the number of binding sites and degree of allosteric shift in the oxygen equilibrium curve.(ABSTRACT TRUNCATED AT 250 WORDS)     

X-ray diffraction study of di and tetra-ligated T-state hemoglobin from high salt crystals.

X-ray diffraction difference electron density maps at 3 A resolution obtained from di and tetra-ligated T-state hemoglobin (Hb) crystals are reported. Crystals isomorphous with native deoxyhemoglobin were obtained from ammonium sulfate solutions incubated with the synthetic allosteric effector RSR-56. RSR-56 binds at two symmetry-related Hb central water cavity sites and each molecule has major interactions with three different subunit side-chains; one effector with Arg141 alpha 2 HC3, Lys99 alpha 1 G6 and Asn108 beta 1 and the other with the symmetry related residues, Arg141 alpha 1 Lys99 alpha 2 and Asn108 beta 2. Crystals mounted in a nitrogen filled glove box were di-ligated as previously found with polyethyleneglycol Hb crystals. Crystals mounted in air under a layer of mother liquor were bright red and showed all four heme groups ligated. The difference electron density from the di-ligated crystals showed atomic movements to be restricted to the immediate neighborhood of the heme groups and the allosteric effector. By contrast, the tetra-ligated structure showed extended difference electron density near amino acid residues around both alpha and beta heme groups and along the alpha 1/beta 2 interface. Ligation of the beta heme group appears to magnify the difference density around the alpha heme groups. There is no evidence of breakage of the Bohr salt bridge, His146 beta HC3----Asp94 beta FG1, in the crystal. The observed difference electron density maps may help to clarify the way the allosteric mechanism is triggered.     

R-state hemoglobin bound to heterotropic effectors: models of the DPG, IHP and RSR13 binding sites

We performed a docking study followed by a 500-ps molecular dynamics simulation of R-state human adult hemoglobin (HbA) complexed to different heterotropic effectors [2,3-diphosphoglycerate (DPG), inositol hexaphosphate (IHP), and 2-[4-[(3,5-dichlorophenylcarbamoyl)-]methyl]-phenoxy]-2-methylpropionic acid (RSR13)) to propose a molecular basis for recently reported interactions of effectors with oxygenated hemoglobin. The simulations were carried out with counterions and explicit solvation. As reported for T-state HbA, the effector binding sites are also located in the central cavity of the R-state and differ depending on effector anionic character. DPG and IHP bind between the alpha-subunits and the RSR13 site spans the alpha1-, alpha2- and beta2-subunits. The generated models provide the first report of the molecular details of R-state HbA bound to heterotropic effectors.     

Hemoglobin Thionville. An alpha-chain variant with a substitution of a glutamate for valine at NA-1 and having an acetylated methionine NH2 terminus.

In hemoglobin (Hb) Thionville, the substitution of a glutamic acid for the alpha-chain NH2-terminal valine inhibits the cleavage of the initiator methionine which is then acetylated. The elongation of the alpha-chain NH2 terminus modifies the three-dimensional structure of hemoglobin at a region that is known to have an important role in the allosteric regulation of oxygen binding. Relative to Hb A, Hb Thionville has a lower affinity for oxygen, and the heterotropic allosteric effects of protons, chloride, and bezafibrate are reduced. In contrast, the response to 2,3-diphosphoglycerate is normal. Analysis of oxygen equilibrium data within the framework of the two-state allosteric model indicates that the structure of deoxy Hb Thionville is stabilized relative to that of deoxy Hb A. The x-ray crystal structure of deoxy Hb Thionville shows that the glutamate side chain extends away from the alpha 1-alpha 2 interface, whereas the methionine side chain (which has two conformations) extends into the alpha 1-alpha 2 interface, physically displacing chloride and bezafibrate. The increased stability of deoxy Hb Thionville is due to new intrasubunit and intersubunit contacts made by the methionine. These interactions replace the indirect contacts, made through bound chloride ions, that Val-1 alpha normally contributes to the alpha 1-alpha 2 interface.     

Polymerization of deoxyhemoglobin CHarlem (β6 Glu → Val, β73 Asp → Asn): The effect of β73 asparagine on the gelation and crystallization of hemoglobin

The kinetics of aggregation and the solubility of deoxy Hb2 C_Harlem (α₂β₂ 6 Val, 73 Asn) in concentrated phosphate buffers were studied in comparison with those of deoxy Hb S and deoxy Hb A. Deoxy Hb C_Harlem aggregated with a clear exhibition of a delay time. The length of the delay and aggregation times and the degree of the aggregation depended upon the initial hemoglobin concentration.The initial hemoglobin concentration required for the aggregation of deoxy Hb C_Harlem was approximately 200% of its solubility, a value much higher than that required for the aggregation of deoxy Hb S (120%). With the same hemoglobin concentration, the delay time for the aggregation of deoxy Hb C_Harlem was approximately 100 times longer than that of deoxy Hb S. The logarithmic plotting of the delay time versus hemoglobin concentration in 1.8 m-phosphate buffer (pH 7.4) showed linear lines with a slope (n) of 4.0 for deoxy Hb C_Harlem. In contrast to the results for the aggregation of deoxy Hb S, n values for deoxy Hb C_Harlem were unchanged with phosphate concentrations varying from 1.2 m to 2.0 m. The solubilities of deoxy Hb S and deoxy Hb C_Harlem were increased exponentially by lowering the pH of the medium, with the increase being more conspicuous for Hb C_Harlem. The gels (or aggregates) of Hb C_Harlem were converted to crystals at a rate much faster than were those of Hb A and Hb S. The kinetics for gelation and crystallization of deoxy Hb C_Harlem can be explained by the following scheme, where nuclei G and nuclei C are formed before gelation and crystallization, respectively. Monomenc deoxy Hb

The hemoglobin concentration required for the crystallization of deoxy Hb C_Harlem was about ten times lower than that required for deoxy Hb A. The solubility of deoxy Hb C_Harlem after aggregation was about twice that of deoxy Hb S, suggesting that the substitution of Asn for Asp at the β73 residue inhibits the formation of nuclei G and accelerates the formation of nuclei C.     

Structure of deoxy hemoglobin C (beta six Glu replaced by Lys) in two crystal forms

Five crystal forms of the abnormal human hemoglobin Hb³ C (beta six Glu → Lys) have been grown. Two of them are grown with liganded Hb C, three with deoxy Hb C. The structures of two of the deoxy crystal forms were determined by the method of molecular replacement, using deoxy Hb A as the model structure. Fourier maps were calculated for each Hb C structure, using data to a resolution of 5 Å in one case and 4 Å in the second case. The structural differences between each deoxy Hb C structure and the deoxy Hb A model are found mostly at the molecular surface. Energetically favorable interactions involving the variant residue, beta six lysine, occur in both Hb C crystal forms, and could explain the lowered solubility and enhanced tendency of deoxy Hb C to crystallize in vivo.     

Indefinite noncooperative self-association of chicken deoxy hemoglobin D

The minor tetrameric hemoglobin (Hb), Hb D, of chicken red blood cells self-associates upon deoxygenation. This self-association enhances the cooperativity of oxygen binding. The maximal Hill coefficient is greater than 4 at high Hb concentrations. Previous measurements at low Hb concentrations were consistent with a monomer-to-dimer equilibrium and an association constant of ～1.3-1.6 × 10(4) M(-1). Here, the Hb tetramer is considered as the monomer. However, new results indicate that the association extends beyond the dimer. We show by combination of Hb oligomer modeling and sedimentation velocity analyses that the data can be well described by an indefinite noncooperative or isodesmic association model. In this model, the deoxy Hb D associates noncooperatively to give a linear oligomeric chain with an equilibrium association constant of 1.42 × 10(4) M(-1) at 20°C for each step. The data are also well described by a monomer-dimer-tetramer equilibrium model with monomer-to-dimer and dimer-to-tetramer association constants of 1.87 and 1.03 × 10(4) M(-1) at 20°C, respectively. A hybrid recombinant Hb D was prepared with recombinant α(D)-globin and native β-globin to give a Hb D tetramer (α(2)(D)β(2)). This rHb D undergoes decreased deoxygenation-dependent self-association compared with the native Hb D. Residue glutamate 138 has previously been proposed to influence intertetramer interactions. Our results with recombinant Hb D show that Glu138 plays no role in deoxy Hb D intertetramer interactions.     

Mutational analysis of phenylalanine beta 85 in the valine beta 6 acceptor pocket during hemoglobin S polymerization.

Hemoglobin (Hb) S containing Leu, Ala, Thr, or Trp substitutions at beta 85 were made and expressed in yeast in an effort to evaluate the role of Phe-beta 85 in the acceptor pocket during polymerization of deoxy Hb S. The four Hb S variants have the same electrophoretic mobility as Hb S, and these beta 85 substitutions do not significantly affect heme-globin interactions and tetramer helix content. Hb S containing Trp-beta 85 had decreased oxygen affinity, whereas those with Leu-, Ala-, and Thr-beta 85 had increased oxygen affinity. All four supersaturated beta 85 variants polymerized with a delay time as does deoxy Hb S. This is in contrast to deoxy Hb S containing Phe-beta 88, Ala-beta 88, Glu-beta 88, or Glu-beta 85, which polymerized with no clear delay time (Adachi K, Konitzer P, Paulraj CG, Surrey S, 1994, J Biol Chem 269:17477-17480; Adachi K, Reddy LR, Surrey S, 1994, J Biol Chem 269:31563-31566). Leu substitution at beta 85 accelerated deoxy Hb S polymerization, whereas Ala, Thr, or Trp substitution inhibited polymerization. The length of the delay time and total polymer formed for these beta 85 Hb S variants depended on hemoglobin concentration in the same fashion as for deoxy Hb S: the higher the concentration, the shorter the delay time and the more polymer formed. Critical concentrations required for polymerization of deoxy Hb SF veta 85L, Hb SF beta 85A, Hb SF beta 85T, and Hb SF beta 85W are 0.65-, 2.2-, 2.5- and 3-fold higher, respectively, than Hb S.(ABSTRACT TRUNCATED AT 250 WORDS)     

Functional consequences of ligand-linked dissociation in hemoglobin from the sea cucumber Molpadia arenicola.

It has been established that Molpadia hemoglobin tends to dissociate into subunits as oxygen is bound. The kinetics and equilibria of carbon monoxide and ethylisocyanide binding reported here show a dependence on protein concentration that supports the conclusion that the aggregated hemoglobin has a lower ligand affinity than the dissociated subunits. This is true for the isolated D-chain as well as for the unfractionated hemolysate that contains four distinct polypeptide chains (A-D). This indicates that even homopolymers of Molpadia hemoglobin have lower ligand affinity than the dissociated subunits. At high protein concentration hemolysates of Molpadia hemoglobin show slight cooperativity. The time course of ligand binding to the deoxy D-chain also suggests cooperative interactions. The low affinity of the aggregated state may have a different molecular explanation than in human hemoglobin where tetramers of identical subunits (as in Hb H) show high oxygen affinity. The absence of tyrosine and histidine at the C-terminal of the Molpadia D-chains also suggests a different stabilization of the low affinity deoxy state. An additional functional difference between Molpadia hemoglobin and human hemoglobin is that organic phosphates do not alter the ligand affinity of the sea cucumber hemoglobin.     

Fixation of aldehydic dextrans onto human deoxyhemoglobin.

A procedure commonly used to transform native adult human hemoglobin (Hb) into a physiological oxygen carrier consists of a pyridoxylation of the protein to lower its oxygen affinity, followed by its polymerization in the presence of glutaraldehyde, with or without further reduction, to increase its circulating half-life. This series of reactions yields derivatives presenting a great molecular heterogeneity that have to be fractionated for use in vivo. Hemoglobin derivatives with low oxygen affinity and a narrow distribution of molecular weights were obtained by linking a dextran polyaldehydic derivative to deoxyhemoglobin at pH 8. From oxygen-binding measurements carried out in the presence of inositolhexaphosphate, a strong effector of hemoglobin, it appeared that the allosteric site of hemoglobin was blocked, probably by crosslinking bonds, which stabilizes its deoxy structure. On the other hand, when the reaction was performed in the presence of inositolhexaphosphate, the resulting conjugates exhibited an oxygen affinity identical to that of unmodified hemoglobin. After treatment with NaBH4, the polymer-hemoglobin derivatives were stable and possessed a reversible oxygen-carrying capacity similar to that of blood. The conjugates prepared from oxyhemoglobin all possessed a lower P50 than native hemoglobin whatever the reaction conditions.     

Heme redox properties of S-nitrosated hemoglobin A0 and hemoglobin S: implications for interactions of nitric oxide with normal and sickle red blood cells

S-Nitrosated hemoglobin is remarkably stable and can be cycled between deoxy, oxygenated, or oxidized forms without significant loss of NO. Here we show that S-nitrosation of adult human hemoglobin (Hb A(0)) or sickle cell Hb (Hb S) results in an increased ease of anaerobic heme oxidation, while anions cause redox shifts in the opposite direction. The negatively charged groups of the cytoplasmic domain of Band 3 protein also produce an allosteric effect on S-nitrosated Hb. Formation and deoxygenation of a SNO-Hb/Band 3 protein assembly does not in itself cause NO release, even in the presence of glutathione; however, this assembly may play a role in the migration of NO from the red blood cells to other targets and may be linked to Heinz body formation. Studies of the anaerobic oxidation of Hb S revealed an altered redox potential relative to Hb A(0) that favors met-Hb formation and may therefore underlie the increased rate of autoxidation of Hb S under aerobic conditions, the increased formation of Heinz bodies in sickle cells, and the decreased lifetime of red cells containing Hb S. A model for the interrelationships between the deoxy, oxy, and met forms of Hb A(0) and Hb S, and their S-nitrosated counterparts, is presented.     

Tetramer-dimer equilibrium of oxyhemoglobin mutants determined from auto-oxidation rates.

One of the main difficulties with blood substitutes based on hemoglobin (Hb) solutions is the auto-oxidation of the hemes, a problem aggravated by the dimerization of Hb tetramers. We have employed a method to study the oxyHb tetramer-dimer equilibrium based on the rate of auto-oxidation as a function of protein concentration. The 16-fold difference in dimer and tetramer auto-oxidation rates (in 20 mM phosphate buffer at pH 7.0, 37 degrees C) was exploited to determine the fraction dimer. The results show a transition of the auto-oxidation rate from low to high protein concentrations, allowing the determination of the tetramer-dimer dissociation coefficient K4,2 = [Dimer] 2/[Tetramer]. A 14-fold increase in K4,2 was observed for addition of 10 mM of the allosteric effector inositol hexaphosphate (IHP). Recombinant hemoglobins (rHb) were genetically engineered to obtain Hb with a lower oxygen affinity than native Hb (Hb A). The rHb alpha2beta2 [(C7) F41Y/(G4) N102Y] shows a fivefold increase in K4,2 at pH 7.0, 37 degrees C. An atmosphere of pure oxygen is necessary in this case to insure fully oxygenated Hb. When this condition is satisfied, this method provides an efficient technique to characterize both the tetramer-dimer equilibrium and the auto-oxidation rates of various oxyHb. For low oxygen affinity Hb equilibrated under air, the presence of deoxy subunits accelerates the auto-oxidation. Although a full analysis is complicated, the auto-oxidation studies for air equilibrated samples are more relevant to the development of a blood substitute based on Hb solutions. The double mutants, rHb alpha2beta2 [(C7) F41Y/(G4) N102A] and rHb alpha2beta2 [(C7) F41Y/(E10) K66T], show a lower oxygen affinity and a higher rate of oxidation than Hb A. Simulations of the auto-oxidation rate versus Hb concentration indicate that very high protein concentrations are required to observe the tetramer auto-oxidation rate. Because the dimers oxidize much more rapidly, even a small fraction dimer will influence the observed oxidation rate.     

pH-dependent Soret difference spectra of the deoxy and carbonmonoxy forms of human hemoglobin and its derivatives.

The transition from deoxy to oxystructure of hemoglobin A (Hb) is accompanied by the breaking of the salt bridges formed by C-terminal residues in deoxy-Hb. This, in turn, changes the state of the heme. The switch between these different allosteric forms can be followed by changes in the optical absorbance spectra (Perutz, M. F., Ladner, J. E., Simon, S. R., and Ho, C. (1974), Biochemistry 13, 2163). Using difference spectroscopy in the soret region, pH-dependent spectral changes of Hb and its derivatives (carbamylated at both the alpha-NH2 groups, alpha2cbeta2c; N-ethylsuccinimide hemoglobin, NES-Hb) in their deoxy and carbonmonoxy forms were measured. From these measurements, the pK values of histidine-146beta and valine-1alpha in deoxy-Hb were determined to be 8.6 +/- 0.2 and 7.7 +/- 0.1, respectively. In carbonmonoxy-Hb a pK value of 6.3 +/- 0.1 was found.     

Proton nuclear magnetic resonance studies of hemoglobin Providence (beta82EF6 Lys replaced by Asn or Asp): a residue involved in anion binding

High-resolution proton nuclear magnetic resonance studies of hemoglobins Providence-Asn (beta82EF6 Lys replaced by Asn) and Providence-Asp (beta82EF6 Lys replaced by Asp) show that different amino acid substitutions at the same position in the hemoglobin molecule have different effects on the structure of the protein molecule. Hemoglobin Providence-Asp appears to be in a low-affinity tertiary structure in both the deoxy and carbonmonoxy forms. Deoxyhemoglobin Providence-Asn has its beta heme resonance shifted downfield slightly from its position in normal adult hemoglobin; however, the tertiary structures of the heme pocket of hemoglobins A and Providence-Asn are very similar when both proteins are in the carbonmonoxy form. These results are consistent with the oxygen equilibrium measurements of Bonaventura, J., et al. [(1976) J. Biol. Chem. 251, 7563] which show that both Hb Providence-Asn and Hb Providence-Asp have oxygen affinities lower than normal adult hemoglobin, with Hb Providence-Asp having the lowest. Our studies of the effects of sodium chloride on the hyperfine shifted proton resonances of deoxyhemoglobins A, Providence-Asn, and Providence-Asp indicate that the beta82EF6 lysine is probably one, but not the only binding site for chloride ions.     

Study on the interaction between hemoglobin Izu (Macaca) and organic phosphates by spin labeling

ESR spectra of the carbonmonoxy, oxy, and deoxy derivatives of hemoglobin Izu [Hb Izu (Macaca): beta 83 (EF 7) Gly leads to Cys] labeled at cysteine beta 83 with maleimide spin label have been observed in the presence and absence of 2,3-diphosphoglycerate and inositol hexaphosphate. The tau c values obtained from the spectra indicated that inositol hexaphosphate binds to all the derivatives of Hb Izu, but 2,3-diphosphoglycerate only to the deoxy derivatives.     

Proton nuclear magnetic resonance studies of hemoglobin Malm?: implications of mutations at homologous positions of the alpha and beta chains.

The abnormal human hemoglobin Malm? (beta97FG4 His leads to Gln) has been studied and its properties are compared with those of normal adult hemoglobin A. The data presented here show that the ring-current shifted proton resonances of both HbCO and HbO2 Malm? are very different from the corresponding forms of Hb A. The hyperfine shifted proton resonances of deoxy-Hb Malm? do not differ drastically from those of deoxy-Hb A. This result, together with the finding that the exchangeable proton resonances of the deoxy form of the two hemoglobins are similar, suggests that unliganded Hb Malm? can assume a deoxy-like quaternary structure both in the absence and presence of organic phosphates We have also compared the properties of Hb Malm? with those of Hb Chesapeake (alpha92FG4 Arg leads to Leu). This allows us to study the properties of two abnormal human hemoglobins with mutations at homologous positions of the alpha and beta chains in the three-dimenstional structure of the hemoglobin molecule. Our present results suggest that the mutaion at betaFG4 has its greatest effect on the teritiary structure of the heme pocket of the liganded forms of the hemoglobin while the mutation at alphaFG4 alters the deoxy structure of the hemoglogin molecule but does not alter the teriary structure of the heme pockets of the liganded form of the hemoglobin molecule. Both hemoglobins undergo a transition from the deoxy (T) to the oxy (R) quaternary structure upon ligation. The abnormally high oxygen affinities and low cooperativities of these two hemoglobins must therefore be due to either the structural differences which we have observed and/or to an altered transition between the T and R structures.     

Hemoglobin Saint Mandé [beta 102 (G4) Asn----Tyr]. Functional studies and structural modeling reveal an altered T state

Oxygen equilibrium studies of purified hemoglobin Saint Mandé (Hb SM) [beta 102 (G4) Asn----Tyr] reveal a decreased oxygen affinity and cooperativity but to a lesser extent than found for Hb Kansas (beta 102 Thr). The low affinity of Hb SM depends on environmental conditions: eliminating chloride or raising the pH greatly elevated the ratio of p50 of Hb SM to that of Hb A. The alkaline Bohr effect is reduced by about 40%. The effects of anions (chloride, organophosphates) binding to deoxy Hb SM are also reduced. These data indicate that the functional properties of Hb SM are intermediary between Hb A and Hb Kansas. In addition, molecular graphics modeling of Hb SM in the oxy and deoxy structures indicate the possibility of a new hydrogen bond in the T state between beta(1)102 Tyr and alpha(2)42 Tyr. Stabilisation of the T state in this manner is a plausible explanation for several of the effects observed.     

Functional consequences of ligand-linked dissociation in hemoglobin from the sea cucumber molpadia arenicola

It has been established that Molpadia hemoglobin tends to dissociate into subunits as oxygen is bound. The kinetics and equilibria of carbon monoxide and ehtylisocyanide binding reported here show a dependence on protein concentration that supports the conclusions that the aggregated hemoglobin has a lower ligand affinity than the dissociated subunits. This is true for the isolated D-chain as well as for the unfractionated hemolysate that contains four distinct polypeptide chains (A-D). This indicates that even homopolymers of Molpadia hemoglobin have lower ligand affinity than the dissociated subunits. At high protein concentration hemolysates of Molpadia hemoglobin show slight cooperativity. The time course of ligand binding to the deoxy D-chain also suggests cooperative interactions, The low affinity of the aggregated state may have a different molecular explanation than in human hemoglobin were tetramers of identical subunits (as in Hb H) show high oxygen affinity. The absence of tyrosine and histidine at the C-tremini of the Molpadia D-chains also suggests a different stabilization of the low affinity deoxy state. An additional functional difference between Molpadia hemoglobin and human hemoglobin is that organic phosphate do not alter the ligand affinity of the sea cucumber hemoglobin.     

The effect of crosslinking by bis(3,5-dibromosalicyl) fumarate on the autoxidation of hemoglobin

Bis(3,5-dibromosalicyl) fumarate was used to crosslink hemoglobin both in the oxy and deoxy states. This double headed diaspirin was known to crosslink oxy Hb A selectively between Lys 82 beta 1 and Lys 82 beta 2 (Walder, J. A., et al. (1979) Biochemistry 18, 4265) and deoxy Hb A between Lys 99 alpha 1 and Lys 99 alpha 2 (Chatterjee R. Y., et al. (1986) J. Biol. Chem. 261, 9929). The autoxidation at 37 degrees C of oxy alpha 99 crosslinked hemoglobin was found to be 1.8 times as fast as that of Hb A while that of the oxy beta 82 crosslinked hemoglobin was only 1.2 times as fast. After 5 hours the formation of methemoglobin in the alpha crosslinked Hb A is 21.3% compared to 10.8% in beta crosslinked Hb A and 6.4% in Hb A. These results may effect the proposed use of alpha 99 crosslinked hemoglobin as a blood substitute by demonstrating the need for protection from autoxidation during storage.